linux/net/ipv4/tcp.c

4778 lines
130 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-or-later
/*
* INET An implementation of the TCP/IP protocol suite for the LINUX
* operating system. INET is implemented using the BSD Socket
* interface as the means of communication with the user level.
*
* Implementation of the Transmission Control Protocol(TCP).
*
* Authors: Ross Biro
* Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
* Mark Evans, <evansmp@uhura.aston.ac.uk>
* Corey Minyard <wf-rch!minyard@relay.EU.net>
* Florian La Roche, <flla@stud.uni-sb.de>
* Charles Hedrick, <hedrick@klinzhai.rutgers.edu>
* Linus Torvalds, <torvalds@cs.helsinki.fi>
* Alan Cox, <gw4pts@gw4pts.ampr.org>
* Matthew Dillon, <dillon@apollo.west.oic.com>
* Arnt Gulbrandsen, <agulbra@nvg.unit.no>
* Jorge Cwik, <jorge@laser.satlink.net>
*
* Fixes:
* Alan Cox : Numerous verify_area() calls
* Alan Cox : Set the ACK bit on a reset
* Alan Cox : Stopped it crashing if it closed while
* sk->inuse=1 and was trying to connect
* (tcp_err()).
* Alan Cox : All icmp error handling was broken
* pointers passed where wrong and the
* socket was looked up backwards. Nobody
* tested any icmp error code obviously.
* Alan Cox : tcp_err() now handled properly. It
* wakes people on errors. poll
* behaves and the icmp error race
* has gone by moving it into sock.c
* Alan Cox : tcp_send_reset() fixed to work for
* everything not just packets for
* unknown sockets.
* Alan Cox : tcp option processing.
* Alan Cox : Reset tweaked (still not 100%) [Had
* syn rule wrong]
* Herp Rosmanith : More reset fixes
* Alan Cox : No longer acks invalid rst frames.
* Acking any kind of RST is right out.
* Alan Cox : Sets an ignore me flag on an rst
* receive otherwise odd bits of prattle
* escape still
* Alan Cox : Fixed another acking RST frame bug.
* Should stop LAN workplace lockups.
* Alan Cox : Some tidyups using the new skb list
* facilities
* Alan Cox : sk->keepopen now seems to work
* Alan Cox : Pulls options out correctly on accepts
* Alan Cox : Fixed assorted sk->rqueue->next errors
* Alan Cox : PSH doesn't end a TCP read. Switched a
* bit to skb ops.
* Alan Cox : Tidied tcp_data to avoid a potential
* nasty.
* Alan Cox : Added some better commenting, as the
* tcp is hard to follow
* Alan Cox : Removed incorrect check for 20 * psh
* Michael O'Reilly : ack < copied bug fix.
* Johannes Stille : Misc tcp fixes (not all in yet).
* Alan Cox : FIN with no memory -> CRASH
* Alan Cox : Added socket option proto entries.
* Also added awareness of them to accept.
* Alan Cox : Added TCP options (SOL_TCP)
* Alan Cox : Switched wakeup calls to callbacks,
* so the kernel can layer network
* sockets.
* Alan Cox : Use ip_tos/ip_ttl settings.
* Alan Cox : Handle FIN (more) properly (we hope).
* Alan Cox : RST frames sent on unsynchronised
* state ack error.
* Alan Cox : Put in missing check for SYN bit.
* Alan Cox : Added tcp_select_window() aka NET2E
* window non shrink trick.
* Alan Cox : Added a couple of small NET2E timer
* fixes
* Charles Hedrick : TCP fixes
* Toomas Tamm : TCP window fixes
* Alan Cox : Small URG fix to rlogin ^C ack fight
* Charles Hedrick : Rewrote most of it to actually work
* Linus : Rewrote tcp_read() and URG handling
* completely
* Gerhard Koerting: Fixed some missing timer handling
* Matthew Dillon : Reworked TCP machine states as per RFC
* Gerhard Koerting: PC/TCP workarounds
* Adam Caldwell : Assorted timer/timing errors
* Matthew Dillon : Fixed another RST bug
* Alan Cox : Move to kernel side addressing changes.
* Alan Cox : Beginning work on TCP fastpathing
* (not yet usable)
* Arnt Gulbrandsen: Turbocharged tcp_check() routine.
* Alan Cox : TCP fast path debugging
* Alan Cox : Window clamping
* Michael Riepe : Bug in tcp_check()
* Matt Dillon : More TCP improvements and RST bug fixes
* Matt Dillon : Yet more small nasties remove from the
* TCP code (Be very nice to this man if
* tcp finally works 100%) 8)
* Alan Cox : BSD accept semantics.
* Alan Cox : Reset on closedown bug.
* Peter De Schrijver : ENOTCONN check missing in tcp_sendto().
* Michael Pall : Handle poll() after URG properly in
* all cases.
* Michael Pall : Undo the last fix in tcp_read_urg()
* (multi URG PUSH broke rlogin).
* Michael Pall : Fix the multi URG PUSH problem in
* tcp_readable(), poll() after URG
* works now.
* Michael Pall : recv(...,MSG_OOB) never blocks in the
* BSD api.
* Alan Cox : Changed the semantics of sk->socket to
* fix a race and a signal problem with
* accept() and async I/O.
* Alan Cox : Relaxed the rules on tcp_sendto().
* Yury Shevchuk : Really fixed accept() blocking problem.
* Craig I. Hagan : Allow for BSD compatible TIME_WAIT for
* clients/servers which listen in on
* fixed ports.
* Alan Cox : Cleaned the above up and shrank it to
* a sensible code size.
* Alan Cox : Self connect lockup fix.
* Alan Cox : No connect to multicast.
* Ross Biro : Close unaccepted children on master
* socket close.
* Alan Cox : Reset tracing code.
* Alan Cox : Spurious resets on shutdown.
* Alan Cox : Giant 15 minute/60 second timer error
* Alan Cox : Small whoops in polling before an
* accept.
* Alan Cox : Kept the state trace facility since
* it's handy for debugging.
* Alan Cox : More reset handler fixes.
* Alan Cox : Started rewriting the code based on
* the RFC's for other useful protocol
* references see: Comer, KA9Q NOS, and
* for a reference on the difference
* between specifications and how BSD
* works see the 4.4lite source.
* A.N.Kuznetsov : Don't time wait on completion of tidy
* close.
* Linus Torvalds : Fin/Shutdown & copied_seq changes.
* Linus Torvalds : Fixed BSD port reuse to work first syn
* Alan Cox : Reimplemented timers as per the RFC
* and using multiple timers for sanity.
* Alan Cox : Small bug fixes, and a lot of new
* comments.
* Alan Cox : Fixed dual reader crash by locking
* the buffers (much like datagram.c)
* Alan Cox : Fixed stuck sockets in probe. A probe
* now gets fed up of retrying without
* (even a no space) answer.
* Alan Cox : Extracted closing code better
* Alan Cox : Fixed the closing state machine to
* resemble the RFC.
* Alan Cox : More 'per spec' fixes.
* Jorge Cwik : Even faster checksumming.
* Alan Cox : tcp_data() doesn't ack illegal PSH
* only frames. At least one pc tcp stack
* generates them.
* Alan Cox : Cache last socket.
* Alan Cox : Per route irtt.
* Matt Day : poll()->select() match BSD precisely on error
* Alan Cox : New buffers
* Marc Tamsky : Various sk->prot->retransmits and
* sk->retransmits misupdating fixed.
* Fixed tcp_write_timeout: stuck close,
* and TCP syn retries gets used now.
* Mark Yarvis : In tcp_read_wakeup(), don't send an
* ack if state is TCP_CLOSED.
* Alan Cox : Look up device on a retransmit - routes may
* change. Doesn't yet cope with MSS shrink right
* but it's a start!
* Marc Tamsky : Closing in closing fixes.
* Mike Shaver : RFC1122 verifications.
* Alan Cox : rcv_saddr errors.
* Alan Cox : Block double connect().
* Alan Cox : Small hooks for enSKIP.
* Alexey Kuznetsov: Path MTU discovery.
* Alan Cox : Support soft errors.
* Alan Cox : Fix MTU discovery pathological case
* when the remote claims no mtu!
* Marc Tamsky : TCP_CLOSE fix.
* Colin (G3TNE) : Send a reset on syn ack replies in
* window but wrong (fixes NT lpd problems)
* Pedro Roque : Better TCP window handling, delayed ack.
* Joerg Reuter : No modification of locked buffers in
* tcp_do_retransmit()
* Eric Schenk : Changed receiver side silly window
* avoidance algorithm to BSD style
* algorithm. This doubles throughput
* against machines running Solaris,
* and seems to result in general
* improvement.
* Stefan Magdalinski : adjusted tcp_readable() to fix FIONREAD
* Willy Konynenberg : Transparent proxying support.
* Mike McLagan : Routing by source
* Keith Owens : Do proper merging with partial SKB's in
* tcp_do_sendmsg to avoid burstiness.
* Eric Schenk : Fix fast close down bug with
* shutdown() followed by close().
* Andi Kleen : Make poll agree with SIGIO
* Salvatore Sanfilippo : Support SO_LINGER with linger == 1 and
* lingertime == 0 (RFC 793 ABORT Call)
* Hirokazu Takahashi : Use copy_from_user() instead of
* csum_and_copy_from_user() if possible.
*
* Description of States:
*
* TCP_SYN_SENT sent a connection request, waiting for ack
*
* TCP_SYN_RECV received a connection request, sent ack,
* waiting for final ack in three-way handshake.
*
* TCP_ESTABLISHED connection established
*
* TCP_FIN_WAIT1 our side has shutdown, waiting to complete
* transmission of remaining buffered data
*
* TCP_FIN_WAIT2 all buffered data sent, waiting for remote
* to shutdown
*
* TCP_CLOSING both sides have shutdown but we still have
* data we have to finish sending
*
* TCP_TIME_WAIT timeout to catch resent junk before entering
* closed, can only be entered from FIN_WAIT2
* or CLOSING. Required because the other end
* may not have gotten our last ACK causing it
* to retransmit the data packet (which we ignore)
*
* TCP_CLOSE_WAIT remote side has shutdown and is waiting for
* us to finish writing our data and to shutdown
* (we have to close() to move on to LAST_ACK)
*
* TCP_LAST_ACK out side has shutdown after remote has
* shutdown. There may still be data in our
* buffer that we have to finish sending
*
* TCP_CLOSE socket is finished
*/
#define pr_fmt(fmt) "TCP: " fmt
#include <crypto/hash.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/types.h>
#include <linux/fcntl.h>
#include <linux/poll.h>
#include <linux/inet_diag.h>
#include <linux/init.h>
#include <linux/fs.h>
#include <linux/skbuff.h>
#include <linux/scatterlist.h>
#include <linux/splice.h>
#include <linux/net.h>
#include <linux/socket.h>
#include <linux/random.h>
mm: remove include/linux/bootmem.h Move remaining definitions and declarations from include/linux/bootmem.h into include/linux/memblock.h and remove the redundant header. The includes were replaced with the semantic patch below and then semi-automated removal of duplicated '#include <linux/memblock.h> @@ @@ - #include <linux/bootmem.h> + #include <linux/memblock.h> [sfr@canb.auug.org.au: dma-direct: fix up for the removal of linux/bootmem.h] Link: http://lkml.kernel.org/r/20181002185342.133d1680@canb.auug.org.au [sfr@canb.auug.org.au: powerpc: fix up for removal of linux/bootmem.h] Link: http://lkml.kernel.org/r/20181005161406.73ef8727@canb.auug.org.au [sfr@canb.auug.org.au: x86/kaslr, ACPI/NUMA: fix for linux/bootmem.h removal] Link: http://lkml.kernel.org/r/20181008190341.5e396491@canb.auug.org.au Link: http://lkml.kernel.org/r/1536927045-23536-30-git-send-email-rppt@linux.vnet.ibm.com Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com> Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Acked-by: Michal Hocko <mhocko@suse.com> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Chris Zankel <chris@zankel.net> Cc: "David S. Miller" <davem@davemloft.net> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Greentime Hu <green.hu@gmail.com> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Guan Xuetao <gxt@pku.edu.cn> Cc: Ingo Molnar <mingo@redhat.com> Cc: "James E.J. Bottomley" <jejb@parisc-linux.org> Cc: Jonas Bonn <jonas@southpole.se> Cc: Jonathan Corbet <corbet@lwn.net> Cc: Ley Foon Tan <lftan@altera.com> Cc: Mark Salter <msalter@redhat.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Matt Turner <mattst88@gmail.com> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Michal Simek <monstr@monstr.eu> Cc: Palmer Dabbelt <palmer@sifive.com> Cc: Paul Burton <paul.burton@mips.com> Cc: Richard Kuo <rkuo@codeaurora.org> Cc: Richard Weinberger <richard@nod.at> Cc: Rich Felker <dalias@libc.org> Cc: Russell King <linux@armlinux.org.uk> Cc: Serge Semin <fancer.lancer@gmail.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Tony Luck <tony.luck@intel.com> Cc: Vineet Gupta <vgupta@synopsys.com> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-10-31 06:09:49 +08:00
#include <linux/memblock.h>
#include <linux/highmem.h>
#include <linux/cache.h>
#include <linux/err.h>
#include <linux/time.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/slab.h>
#include <linux/errqueue.h>
#include <linux/static_key.h>
bpf: net: Emit anonymous enum with BPF_TCP_CLOSE value explicitly The selftest failed to compile with clang-built bpf-next. Adding LLVM=1 to your vmlinux and selftest build will use clang. The error message is: progs/test_sk_storage_tracing.c:38:18: error: use of undeclared identifier 'BPF_TCP_CLOSE' if (newstate == BPF_TCP_CLOSE) ^ 1 error generated. make: *** [Makefile:423: /bpf-next/tools/testing/selftests/bpf/test_sk_storage_tracing.o] Error 1 The reason for the failure is that BPF_TCP_CLOSE, a value of an anonymous enum defined in uapi bpf.h, is not defined in vmlinux.h. gcc does not have this problem. Since vmlinux.h is derived from BTF which is derived from vmlinux DWARF, that means gcc-produced vmlinux DWARF has BPF_TCP_CLOSE while llvm-produced vmlinux DWARF does not have. BPF_TCP_CLOSE is referenced in net/ipv4/tcp.c as BUILD_BUG_ON((int)BPF_TCP_CLOSE != (int)TCP_CLOSE); The following test mimics the above BUILD_BUG_ON, preprocessed with clang compiler, and shows gcc DWARF contains BPF_TCP_CLOSE while llvm DWARF does not. $ cat t.c enum { BPF_TCP_ESTABLISHED = 1, BPF_TCP_CLOSE = 7, }; enum { TCP_ESTABLISHED = 1, TCP_CLOSE = 7, }; int test() { do { extern void __compiletime_assert_767(void) ; if ((int)BPF_TCP_CLOSE != (int)TCP_CLOSE) __compiletime_assert_767(); } while (0); return 0; } $ clang t.c -O2 -c -g && llvm-dwarfdump t.o | grep BPF_TCP_CLOSE $ gcc t.c -O2 -c -g && llvm-dwarfdump t.o | grep BPF_TCP_CLOSE DW_AT_name ("BPF_TCP_CLOSE") Further checking clang code find clang actually tried to evaluate condition at compile time. If it is definitely true/false, it will perform optimization and the whole if condition will be removed before generating IR/debuginfo. This patch explicited add an expression after the above mentioned BUILD_BUG_ON in net/ipv4/tcp.c like (void)BPF_TCP_ESTABLISHED to enable generation of debuginfo for the anonymous enum which also includes BPF_TCP_CLOSE. Signed-off-by: Yonghong Song <yhs@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20210317174132.589276-1-yhs@fb.com
2021-03-18 01:41:32 +08:00
#include <linux/btf.h>
#include <net/icmp.h>
#include <net/inet_common.h>
#include <net/tcp.h>
#include <net/mptcp.h>
#include <net/xfrm.h>
#include <net/ip.h>
#include <net/sock.h>
#include <linux/uaccess.h>
#include <asm/ioctls.h>
#include <net/busy_poll.h>
/* Track pending CMSGs. */
enum {
TCP_CMSG_INQ = 1,
TCP_CMSG_TS = 2
};
tcp: switch orphan_count to bare per-cpu counters Use of percpu_counter structure to track count of orphaned sockets is causing problems on modern hosts with 256 cpus or more. Stefan Bach reported a serious spinlock contention in real workloads, that I was able to reproduce with a netfilter rule dropping incoming FIN packets. 53.56% server [kernel.kallsyms] [k] queued_spin_lock_slowpath | ---queued_spin_lock_slowpath | --53.51%--_raw_spin_lock_irqsave | --53.51%--__percpu_counter_sum tcp_check_oom | |--39.03%--__tcp_close | tcp_close | inet_release | inet6_release | sock_close | __fput | ____fput | task_work_run | exit_to_usermode_loop | do_syscall_64 | entry_SYSCALL_64_after_hwframe | __GI___libc_close | --14.48%--tcp_out_of_resources tcp_write_timeout tcp_retransmit_timer tcp_write_timer_handler tcp_write_timer call_timer_fn expire_timers __run_timers run_timer_softirq __softirqentry_text_start As explained in commit cf86a086a180 ("net/dst: use a smaller percpu_counter batch for dst entries accounting"), default batch size is too big for the default value of tcp_max_orphans (262144). But even if we reduce batch sizes, there would still be cases where the estimated count of orphans is beyond the limit, and where tcp_too_many_orphans() has to call the expensive percpu_counter_sum_positive(). One solution is to use plain per-cpu counters, and have a timer to periodically refresh this cache. Updating this cache every 100ms seems about right, tcp pressure state is not radically changing over shorter periods. percpu_counter was nice 15 years ago while hosts had less than 16 cpus, not anymore by current standards. v2: Fix the build issue for CONFIG_CRYPTO_DEV_CHELSIO_TLS=m, reported by kernel test robot <lkp@intel.com> Remove unused socket argument from tcp_too_many_orphans() Fixes: dd24c00191d5 ("net: Use a percpu_counter for orphan_count") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Stefan Bach <sfb@google.com> Cc: Neal Cardwell <ncardwell@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-10-14 21:41:26 +08:00
DEFINE_PER_CPU(unsigned int, tcp_orphan_count);
EXPORT_PER_CPU_SYMBOL_GPL(tcp_orphan_count);
long sysctl_tcp_mem[3] __read_mostly;
EXPORT_SYMBOL(sysctl_tcp_mem);
atomic_long_t tcp_memory_allocated ____cacheline_aligned_in_smp; /* Current allocated memory. */
EXPORT_SYMBOL(tcp_memory_allocated);
DEFINE_PER_CPU(int, tcp_memory_per_cpu_fw_alloc);
EXPORT_PER_CPU_SYMBOL_GPL(tcp_memory_per_cpu_fw_alloc);
#if IS_ENABLED(CONFIG_SMC)
DEFINE_STATIC_KEY_FALSE(tcp_have_smc);
EXPORT_SYMBOL(tcp_have_smc);
#endif
/*
* Current number of TCP sockets.
*/
struct percpu_counter tcp_sockets_allocated ____cacheline_aligned_in_smp;
EXPORT_SYMBOL(tcp_sockets_allocated);
/*
* TCP splice context
*/
struct tcp_splice_state {
struct pipe_inode_info *pipe;
size_t len;
unsigned int flags;
};
/*
* Pressure flag: try to collapse.
* Technical note: it is used by multiple contexts non atomically.
[NET] CORE: Introducing new memory accounting interface. This patch introduces new memory accounting functions for each network protocol. Most of them are renamed from memory accounting functions for stream protocols. At the same time, some stream memory accounting functions are removed since other functions do same thing. Renaming: sk_stream_free_skb() -> sk_wmem_free_skb() __sk_stream_mem_reclaim() -> __sk_mem_reclaim() sk_stream_mem_reclaim() -> sk_mem_reclaim() sk_stream_mem_schedule -> __sk_mem_schedule() sk_stream_pages() -> sk_mem_pages() sk_stream_rmem_schedule() -> sk_rmem_schedule() sk_stream_wmem_schedule() -> sk_wmem_schedule() sk_charge_skb() -> sk_mem_charge() Removeing sk_stream_rfree(): consolidates into sock_rfree() sk_stream_set_owner_r(): consolidates into skb_set_owner_r() sk_stream_mem_schedule() The following functions are added. sk_has_account(): check if the protocol supports accounting sk_mem_uncharge(): do the opposite of sk_mem_charge() In addition, to achieve consolidation, updating sk_wmem_queued is removed from sk_mem_charge(). Next, to consolidate memory accounting functions, this patch adds memory accounting calls to network core functions. Moreover, present memory accounting call is renamed to new accounting call. Finally we replace present memory accounting calls with new interface in TCP and SCTP. Signed-off-by: Takahiro Yasui <tyasui@redhat.com> Signed-off-by: Hideo Aoki <haoki@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-12-31 16:11:19 +08:00
* All the __sk_mem_schedule() is of this nature: accounting
* is strict, actions are advisory and have some latency.
*/
unsigned long tcp_memory_pressure __read_mostly;
EXPORT_SYMBOL_GPL(tcp_memory_pressure);
void tcp_enter_memory_pressure(struct sock *sk)
{
unsigned long val;
if (READ_ONCE(tcp_memory_pressure))
return;
val = jiffies;
if (!val)
val--;
if (!cmpxchg(&tcp_memory_pressure, 0, val))
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMEMORYPRESSURES);
}
EXPORT_SYMBOL_GPL(tcp_enter_memory_pressure);
void tcp_leave_memory_pressure(struct sock *sk)
{
unsigned long val;
if (!READ_ONCE(tcp_memory_pressure))
return;
val = xchg(&tcp_memory_pressure, 0);
if (val)
NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPMEMORYPRESSURESCHRONO,
jiffies_to_msecs(jiffies - val));
}
EXPORT_SYMBOL_GPL(tcp_leave_memory_pressure);
/* Convert seconds to retransmits based on initial and max timeout */
static u8 secs_to_retrans(int seconds, int timeout, int rto_max)
{
u8 res = 0;
if (seconds > 0) {
int period = timeout;
res = 1;
while (seconds > period && res < 255) {
res++;
timeout <<= 1;
if (timeout > rto_max)
timeout = rto_max;
period += timeout;
}
}
return res;
}
/* Convert retransmits to seconds based on initial and max timeout */
static int retrans_to_secs(u8 retrans, int timeout, int rto_max)
{
int period = 0;
if (retrans > 0) {
period = timeout;
while (--retrans) {
timeout <<= 1;
if (timeout > rto_max)
timeout = rto_max;
period += timeout;
}
}
return period;
}
static u64 tcp_compute_delivery_rate(const struct tcp_sock *tp)
{
u32 rate = READ_ONCE(tp->rate_delivered);
u32 intv = READ_ONCE(tp->rate_interval_us);
u64 rate64 = 0;
if (rate && intv) {
rate64 = (u64)rate * tp->mss_cache * USEC_PER_SEC;
do_div(rate64, intv);
}
return rate64;
}
/* Address-family independent initialization for a tcp_sock.
*
* NOTE: A lot of things set to zero explicitly by call to
* sk_alloc() so need not be done here.
*/
void tcp_init_sock(struct sock *sk)
{
struct inet_connection_sock *icsk = inet_csk(sk);
struct tcp_sock *tp = tcp_sk(sk);
2016-09-08 05:49:28 +08:00
tp->out_of_order_queue = RB_ROOT;
tcp: implement rb-tree based retransmit queue Using a linear list to store all skbs in write queue has been okay for quite a while : O(N) is not too bad when N < 500. Things get messy when N is the order of 100,000 : Modern TCP stacks want 10Gbit+ of throughput even with 200 ms RTT flows. 40 ns per cache line miss means a full scan can use 4 ms, blowing away CPU caches. SACK processing often can use various hints to avoid parsing whole retransmit queue. But with high packet losses and/or high reordering, hints no longer work. Sender has to process thousands of unfriendly SACK, accumulating a huge socket backlog, burning a cpu and massively dropping packets. Using an rb-tree for retransmit queue has been avoided for years because it added complexity and overhead, but now is the time to be more resistant and say no to quadratic behavior. 1) RTX queue is no longer part of the write queue : already sent skbs are stored in one rb-tree. 2) Since reaching the head of write queue no longer needs sk->sk_send_head, we added an union of sk_send_head and tcp_rtx_queue Tested: On receiver : netem on ingress : delay 150ms 200us loss 1 GRO disabled to force stress and SACK storms. for f in `seq 1 10` do ./netperf -H lpaa6 -l30 -- -K bbr -o THROUGHPUT|tail -1 done | awk '{print $0} {sum += $0} END {printf "%7u\n",sum}' Before patch : 323.87 351.48 339.59 338.62 306.72 204.07 304.93 291.88 202.47 176.88 2840 After patch: 1700.83 2207.98 2070.17 1544.26 2114.76 2124.89 1693.14 1080.91 2216.82 1299.94 18053 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-06 13:21:27 +08:00
sk->tcp_rtx_queue = RB_ROOT;
tcp_init_xmit_timers(sk);
tcp: TCP Small Queues This introduce TSQ (TCP Small Queues) TSQ goal is to reduce number of TCP packets in xmit queues (qdisc & device queues), to reduce RTT and cwnd bias, part of the bufferbloat problem. sk->sk_wmem_alloc not allowed to grow above a given limit, allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a given time. TSO packets are sized/capped to half the limit, so that we have two TSO packets in flight, allowing better bandwidth use. As a side effect, setting the limit to 40000 automatically reduces the standard gso max limit (65536) to 40000/2 : It can help to reduce latencies of high prio packets, having smaller TSO packets. This means we divert sock_wfree() to a tcp_wfree() handler, to queue/send following frames when skb_orphan() [2] is called for the already queued skbs. Results on my dev machines (tg3/ixgbe nics) are really impressive, using standard pfifo_fast, and with or without TSO/GSO. Without reduction of nominal bandwidth, we have reduction of buffering per bulk sender : < 1ms on Gbit (instead of 50ms with TSO) < 8ms on 100Mbit (instead of 132 ms) I no longer have 4 MBytes backlogged in qdisc by a single netperf session, and both side socket autotuning no longer use 4 Mbytes. As skb destructor cannot restart xmit itself ( as qdisc lock might be taken at this point ), we delegate the work to a tasklet. We use one tasklest per cpu for performance reasons. If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag. This flag is tested in a new protocol method called from release_sock(), to eventually send new segments. [1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable [2] skb_orphan() is usually called at TX completion time, but some drivers call it in their start_xmit() handler. These drivers should at least use BQL, or else a single TCP session can still fill the whole NIC TX ring, since TSQ will have no effect. Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Dave Taht <dave.taht@bufferbloat.net> Cc: Tom Herbert <therbert@google.com> Cc: Matt Mathis <mattmathis@google.com> Cc: Yuchung Cheng <ycheng@google.com> Cc: Nandita Dukkipati <nanditad@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
INIT_LIST_HEAD(&tp->tsq_node);
tcp: new list for sent but unacked skbs for RACK recovery This patch adds a new queue (list) that tracks the sent but not yet acked or SACKed skbs for a TCP connection. The list is chronologically ordered by skb->skb_mstamp (the head is the oldest sent skb). This list will be used to optimize TCP Rack recovery, which checks an skb's timestamp to judge if it has been lost and needs to be retransmitted. Since TCP write queue is ordered by sequence instead of sent time, RACK has to scan over the write queue to catch all eligible packets to detect lost retransmission, and iterates through SACKed skbs repeatedly. Special cares for rare events: 1. TCP repair fakes skb transmission so the send queue needs adjusted 2. SACK reneging would require re-inserting SACKed skbs into the send queue. For now I believe it's not worth the complexity to make RACK work perfectly on SACK reneging, so we do nothing here. 3. Fast Open: currently for non-TFO, send-queue correctly queues the pure SYN packet. For TFO which queues a pure SYN and then a data packet, send-queue only queues the data packet but not the pure SYN due to the structure of TFO code. This is okay because the SYN receiver would never respond with a SACK on a missing SYN (i.e. SYN is never fast-retransmitted by SACK/RACK). In order to not grow sk_buff, we use an union for the new list and _skb_refdst/destructor fields. This is a bit complicated because we need to make sure _skb_refdst and destructor are properly zeroed before skb is cloned/copied at transmit, and before being freed. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-05 03:59:58 +08:00
INIT_LIST_HEAD(&tp->tsorted_sent_queue);
icsk->icsk_rto = TCP_TIMEOUT_INIT;
icsk->icsk_rto_min = TCP_RTO_MIN;
icsk->icsk_delack_max = TCP_DELACK_MAX;
tcp: switch rtt estimations to usec resolution Upcoming congestion controls for TCP require usec resolution for RTT estimations. Millisecond resolution is simply not enough these days. FQ/pacing in DC environments also require this change for finer control and removal of bimodal behavior due to the current hack in tcp_update_pacing_rate() for 'small rtt' TCP_CONG_RTT_STAMP is no longer needed. As Julian Anastasov pointed out, we need to keep user compatibility : tcp_metrics used to export RTT and RTTVAR in msec resolution, so we added RTT_US and RTTVAR_US. An iproute2 patch is needed to use the new attributes if provided by the kernel. In this example ss command displays a srtt of 32 usecs (10Gbit link) lpk51:~# ./ss -i dst lpk52 Netid State Recv-Q Send-Q Local Address:Port Peer Address:Port tcp ESTAB 0 1 10.246.11.51:42959 10.246.11.52:64614 cubic wscale:6,6 rto:201 rtt:0.032/0.001 ato:40 mss:1448 cwnd:10 send 3620.0Mbps pacing_rate 7240.0Mbps unacked:1 rcv_rtt:993 rcv_space:29559 Updated iproute2 ip command displays : lpk51:~# ./ip tcp_metrics | grep 10.246.11.52 10.246.11.52 age 561.914sec cwnd 10 rtt 274us rttvar 213us source 10.246.11.51 Old binary displays : lpk51:~# ip tcp_metrics | grep 10.246.11.52 10.246.11.52 age 561.914sec cwnd 10 rtt 250us rttvar 125us source 10.246.11.51 With help from Julian Anastasov, Stephen Hemminger and Yuchung Cheng Signed-off-by: Eric Dumazet <edumazet@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Cc: Stephen Hemminger <stephen@networkplumber.org> Cc: Yuchung Cheng <ycheng@google.com> Cc: Larry Brakmo <brakmo@google.com> Cc: Julian Anastasov <ja@ssi.bg> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-27 06:02:48 +08:00
tp->mdev_us = jiffies_to_usecs(TCP_TIMEOUT_INIT);
minmax_reset(&tp->rtt_min, tcp_jiffies32, ~0U);
/* So many TCP implementations out there (incorrectly) count the
* initial SYN frame in their delayed-ACK and congestion control
* algorithms that we must have the following bandaid to talk
* efficiently to them. -DaveM
*/
tcp_snd_cwnd_set(tp, TCP_INIT_CWND);
tcp: track application-limited rate samples This commit adds code to track whether the delivery rate represented by each rate_sample was limited by the application. Upon each transmit, we store in the is_app_limited field in the skb a boolean bit indicating whether there is a known "bubble in the pipe": a point in the rate sample interval where the sender was application-limited, and did not transmit even though the cwnd and pacing rate allowed it. This logic marks the flow app-limited on a write if *all* of the following are true: 1) There is less than 1 MSS of unsent data in the write queue available to transmit. 2) There is no packet in the sender's queues (e.g. in fq or the NIC tx queue). 3) The connection is not limited by cwnd. 4) There are no lost packets to retransmit. The tcp_rate_check_app_limited() code in tcp_rate.c determines whether the connection is application-limited at the moment. If the flow is application-limited, it sets the tp->app_limited field. If the flow is application-limited then that means there is effectively a "bubble" of silence in the pipe now, and this silence will be reflected in a lower bandwidth sample for any rate samples from now until we get an ACK indicating this bubble has exited the pipe: specifically, until we get an ACK for the next packet we transmit. When we send every skb we record in scb->tx.is_app_limited whether the resulting rate sample will be application-limited. The code in tcp_rate_gen() checks to see when it is safe to mark all known application-limited bubbles of silence as having exited the pipe. It does this by checking to see when the delivered count moves past the tp->app_limited marker. At this point it zeroes the tp->app_limited marker, as all known bubbles are out of the pipe. We make room for the tx.is_app_limited bit in the skb by borrowing a bit from the in_flight field used by NV to record the number of bytes in flight. The receive window in the TCP header is 16 bits, and the max receive window scaling shift factor is 14 (RFC 1323). So the max receive window offered by the TCP protocol is 2^(16+14) = 2^30. So we only need 30 bits for the tx.in_flight used by NV. Signed-off-by: Van Jacobson <vanj@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Nandita Dukkipati <nanditad@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-09-20 11:39:15 +08:00
/* There's a bubble in the pipe until at least the first ACK. */
tp->app_limited = ~0U;
tp->rate_app_limited = 1;
tcp: track application-limited rate samples This commit adds code to track whether the delivery rate represented by each rate_sample was limited by the application. Upon each transmit, we store in the is_app_limited field in the skb a boolean bit indicating whether there is a known "bubble in the pipe": a point in the rate sample interval where the sender was application-limited, and did not transmit even though the cwnd and pacing rate allowed it. This logic marks the flow app-limited on a write if *all* of the following are true: 1) There is less than 1 MSS of unsent data in the write queue available to transmit. 2) There is no packet in the sender's queues (e.g. in fq or the NIC tx queue). 3) The connection is not limited by cwnd. 4) There are no lost packets to retransmit. The tcp_rate_check_app_limited() code in tcp_rate.c determines whether the connection is application-limited at the moment. If the flow is application-limited, it sets the tp->app_limited field. If the flow is application-limited then that means there is effectively a "bubble" of silence in the pipe now, and this silence will be reflected in a lower bandwidth sample for any rate samples from now until we get an ACK indicating this bubble has exited the pipe: specifically, until we get an ACK for the next packet we transmit. When we send every skb we record in scb->tx.is_app_limited whether the resulting rate sample will be application-limited. The code in tcp_rate_gen() checks to see when it is safe to mark all known application-limited bubbles of silence as having exited the pipe. It does this by checking to see when the delivered count moves past the tp->app_limited marker. At this point it zeroes the tp->app_limited marker, as all known bubbles are out of the pipe. We make room for the tx.is_app_limited bit in the skb by borrowing a bit from the in_flight field used by NV to record the number of bytes in flight. The receive window in the TCP header is 16 bits, and the max receive window scaling shift factor is 14 (RFC 1323). So the max receive window offered by the TCP protocol is 2^(16+14) = 2^30. So we only need 30 bits for the tx.in_flight used by NV. Signed-off-by: Van Jacobson <vanj@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Nandita Dukkipati <nanditad@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-09-20 11:39:15 +08:00
/* See draft-stevens-tcpca-spec-01 for discussion of the
* initialization of these values.
*/
tp->snd_ssthresh = TCP_INFINITE_SSTHRESH;
tp->snd_cwnd_clamp = ~0;
tp->mss_cache = TCP_MSS_DEFAULT;
tp->reordering = READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_reordering);
tcp_assign_congestion_control(sk);
tp->tsoffset = 0;
tcp: higher throughput under reordering with adaptive RACK reordering wnd Currently TCP RACK loss detection does not work well if packets are being reordered beyond its static reordering window (min_rtt/4).Under such reordering it may falsely trigger loss recoveries and reduce TCP throughput significantly. This patch improves that by increasing and reducing the reordering window based on DSACK, which is now supported in major TCP implementations. It makes RACK's reo_wnd adaptive based on DSACK and no. of recoveries. - If DSACK is received, increment reo_wnd by min_rtt/4 (upper bounded by srtt), since there is possibility that spurious retransmission was due to reordering delay longer than reo_wnd. - Persist the current reo_wnd value for TCP_RACK_RECOVERY_THRESH (16) no. of successful recoveries (accounts for full DSACK-based loss recovery undo). After that, reset it to default (min_rtt/4). - At max, reo_wnd is incremented only once per rtt. So that the new DSACK on which we are reacting, is due to the spurious retx (approx) after the reo_wnd has been updated last time. - reo_wnd is tracked in terms of steps (of min_rtt/4), rather than absolute value to account for change in rtt. In our internal testing, we observed significant increase in throughput, in scenarios where reordering exceeds min_rtt/4 (previous static value). Signed-off-by: Priyaranjan Jha <priyarjha@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-11-04 07:38:48 +08:00
tp->rack.reo_wnd_steps = 1;
sk->sk_write_space = sk_stream_write_space;
sock_set_flag(sk, SOCK_USE_WRITE_QUEUE);
icsk->icsk_sync_mss = tcp_sync_mss;
WRITE_ONCE(sk->sk_sndbuf, READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_wmem[1]));
WRITE_ONCE(sk->sk_rcvbuf, READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_rmem[1]));
tcp_scaling_ratio_init(sk);
set_bit(SOCK_SUPPORT_ZC, &sk->sk_socket->flags);
sk_sockets_allocated_inc(sk);
}
EXPORT_SYMBOL(tcp_init_sock);
static void tcp_tx_timestamp(struct sock *sk, u16 tsflags)
net-timestamp: TCP timestamping TCP timestamping extends SO_TIMESTAMPING to bytestreams. Bytestreams do not have a 1:1 relationship between send() buffers and network packets. The feature interprets a send call on a bytestream as a request for a timestamp for the last byte in that send() buffer. The choice corresponds to a request for a timestamp when all bytes in the buffer have been sent. That assumption depends on in-order kernel transmission. This is the common case. That said, it is possible to construct a traffic shaping tree that would result in reordering. The guarantee is strong, then, but not ironclad. This implementation supports send and sendpages (splice). GSO replaces one large packet with multiple smaller packets. This patch also copies the option into the correct smaller packet. This patch does not yet support timestamping on data in an initial TCP Fast Open SYN, because that takes a very different data path. If ID generation in ee_data is enabled, bytestream timestamps return a byte offset, instead of the packet counter for datagrams. The implementation supports a single timestamp per packet. It silenty replaces requests for previous timestamps. To avoid missing tstamps, flush the tcp queue by disabling Nagle, cork and autocork. Missing tstamps can be detected by offset when the ee_data ID is enabled. Implementation details: - On GSO, the timestamping code can be included in the main loop. I moved it into its own loop to reduce the impact on the common case to a single branch. - To avoid leaking the absolute seqno to userspace, the offset returned in ee_data must always be relative. It is an offset between an skb and sk field. The first is always set (also for GSO & ACK). The second must also never be uninitialized. Only allow the ID option on sockets in the ESTABLISHED state, for which the seqno is available. Never reset it to zero (instead, move it to the current seqno when reenabling the option). Signed-off-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-08-05 10:11:49 +08:00
{
struct sk_buff *skb = tcp_write_queue_tail(sk);
if (tsflags && skb) {
struct skb_shared_info *shinfo = skb_shinfo(skb);
struct tcp_skb_cb *tcb = TCP_SKB_CB(skb);
net-timestamp: TCP timestamping TCP timestamping extends SO_TIMESTAMPING to bytestreams. Bytestreams do not have a 1:1 relationship between send() buffers and network packets. The feature interprets a send call on a bytestream as a request for a timestamp for the last byte in that send() buffer. The choice corresponds to a request for a timestamp when all bytes in the buffer have been sent. That assumption depends on in-order kernel transmission. This is the common case. That said, it is possible to construct a traffic shaping tree that would result in reordering. The guarantee is strong, then, but not ironclad. This implementation supports send and sendpages (splice). GSO replaces one large packet with multiple smaller packets. This patch also copies the option into the correct smaller packet. This patch does not yet support timestamping on data in an initial TCP Fast Open SYN, because that takes a very different data path. If ID generation in ee_data is enabled, bytestream timestamps return a byte offset, instead of the packet counter for datagrams. The implementation supports a single timestamp per packet. It silenty replaces requests for previous timestamps. To avoid missing tstamps, flush the tcp queue by disabling Nagle, cork and autocork. Missing tstamps can be detected by offset when the ee_data ID is enabled. Implementation details: - On GSO, the timestamping code can be included in the main loop. I moved it into its own loop to reduce the impact on the common case to a single branch. - To avoid leaking the absolute seqno to userspace, the offset returned in ee_data must always be relative. It is an offset between an skb and sk field. The first is always set (also for GSO & ACK). The second must also never be uninitialized. Only allow the ID option on sockets in the ESTABLISHED state, for which the seqno is available. Never reset it to zero (instead, move it to the current seqno when reenabling the option). Signed-off-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-08-05 10:11:49 +08:00
sock_tx_timestamp(sk, tsflags, &shinfo->tx_flags);
if (tsflags & SOF_TIMESTAMPING_TX_ACK)
tcb->txstamp_ack = 1;
if (tsflags & SOF_TIMESTAMPING_TX_RECORD_MASK)
shinfo->tskey = TCP_SKB_CB(skb)->seq + skb->len - 1;
}
net-timestamp: TCP timestamping TCP timestamping extends SO_TIMESTAMPING to bytestreams. Bytestreams do not have a 1:1 relationship between send() buffers and network packets. The feature interprets a send call on a bytestream as a request for a timestamp for the last byte in that send() buffer. The choice corresponds to a request for a timestamp when all bytes in the buffer have been sent. That assumption depends on in-order kernel transmission. This is the common case. That said, it is possible to construct a traffic shaping tree that would result in reordering. The guarantee is strong, then, but not ironclad. This implementation supports send and sendpages (splice). GSO replaces one large packet with multiple smaller packets. This patch also copies the option into the correct smaller packet. This patch does not yet support timestamping on data in an initial TCP Fast Open SYN, because that takes a very different data path. If ID generation in ee_data is enabled, bytestream timestamps return a byte offset, instead of the packet counter for datagrams. The implementation supports a single timestamp per packet. It silenty replaces requests for previous timestamps. To avoid missing tstamps, flush the tcp queue by disabling Nagle, cork and autocork. Missing tstamps can be detected by offset when the ee_data ID is enabled. Implementation details: - On GSO, the timestamping code can be included in the main loop. I moved it into its own loop to reduce the impact on the common case to a single branch. - To avoid leaking the absolute seqno to userspace, the offset returned in ee_data must always be relative. It is an offset between an skb and sk field. The first is always set (also for GSO & ACK). The second must also never be uninitialized. Only allow the ID option on sockets in the ESTABLISHED state, for which the seqno is available. Never reset it to zero (instead, move it to the current seqno when reenabling the option). Signed-off-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-08-05 10:11:49 +08:00
}
static bool tcp_stream_is_readable(struct sock *sk, int target)
{
if (tcp_epollin_ready(sk, target))
return true;
return sk_is_readable(sk);
}
/*
* Wait for a TCP event.
*
* Note that we don't need to lock the socket, as the upper poll layers
* take care of normal races (between the test and the event) and we don't
* go look at any of the socket buffers directly.
*/
__poll_t tcp_poll(struct file *file, struct socket *sock, poll_table *wait)
{
__poll_t mask;
struct sock *sk = sock->sk;
const struct tcp_sock *tp = tcp_sk(sk);
tcp: add annotations around sk->sk_shutdown accesses Now sk->sk_shutdown is no longer a bitfield, we can add standard READ_ONCE()/WRITE_ONCE() annotations to silence KCSAN reports like the following: BUG: KCSAN: data-race in tcp_disconnect / tcp_poll write to 0xffff88814588582c of 1 bytes by task 3404 on cpu 1: tcp_disconnect+0x4d6/0xdb0 net/ipv4/tcp.c:3121 __inet_stream_connect+0x5dd/0x6e0 net/ipv4/af_inet.c:715 inet_stream_connect+0x48/0x70 net/ipv4/af_inet.c:727 __sys_connect_file net/socket.c:2001 [inline] __sys_connect+0x19b/0x1b0 net/socket.c:2018 __do_sys_connect net/socket.c:2028 [inline] __se_sys_connect net/socket.c:2025 [inline] __x64_sys_connect+0x41/0x50 net/socket.c:2025 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd read to 0xffff88814588582c of 1 bytes by task 3374 on cpu 0: tcp_poll+0x2e6/0x7d0 net/ipv4/tcp.c:562 sock_poll+0x253/0x270 net/socket.c:1383 vfs_poll include/linux/poll.h:88 [inline] io_poll_check_events io_uring/poll.c:281 [inline] io_poll_task_func+0x15a/0x820 io_uring/poll.c:333 handle_tw_list io_uring/io_uring.c:1184 [inline] tctx_task_work+0x1fe/0x4d0 io_uring/io_uring.c:1246 task_work_run+0x123/0x160 kernel/task_work.c:179 get_signal+0xe64/0xff0 kernel/signal.c:2635 arch_do_signal_or_restart+0x89/0x2a0 arch/x86/kernel/signal.c:306 exit_to_user_mode_loop+0x6f/0xe0 kernel/entry/common.c:168 exit_to_user_mode_prepare+0x6c/0xb0 kernel/entry/common.c:204 __syscall_exit_to_user_mode_work kernel/entry/common.c:286 [inline] syscall_exit_to_user_mode+0x26/0x140 kernel/entry/common.c:297 do_syscall_64+0x4d/0xc0 arch/x86/entry/common.c:86 entry_SYSCALL_64_after_hwframe+0x63/0xcd value changed: 0x03 -> 0x00 Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-05-10 04:36:56 +08:00
u8 shutdown;
int state;
sock_poll_wait(file, sock, wait);
state = inet_sk_state_load(sk);
if (state == TCP_LISTEN)
return inet_csk_listen_poll(sk);
/* Socket is not locked. We are protected from async events
* by poll logic and correct handling of state changes
* made by other threads is impossible in any case.
*/
mask = 0;
/*
* EPOLLHUP is certainly not done right. But poll() doesn't
* have a notion of HUP in just one direction, and for a
* socket the read side is more interesting.
*
* Some poll() documentation says that EPOLLHUP is incompatible
* with the EPOLLOUT/POLLWR flags, so somebody should check this
* all. But careful, it tends to be safer to return too many
* bits than too few, and you can easily break real applications
* if you don't tell them that something has hung up!
*
* Check-me.
*
* Check number 1. EPOLLHUP is _UNMASKABLE_ event (see UNIX98 and
* our fs/select.c). It means that after we received EOF,
* poll always returns immediately, making impossible poll() on write()
* in state CLOSE_WAIT. One solution is evident --- to set EPOLLHUP
* if and only if shutdown has been made in both directions.
* Actually, it is interesting to look how Solaris and DUX
* solve this dilemma. I would prefer, if EPOLLHUP were maskable,
* then we could set it on SND_SHUTDOWN. BTW examples given
* in Stevens' books assume exactly this behaviour, it explains
* why EPOLLHUP is incompatible with EPOLLOUT. --ANK
*
* NOTE. Check for TCP_CLOSE is added. The goal is to prevent
* blocking on fresh not-connected or disconnected socket. --ANK
*/
tcp: add annotations around sk->sk_shutdown accesses Now sk->sk_shutdown is no longer a bitfield, we can add standard READ_ONCE()/WRITE_ONCE() annotations to silence KCSAN reports like the following: BUG: KCSAN: data-race in tcp_disconnect / tcp_poll write to 0xffff88814588582c of 1 bytes by task 3404 on cpu 1: tcp_disconnect+0x4d6/0xdb0 net/ipv4/tcp.c:3121 __inet_stream_connect+0x5dd/0x6e0 net/ipv4/af_inet.c:715 inet_stream_connect+0x48/0x70 net/ipv4/af_inet.c:727 __sys_connect_file net/socket.c:2001 [inline] __sys_connect+0x19b/0x1b0 net/socket.c:2018 __do_sys_connect net/socket.c:2028 [inline] __se_sys_connect net/socket.c:2025 [inline] __x64_sys_connect+0x41/0x50 net/socket.c:2025 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd read to 0xffff88814588582c of 1 bytes by task 3374 on cpu 0: tcp_poll+0x2e6/0x7d0 net/ipv4/tcp.c:562 sock_poll+0x253/0x270 net/socket.c:1383 vfs_poll include/linux/poll.h:88 [inline] io_poll_check_events io_uring/poll.c:281 [inline] io_poll_task_func+0x15a/0x820 io_uring/poll.c:333 handle_tw_list io_uring/io_uring.c:1184 [inline] tctx_task_work+0x1fe/0x4d0 io_uring/io_uring.c:1246 task_work_run+0x123/0x160 kernel/task_work.c:179 get_signal+0xe64/0xff0 kernel/signal.c:2635 arch_do_signal_or_restart+0x89/0x2a0 arch/x86/kernel/signal.c:306 exit_to_user_mode_loop+0x6f/0xe0 kernel/entry/common.c:168 exit_to_user_mode_prepare+0x6c/0xb0 kernel/entry/common.c:204 __syscall_exit_to_user_mode_work kernel/entry/common.c:286 [inline] syscall_exit_to_user_mode+0x26/0x140 kernel/entry/common.c:297 do_syscall_64+0x4d/0xc0 arch/x86/entry/common.c:86 entry_SYSCALL_64_after_hwframe+0x63/0xcd value changed: 0x03 -> 0x00 Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-05-10 04:36:56 +08:00
shutdown = READ_ONCE(sk->sk_shutdown);
if (shutdown == SHUTDOWN_MASK || state == TCP_CLOSE)
mask |= EPOLLHUP;
tcp: add annotations around sk->sk_shutdown accesses Now sk->sk_shutdown is no longer a bitfield, we can add standard READ_ONCE()/WRITE_ONCE() annotations to silence KCSAN reports like the following: BUG: KCSAN: data-race in tcp_disconnect / tcp_poll write to 0xffff88814588582c of 1 bytes by task 3404 on cpu 1: tcp_disconnect+0x4d6/0xdb0 net/ipv4/tcp.c:3121 __inet_stream_connect+0x5dd/0x6e0 net/ipv4/af_inet.c:715 inet_stream_connect+0x48/0x70 net/ipv4/af_inet.c:727 __sys_connect_file net/socket.c:2001 [inline] __sys_connect+0x19b/0x1b0 net/socket.c:2018 __do_sys_connect net/socket.c:2028 [inline] __se_sys_connect net/socket.c:2025 [inline] __x64_sys_connect+0x41/0x50 net/socket.c:2025 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd read to 0xffff88814588582c of 1 bytes by task 3374 on cpu 0: tcp_poll+0x2e6/0x7d0 net/ipv4/tcp.c:562 sock_poll+0x253/0x270 net/socket.c:1383 vfs_poll include/linux/poll.h:88 [inline] io_poll_check_events io_uring/poll.c:281 [inline] io_poll_task_func+0x15a/0x820 io_uring/poll.c:333 handle_tw_list io_uring/io_uring.c:1184 [inline] tctx_task_work+0x1fe/0x4d0 io_uring/io_uring.c:1246 task_work_run+0x123/0x160 kernel/task_work.c:179 get_signal+0xe64/0xff0 kernel/signal.c:2635 arch_do_signal_or_restart+0x89/0x2a0 arch/x86/kernel/signal.c:306 exit_to_user_mode_loop+0x6f/0xe0 kernel/entry/common.c:168 exit_to_user_mode_prepare+0x6c/0xb0 kernel/entry/common.c:204 __syscall_exit_to_user_mode_work kernel/entry/common.c:286 [inline] syscall_exit_to_user_mode+0x26/0x140 kernel/entry/common.c:297 do_syscall_64+0x4d/0xc0 arch/x86/entry/common.c:86 entry_SYSCALL_64_after_hwframe+0x63/0xcd value changed: 0x03 -> 0x00 Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-05-10 04:36:56 +08:00
if (shutdown & RCV_SHUTDOWN)
mask |= EPOLLIN | EPOLLRDNORM | EPOLLRDHUP;
/* Connected or passive Fast Open socket? */
if (state != TCP_SYN_SENT &&
(state != TCP_SYN_RECV || rcu_access_pointer(tp->fastopen_rsk))) {
int target = sock_rcvlowat(sk, 0, INT_MAX);
u16 urg_data = READ_ONCE(tp->urg_data);
if (unlikely(urg_data) &&
READ_ONCE(tp->urg_seq) == READ_ONCE(tp->copied_seq) &&
!sock_flag(sk, SOCK_URGINLINE))
target++;
if (tcp_stream_is_readable(sk, target))
mask |= EPOLLIN | EPOLLRDNORM;
tcp: add annotations around sk->sk_shutdown accesses Now sk->sk_shutdown is no longer a bitfield, we can add standard READ_ONCE()/WRITE_ONCE() annotations to silence KCSAN reports like the following: BUG: KCSAN: data-race in tcp_disconnect / tcp_poll write to 0xffff88814588582c of 1 bytes by task 3404 on cpu 1: tcp_disconnect+0x4d6/0xdb0 net/ipv4/tcp.c:3121 __inet_stream_connect+0x5dd/0x6e0 net/ipv4/af_inet.c:715 inet_stream_connect+0x48/0x70 net/ipv4/af_inet.c:727 __sys_connect_file net/socket.c:2001 [inline] __sys_connect+0x19b/0x1b0 net/socket.c:2018 __do_sys_connect net/socket.c:2028 [inline] __se_sys_connect net/socket.c:2025 [inline] __x64_sys_connect+0x41/0x50 net/socket.c:2025 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd read to 0xffff88814588582c of 1 bytes by task 3374 on cpu 0: tcp_poll+0x2e6/0x7d0 net/ipv4/tcp.c:562 sock_poll+0x253/0x270 net/socket.c:1383 vfs_poll include/linux/poll.h:88 [inline] io_poll_check_events io_uring/poll.c:281 [inline] io_poll_task_func+0x15a/0x820 io_uring/poll.c:333 handle_tw_list io_uring/io_uring.c:1184 [inline] tctx_task_work+0x1fe/0x4d0 io_uring/io_uring.c:1246 task_work_run+0x123/0x160 kernel/task_work.c:179 get_signal+0xe64/0xff0 kernel/signal.c:2635 arch_do_signal_or_restart+0x89/0x2a0 arch/x86/kernel/signal.c:306 exit_to_user_mode_loop+0x6f/0xe0 kernel/entry/common.c:168 exit_to_user_mode_prepare+0x6c/0xb0 kernel/entry/common.c:204 __syscall_exit_to_user_mode_work kernel/entry/common.c:286 [inline] syscall_exit_to_user_mode+0x26/0x140 kernel/entry/common.c:297 do_syscall_64+0x4d/0xc0 arch/x86/entry/common.c:86 entry_SYSCALL_64_after_hwframe+0x63/0xcd value changed: 0x03 -> 0x00 Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-05-10 04:36:56 +08:00
if (!(shutdown & SEND_SHUTDOWN)) {
if (__sk_stream_is_writeable(sk, 1)) {
mask |= EPOLLOUT | EPOLLWRNORM;
} else { /* send SIGIO later */
sk_set_bit(SOCKWQ_ASYNC_NOSPACE, sk);
set_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
/* Race breaker. If space is freed after
* wspace test but before the flags are set,
* IO signal will be lost. Memory barrier
* pairs with the input side.
*/
smp_mb__after_atomic();
if (__sk_stream_is_writeable(sk, 1))
mask |= EPOLLOUT | EPOLLWRNORM;
}
2010-08-25 00:05:48 +08:00
} else
mask |= EPOLLOUT | EPOLLWRNORM;
if (urg_data & TCP_URG_VALID)
mask |= EPOLLPRI;
} else if (state == TCP_SYN_SENT &&
inet_test_bit(DEFER_CONNECT, sk)) {
net/tcp-fastopen: Add new API support This patch adds a new socket option, TCP_FASTOPEN_CONNECT, as an alternative way to perform Fast Open on the active side (client). Prior to this patch, a client needs to replace the connect() call with sendto(MSG_FASTOPEN). This can be cumbersome for applications who want to use Fast Open: these socket operations are often done in lower layer libraries used by many other applications. Changing these libraries and/or the socket call sequences are not trivial. A more convenient approach is to perform Fast Open by simply enabling a socket option when the socket is created w/o changing other socket calls sequence: s = socket() create a new socket setsockopt(s, IPPROTO_TCP, TCP_FASTOPEN_CONNECT …); newly introduced sockopt If set, new functionality described below will be used. Return ENOTSUPP if TFO is not supported or not enabled in the kernel. connect() With cookie present, return 0 immediately. With no cookie, initiate 3WHS with TFO cookie-request option and return -1 with errno = EINPROGRESS. write()/sendmsg() With cookie present, send out SYN with data and return the number of bytes buffered. With no cookie, and 3WHS not yet completed, return -1 with errno = EINPROGRESS. No MSG_FASTOPEN flag is needed. read() Return -1 with errno = EWOULDBLOCK/EAGAIN if connect() is called but write() is not called yet. Return -1 with errno = EWOULDBLOCK/EAGAIN if connection is established but no msg is received yet. Return number of bytes read if socket is established and there is msg received. The new API simplifies life for applications that always perform a write() immediately after a successful connect(). Such applications can now take advantage of Fast Open by merely making one new setsockopt() call at the time of creating the socket. Nothing else about the application's socket call sequence needs to change. Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-24 02:59:22 +08:00
/* Active TCP fastopen socket with defer_connect
* Return EPOLLOUT so application can call write()
net/tcp-fastopen: Add new API support This patch adds a new socket option, TCP_FASTOPEN_CONNECT, as an alternative way to perform Fast Open on the active side (client). Prior to this patch, a client needs to replace the connect() call with sendto(MSG_FASTOPEN). This can be cumbersome for applications who want to use Fast Open: these socket operations are often done in lower layer libraries used by many other applications. Changing these libraries and/or the socket call sequences are not trivial. A more convenient approach is to perform Fast Open by simply enabling a socket option when the socket is created w/o changing other socket calls sequence: s = socket() create a new socket setsockopt(s, IPPROTO_TCP, TCP_FASTOPEN_CONNECT …); newly introduced sockopt If set, new functionality described below will be used. Return ENOTSUPP if TFO is not supported or not enabled in the kernel. connect() With cookie present, return 0 immediately. With no cookie, initiate 3WHS with TFO cookie-request option and return -1 with errno = EINPROGRESS. write()/sendmsg() With cookie present, send out SYN with data and return the number of bytes buffered. With no cookie, and 3WHS not yet completed, return -1 with errno = EINPROGRESS. No MSG_FASTOPEN flag is needed. read() Return -1 with errno = EWOULDBLOCK/EAGAIN if connect() is called but write() is not called yet. Return -1 with errno = EWOULDBLOCK/EAGAIN if connection is established but no msg is received yet. Return number of bytes read if socket is established and there is msg received. The new API simplifies life for applications that always perform a write() immediately after a successful connect(). Such applications can now take advantage of Fast Open by merely making one new setsockopt() call at the time of creating the socket. Nothing else about the application's socket call sequence needs to change. Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-24 02:59:22 +08:00
* in order for kernel to generate SYN+data
*/
mask |= EPOLLOUT | EPOLLWRNORM;
}
/* This barrier is coupled with smp_wmb() in tcp_reset() */
smp_rmb();
if (READ_ONCE(sk->sk_err) ||
!skb_queue_empty_lockless(&sk->sk_error_queue))
mask |= EPOLLERR;
return mask;
}
EXPORT_SYMBOL(tcp_poll);
net: ioctl: Use kernel memory on protocol ioctl callbacks Most of the ioctls to net protocols operates directly on userspace argument (arg). Usually doing get_user()/put_user() directly in the ioctl callback. This is not flexible, because it is hard to reuse these functions without passing userspace buffers. Change the "struct proto" ioctls to avoid touching userspace memory and operate on kernel buffers, i.e., all protocol's ioctl callbacks is adapted to operate on a kernel memory other than on userspace (so, no more {put,get}_user() and friends being called in the ioctl callback). This changes the "struct proto" ioctl format in the following way: int (*ioctl)(struct sock *sk, int cmd, - unsigned long arg); + int *karg); (Important to say that this patch does not touch the "struct proto_ops" protocols) So, the "karg" argument, which is passed to the ioctl callback, is a pointer allocated to kernel space memory (inside a function wrapper). This buffer (karg) may contain input argument (copied from userspace in a prep function) and it might return a value/buffer, which is copied back to userspace if necessary. There is not one-size-fits-all format (that is I am using 'may' above), but basically, there are three type of ioctls: 1) Do not read from userspace, returns a result to userspace 2) Read an input parameter from userspace, and does not return anything to userspace 3) Read an input from userspace, and return a buffer to userspace. The default case (1) (where no input parameter is given, and an "int" is returned to userspace) encompasses more than 90% of the cases, but there are two other exceptions. Here is a list of exceptions: * Protocol RAW: * cmd = SIOCGETVIFCNT: * input and output = struct sioc_vif_req * cmd = SIOCGETSGCNT * input and output = struct sioc_sg_req * Explanation: for the SIOCGETVIFCNT case, userspace passes the input argument, which is struct sioc_vif_req. Then the callback populates the struct, which is copied back to userspace. * Protocol RAW6: * cmd = SIOCGETMIFCNT_IN6 * input and output = struct sioc_mif_req6 * cmd = SIOCGETSGCNT_IN6 * input and output = struct sioc_sg_req6 * Protocol PHONET: * cmd == SIOCPNADDRESOURCE | SIOCPNDELRESOURCE * input int (4 bytes) * Nothing is copied back to userspace. For the exception cases, functions sock_sk_ioctl_inout() will copy the userspace input, and copy it back to kernel space. The wrapper that prepare the buffer and put the buffer back to user is sk_ioctl(), so, instead of calling sk->sk_prot->ioctl(), the callee now calls sk_ioctl(), which will handle all cases. Signed-off-by: Breno Leitao <leitao@debian.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Reviewed-by: David Ahern <dsahern@kernel.org> Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com> Link: https://lore.kernel.org/r/20230609152800.830401-1-leitao@debian.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-06-09 23:27:42 +08:00
int tcp_ioctl(struct sock *sk, int cmd, int *karg)
{
struct tcp_sock *tp = tcp_sk(sk);
int answ;
bool slow;
switch (cmd) {
case SIOCINQ:
if (sk->sk_state == TCP_LISTEN)
return -EINVAL;
slow = lock_sock_fast(sk);
answ = tcp_inq(sk);
unlock_sock_fast(sk, slow);
break;
case SIOCATMARK:
answ = READ_ONCE(tp->urg_data) &&
READ_ONCE(tp->urg_seq) == READ_ONCE(tp->copied_seq);
break;
case SIOCOUTQ:
if (sk->sk_state == TCP_LISTEN)
return -EINVAL;
if ((1 << sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV))
answ = 0;
else
answ = READ_ONCE(tp->write_seq) - tp->snd_una;
break;
case SIOCOUTQNSD:
if (sk->sk_state == TCP_LISTEN)
return -EINVAL;
if ((1 << sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV))
answ = 0;
else
answ = READ_ONCE(tp->write_seq) -
READ_ONCE(tp->snd_nxt);
break;
default:
return -ENOIOCTLCMD;
}
net: ioctl: Use kernel memory on protocol ioctl callbacks Most of the ioctls to net protocols operates directly on userspace argument (arg). Usually doing get_user()/put_user() directly in the ioctl callback. This is not flexible, because it is hard to reuse these functions without passing userspace buffers. Change the "struct proto" ioctls to avoid touching userspace memory and operate on kernel buffers, i.e., all protocol's ioctl callbacks is adapted to operate on a kernel memory other than on userspace (so, no more {put,get}_user() and friends being called in the ioctl callback). This changes the "struct proto" ioctl format in the following way: int (*ioctl)(struct sock *sk, int cmd, - unsigned long arg); + int *karg); (Important to say that this patch does not touch the "struct proto_ops" protocols) So, the "karg" argument, which is passed to the ioctl callback, is a pointer allocated to kernel space memory (inside a function wrapper). This buffer (karg) may contain input argument (copied from userspace in a prep function) and it might return a value/buffer, which is copied back to userspace if necessary. There is not one-size-fits-all format (that is I am using 'may' above), but basically, there are three type of ioctls: 1) Do not read from userspace, returns a result to userspace 2) Read an input parameter from userspace, and does not return anything to userspace 3) Read an input from userspace, and return a buffer to userspace. The default case (1) (where no input parameter is given, and an "int" is returned to userspace) encompasses more than 90% of the cases, but there are two other exceptions. Here is a list of exceptions: * Protocol RAW: * cmd = SIOCGETVIFCNT: * input and output = struct sioc_vif_req * cmd = SIOCGETSGCNT * input and output = struct sioc_sg_req * Explanation: for the SIOCGETVIFCNT case, userspace passes the input argument, which is struct sioc_vif_req. Then the callback populates the struct, which is copied back to userspace. * Protocol RAW6: * cmd = SIOCGETMIFCNT_IN6 * input and output = struct sioc_mif_req6 * cmd = SIOCGETSGCNT_IN6 * input and output = struct sioc_sg_req6 * Protocol PHONET: * cmd == SIOCPNADDRESOURCE | SIOCPNDELRESOURCE * input int (4 bytes) * Nothing is copied back to userspace. For the exception cases, functions sock_sk_ioctl_inout() will copy the userspace input, and copy it back to kernel space. The wrapper that prepare the buffer and put the buffer back to user is sk_ioctl(), so, instead of calling sk->sk_prot->ioctl(), the callee now calls sk_ioctl(), which will handle all cases. Signed-off-by: Breno Leitao <leitao@debian.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Reviewed-by: David Ahern <dsahern@kernel.org> Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com> Link: https://lore.kernel.org/r/20230609152800.830401-1-leitao@debian.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2023-06-09 23:27:42 +08:00
*karg = answ;
return 0;
}
EXPORT_SYMBOL(tcp_ioctl);
void tcp_mark_push(struct tcp_sock *tp, struct sk_buff *skb)
{
TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_PSH;
tp->pushed_seq = tp->write_seq;
}
static inline bool forced_push(const struct tcp_sock *tp)
{
return after(tp->write_seq, tp->pushed_seq + (tp->max_window >> 1));
}
void tcp_skb_entail(struct sock *sk, struct sk_buff *skb)
{
[TCP]: Sed magic converts func(sk, tp, ...) -> func(sk, ...) This is (mostly) automated change using magic: sed -e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N' -e 's|struct sock \*sk,[\n\t ]*struct tcp_sock \*tp\([^{]*\n{\n\)| struct sock \*sk\1\tstruct tcp_sock *tp = tcp_sk(sk);\n|g' -e 's|struct sock \*sk, struct tcp_sock \*tp| struct sock \*sk|g' -e 's|sk, tp\([^-]\)|sk\1|g' Fixed four unused variable (tp) warnings that were introduced. In addition, manually added newlines after local variables and tweaked function arguments positioning. $ gcc --version gcc (GCC) 4.1.1 20060525 (Red Hat 4.1.1-1) ... $ codiff -fV built-in.o.old built-in.o.new net/ipv4/route.c: rt_cache_flush | +14 1 function changed, 14 bytes added net/ipv4/tcp.c: tcp_setsockopt | -5 tcp_sendpage | -25 tcp_sendmsg | -16 3 functions changed, 46 bytes removed net/ipv4/tcp_input.c: tcp_try_undo_recovery | +3 tcp_try_undo_dsack | +2 tcp_mark_head_lost | -12 tcp_ack | -15 tcp_event_data_recv | -32 tcp_rcv_state_process | -10 tcp_rcv_established | +1 7 functions changed, 6 bytes added, 69 bytes removed, diff: -63 net/ipv4/tcp_output.c: update_send_head | -9 tcp_transmit_skb | +19 tcp_cwnd_validate | +1 tcp_write_wakeup | -17 __tcp_push_pending_frames | -25 tcp_push_one | -8 tcp_send_fin | -4 7 functions changed, 20 bytes added, 63 bytes removed, diff: -43 built-in.o.new: 18 functions changed, 40 bytes added, 178 bytes removed, diff: -138 Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-21 13:18:02 +08:00
struct tcp_sock *tp = tcp_sk(sk);
struct tcp_skb_cb *tcb = TCP_SKB_CB(skb);
tcb->seq = tcb->end_seq = tp->write_seq;
tcb->tcp_flags = TCPHDR_ACK;
__skb_header_release(skb);
tcp_add_write_queue_tail(sk, skb);
sk_wmem_queued_add(sk, skb->truesize);
[NET] CORE: Introducing new memory accounting interface. This patch introduces new memory accounting functions for each network protocol. Most of them are renamed from memory accounting functions for stream protocols. At the same time, some stream memory accounting functions are removed since other functions do same thing. Renaming: sk_stream_free_skb() -> sk_wmem_free_skb() __sk_stream_mem_reclaim() -> __sk_mem_reclaim() sk_stream_mem_reclaim() -> sk_mem_reclaim() sk_stream_mem_schedule -> __sk_mem_schedule() sk_stream_pages() -> sk_mem_pages() sk_stream_rmem_schedule() -> sk_rmem_schedule() sk_stream_wmem_schedule() -> sk_wmem_schedule() sk_charge_skb() -> sk_mem_charge() Removeing sk_stream_rfree(): consolidates into sock_rfree() sk_stream_set_owner_r(): consolidates into skb_set_owner_r() sk_stream_mem_schedule() The following functions are added. sk_has_account(): check if the protocol supports accounting sk_mem_uncharge(): do the opposite of sk_mem_charge() In addition, to achieve consolidation, updating sk_wmem_queued is removed from sk_mem_charge(). Next, to consolidate memory accounting functions, this patch adds memory accounting calls to network core functions. Moreover, present memory accounting call is renamed to new accounting call. Finally we replace present memory accounting calls with new interface in TCP and SCTP. Signed-off-by: Takahiro Yasui <tyasui@redhat.com> Signed-off-by: Hideo Aoki <haoki@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-12-31 16:11:19 +08:00
sk_mem_charge(sk, skb->truesize);
if (tp->nonagle & TCP_NAGLE_PUSH)
tp->nonagle &= ~TCP_NAGLE_PUSH;
tcp: fix slow start after idle vs TSO/GSO slow start after idle might reduce cwnd, but we perform this after first packet was cooked and sent. With TSO/GSO, it means that we might send a full TSO packet even if cwnd should have been reduced to IW10. Moving the SSAI check in skb_entail() makes sense, because we slightly reduce number of times this check is done, especially for large send() and TCP Small queue callbacks from softirq context. As Neal pointed out, we also need to perform the check if/when receive window opens. Tested: Following packetdrill test demonstrates the problem // Test of slow start after idle `sysctl -q net.ipv4.tcp_slow_start_after_idle=1` 0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3 +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0 +0 bind(3, ..., ...) = 0 +0 listen(3, 1) = 0 +0 < S 0:0(0) win 65535 <mss 1000,sackOK,nop,nop,nop,wscale 7> +0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 6> +.100 < . 1:1(0) ack 1 win 511 +0 accept(3, ..., ...) = 4 +0 setsockopt(4, SOL_SOCKET, SO_SNDBUF, [200000], 4) = 0 +0 write(4, ..., 26000) = 26000 +0 > . 1:5001(5000) ack 1 +0 > . 5001:10001(5000) ack 1 +0 %{ assert tcpi_snd_cwnd == 10 }% +.100 < . 1:1(0) ack 10001 win 511 +0 %{ assert tcpi_snd_cwnd == 20, tcpi_snd_cwnd }% +0 > . 10001:20001(10000) ack 1 +0 > P. 20001:26001(6000) ack 1 +.100 < . 1:1(0) ack 26001 win 511 +0 %{ assert tcpi_snd_cwnd == 36, tcpi_snd_cwnd }% +4 write(4, ..., 20000) = 20000 // If slow start after idle works properly, we should send 5 MSS here (cwnd/2) +0 > . 26001:31001(5000) ack 1 +0 %{ assert tcpi_snd_cwnd == 10, tcpi_snd_cwnd }% +0 > . 31001:36001(5000) ack 1 Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-08-22 03:30:00 +08:00
tcp_slow_start_after_idle_check(sk);
}
static inline void tcp_mark_urg(struct tcp_sock *tp, int flags)
{
if (flags & MSG_OOB)
tp->snd_up = tp->write_seq;
}
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
/* If a not yet filled skb is pushed, do not send it if
* we have data packets in Qdisc or NIC queues :
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
* Because TX completion will happen shortly, it gives a chance
* to coalesce future sendmsg() payload into this skb, without
* need for a timer, and with no latency trade off.
* As packets containing data payload have a bigger truesize
* than pure acks (dataless) packets, the last checks prevent
* autocorking if we only have an ACK in Qdisc/NIC queues,
* or if TX completion was delayed after we processed ACK packet.
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
*/
static bool tcp_should_autocork(struct sock *sk, struct sk_buff *skb,
int size_goal)
{
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
return skb->len < size_goal &&
READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_autocorking) &&
tcp: restore autocorking When adding rb-tree for TCP retransmit queue, we inadvertently broke TCP autocorking. tcp_should_autocork() should really check if the rtx queue is not empty. Tested: Before the fix : $ nstat -n;./netperf -H 10.246.7.152 -Cc -- -m 500;nstat | grep AutoCork MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.152 () port 0 AF_INET Recv Send Send Utilization Service Demand Socket Socket Message Elapsed Send Recv Send Recv Size Size Size Time Throughput local remote local remote bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB 540000 262144 500 10.00 2682.85 2.47 1.59 3.618 2.329 TcpExtTCPAutoCorking 33 0.0 // Same test, but forcing TCP_NODELAY $ nstat -n;./netperf -H 10.246.7.152 -Cc -- -D -m 500;nstat | grep AutoCork MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.152 () port 0 AF_INET : nodelay Recv Send Send Utilization Service Demand Socket Socket Message Elapsed Send Recv Send Recv Size Size Size Time Throughput local remote local remote bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB 540000 262144 500 10.00 1408.75 2.44 2.96 6.802 8.259 TcpExtTCPAutoCorking 1 0.0 After the fix : $ nstat -n;./netperf -H 10.246.7.152 -Cc -- -m 500;nstat | grep AutoCork MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.152 () port 0 AF_INET Recv Send Send Utilization Service Demand Socket Socket Message Elapsed Send Recv Send Recv Size Size Size Time Throughput local remote local remote bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB 540000 262144 500 10.00 5472.46 2.45 1.43 1.761 1.027 TcpExtTCPAutoCorking 361293 0.0 // With TCP_NODELAY option $ nstat -n;./netperf -H 10.246.7.152 -Cc -- -D -m 500;nstat | grep AutoCork MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.152 () port 0 AF_INET : nodelay Recv Send Send Utilization Service Demand Socket Socket Message Elapsed Send Recv Send Recv Size Size Size Time Throughput local remote local remote bytes bytes bytes secs. 10^6bits/s % S % S us/KB us/KB 540000 262144 500 10.00 5454.96 2.46 1.63 1.775 1.174 TcpExtTCPAutoCorking 315448 0.0 Fixes: 75c119afe14f ("tcp: implement rb-tree based retransmit queue") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Michael Wenig <mwenig@vmware.com> Tested-by: Michael Wenig <mwenig@vmware.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Michael Wenig <mwenig@vmware.com> Tested-by: Michael Wenig <mwenig@vmware.com> Acked-by: Neal Cardwell <ncardwell@google.com> Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-05-03 11:25:13 +08:00
!tcp_rtx_queue_empty(sk) &&
refcount_read(&sk->sk_wmem_alloc) > skb->truesize &&
tcp_skb_can_collapse_to(skb);
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
}
void tcp_push(struct sock *sk, int flags, int mss_now,
int nonagle, int size_goal)
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
{
struct tcp_sock *tp = tcp_sk(sk);
struct sk_buff *skb;
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
skb = tcp_write_queue_tail(sk);
tcp: implement rb-tree based retransmit queue Using a linear list to store all skbs in write queue has been okay for quite a while : O(N) is not too bad when N < 500. Things get messy when N is the order of 100,000 : Modern TCP stacks want 10Gbit+ of throughput even with 200 ms RTT flows. 40 ns per cache line miss means a full scan can use 4 ms, blowing away CPU caches. SACK processing often can use various hints to avoid parsing whole retransmit queue. But with high packet losses and/or high reordering, hints no longer work. Sender has to process thousands of unfriendly SACK, accumulating a huge socket backlog, burning a cpu and massively dropping packets. Using an rb-tree for retransmit queue has been avoided for years because it added complexity and overhead, but now is the time to be more resistant and say no to quadratic behavior. 1) RTX queue is no longer part of the write queue : already sent skbs are stored in one rb-tree. 2) Since reaching the head of write queue no longer needs sk->sk_send_head, we added an union of sk_send_head and tcp_rtx_queue Tested: On receiver : netem on ingress : delay 150ms 200us loss 1 GRO disabled to force stress and SACK storms. for f in `seq 1 10` do ./netperf -H lpaa6 -l30 -- -K bbr -o THROUGHPUT|tail -1 done | awk '{print $0} {sum += $0} END {printf "%7u\n",sum}' Before patch : 323.87 351.48 339.59 338.62 306.72 204.07 304.93 291.88 202.47 176.88 2840 After patch: 1700.83 2207.98 2070.17 1544.26 2114.76 2124.89 1693.14 1080.91 2216.82 1299.94 18053 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-06 13:21:27 +08:00
if (!skb)
return;
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
if (!(flags & MSG_MORE) || forced_push(tp))
tcp_mark_push(tp, skb);
tcp_mark_urg(tp, flags);
if (tcp_should_autocork(sk, skb, size_goal)) {
/* avoid atomic op if TSQ_THROTTLED bit is already set */
if (!test_bit(TSQ_THROTTLED, &sk->sk_tsq_flags)) {
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPAUTOCORKING);
set_bit(TSQ_THROTTLED, &sk->sk_tsq_flags);
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
}
/* It is possible TX completion already happened
* before we set TSQ_THROTTLED.
*/
if (refcount_read(&sk->sk_wmem_alloc) > skb->truesize)
return;
}
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
if (flags & MSG_MORE)
nonagle = TCP_NAGLE_CORK;
__tcp_push_pending_frames(sk, mss_now, nonagle);
}
static int tcp_splice_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb,
unsigned int offset, size_t len)
{
struct tcp_splice_state *tss = rd_desc->arg.data;
int ret;
ret = skb_splice_bits(skb, skb->sk, offset, tss->pipe,
min(rd_desc->count, len), tss->flags);
if (ret > 0)
rd_desc->count -= ret;
return ret;
}
static int __tcp_splice_read(struct sock *sk, struct tcp_splice_state *tss)
{
/* Store TCP splice context information in read_descriptor_t. */
read_descriptor_t rd_desc = {
.arg.data = tss,
.count = tss->len,
};
return tcp_read_sock(sk, &rd_desc, tcp_splice_data_recv);
}
/**
* tcp_splice_read - splice data from TCP socket to a pipe
* @sock: socket to splice from
* @ppos: position (not valid)
* @pipe: pipe to splice to
* @len: number of bytes to splice
* @flags: splice modifier flags
*
* Description:
* Will read pages from given socket and fill them into a pipe.
*
**/
ssize_t tcp_splice_read(struct socket *sock, loff_t *ppos,
struct pipe_inode_info *pipe, size_t len,
unsigned int flags)
{
struct sock *sk = sock->sk;
struct tcp_splice_state tss = {
.pipe = pipe,
.len = len,
.flags = flags,
};
long timeo;
ssize_t spliced;
int ret;
sock_rps_record_flow(sk);
/*
* We can't seek on a socket input
*/
if (unlikely(*ppos))
return -ESPIPE;
ret = spliced = 0;
lock_sock(sk);
net: splice() from tcp to pipe should take into account O_NONBLOCK tcp_splice_read() doesnt take into account socket's O_NONBLOCK flag Before this patch : splice(socket,0,pipe,0,128*1024,SPLICE_F_MOVE); causes a random endless block (if pipe is full) and splice(socket,0,pipe,0,128*1024,SPLICE_F_MOVE | SPLICE_F_NONBLOCK); will return 0 immediately if the TCP buffer is empty. User application has no way to instruct splice() that socket should be in blocking mode but pipe in nonblock more. Many projects cannot use splice(tcp -> pipe) because of this flaw. http://git.samba.org/?p=samba.git;a=history;f=source3/lib/recvfile.c;h=ea0159642137390a0f7e57a123684e6e63e47581;hb=HEAD http://lkml.indiana.edu/hypermail/linux/kernel/0807.2/0687.html Linus introduced SPLICE_F_NONBLOCK in commit 29e350944fdc2dfca102500790d8ad6d6ff4f69d (splice: add SPLICE_F_NONBLOCK flag ) It doesn't make the splice itself necessarily nonblocking (because the actual file descriptors that are spliced from/to may block unless they have the O_NONBLOCK flag set), but it makes the splice pipe operations nonblocking. Linus intention was clear : let SPLICE_F_NONBLOCK control the splice pipe mode only This patch instruct tcp_splice_read() to use the underlying file O_NONBLOCK flag, as other socket operations do. Users will then call : splice(socket,0,pipe,0,128*1024,SPLICE_F_MOVE | SPLICE_F_NONBLOCK ); to block on data coming from socket (if file is in blocking mode), and not block on pipe output (to avoid deadlock) First version of this patch was submitted by Octavian Purdila Reported-by: Volker Lendecke <vl@samba.org> Reported-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Acked-by: Linus Torvalds <torvalds@linux-foundation.org> Acked-by: Jens Axboe <jens.axboe@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-02 06:26:00 +08:00
timeo = sock_rcvtimeo(sk, sock->file->f_flags & O_NONBLOCK);
while (tss.len) {
ret = __tcp_splice_read(sk, &tss);
if (ret < 0)
break;
else if (!ret) {
if (spliced)
break;
if (sock_flag(sk, SOCK_DONE))
break;
if (sk->sk_err) {
ret = sock_error(sk);
break;
}
if (sk->sk_shutdown & RCV_SHUTDOWN)
break;
if (sk->sk_state == TCP_CLOSE) {
/*
* This occurs when user tries to read
* from never connected socket.
*/
ret = -ENOTCONN;
break;
}
if (!timeo) {
ret = -EAGAIN;
break;
}
/* if __tcp_splice_read() got nothing while we have
* an skb in receive queue, we do not want to loop.
* This might happen with URG data.
*/
if (!skb_queue_empty(&sk->sk_receive_queue))
break;
ret = sk_wait_data(sk, &timeo, NULL);
if (ret < 0)
break;
if (signal_pending(current)) {
ret = sock_intr_errno(timeo);
break;
}
continue;
}
tss.len -= ret;
spliced += ret;
if (!tss.len || !timeo)
break;
release_sock(sk);
lock_sock(sk);
if (sk->sk_err || sk->sk_state == TCP_CLOSE ||
(sk->sk_shutdown & RCV_SHUTDOWN) ||
signal_pending(current))
break;
}
release_sock(sk);
if (spliced)
return spliced;
return ret;
}
EXPORT_SYMBOL(tcp_splice_read);
struct sk_buff *tcp_stream_alloc_skb(struct sock *sk, gfp_t gfp,
bool force_schedule)
{
struct sk_buff *skb;
skb = alloc_skb_fclone(MAX_TCP_HEADER, gfp);
if (likely(skb)) {
bool mem_scheduled;
net: avoid double accounting for pure zerocopy skbs Track skbs containing only zerocopy data and avoid charging them to kernel memory to correctly account the memory utilization for msg_zerocopy. All of the data in such skbs is held in user pages which are already accounted to user. Before this change, they are charged again in kernel in __zerocopy_sg_from_iter. The charging in kernel is excessive because data is not being copied into skb frags. This excessive charging can lead to kernel going into memory pressure state which impacts all sockets in the system adversely. Mark pure zerocopy skbs with a SKBFL_PURE_ZEROCOPY flag and remove charge/uncharge for data in such skbs. Initially, an skb is marked pure zerocopy when it is empty and in zerocopy path. skb can then change from a pure zerocopy skb to mixed data skb (zerocopy and copy data) if it is at tail of write queue and there is room available in it and non-zerocopy data is being sent in the next sendmsg call. At this time sk_mem_charge is done for the pure zerocopied data and the pure zerocopy flag is unmarked. We found that this happens very rarely on workloads that pass MSG_ZEROCOPY. A pure zerocopy skb can later be coalesced into normal skb if they are next to each other in queue but this patch prevents coalescing from happening. This avoids complexity of charging when skb downgrades from pure zerocopy to mixed. This is also rare. In sk_wmem_free_skb, if it is a pure zerocopy skb, an sk_mem_uncharge for SKB_TRUESIZE(skb_end_offset(skb)) is done for sk_mem_charge in tcp_skb_entail for an skb without data. Testing with the msg_zerocopy.c benchmark between two hosts(100G nics) with zerocopy showed that before this patch the 'sock' variable in memory.stat for cgroup2 that tracks sum of sk_forward_alloc, sk_rmem_alloc and sk_wmem_queued is around 1822720 and with this change it is 0. This is due to no charge to sk_forward_alloc for zerocopy data and shows memory utilization for kernel is lowered. With this commit we don't see the warning we saw in previous commit which resulted in commit 84882cf72cd774cf16fd338bdbf00f69ac9f9194. Signed-off-by: Talal Ahmad <talalahmad@google.com> Acked-by: Arjun Roy <arjunroy@google.com> Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-11-03 10:58:44 +08:00
skb->truesize = SKB_TRUESIZE(skb_end_offset(skb));
if (force_schedule) {
mem_scheduled = true;
sk_forced_mem_schedule(sk, skb->truesize);
} else {
mem_scheduled = sk_wmem_schedule(sk, skb->truesize);
}
if (likely(mem_scheduled)) {
skb_reserve(skb, MAX_TCP_HEADER);
skb->ip_summed = CHECKSUM_PARTIAL;
tcp: new list for sent but unacked skbs for RACK recovery This patch adds a new queue (list) that tracks the sent but not yet acked or SACKed skbs for a TCP connection. The list is chronologically ordered by skb->skb_mstamp (the head is the oldest sent skb). This list will be used to optimize TCP Rack recovery, which checks an skb's timestamp to judge if it has been lost and needs to be retransmitted. Since TCP write queue is ordered by sequence instead of sent time, RACK has to scan over the write queue to catch all eligible packets to detect lost retransmission, and iterates through SACKed skbs repeatedly. Special cares for rare events: 1. TCP repair fakes skb transmission so the send queue needs adjusted 2. SACK reneging would require re-inserting SACKed skbs into the send queue. For now I believe it's not worth the complexity to make RACK work perfectly on SACK reneging, so we do nothing here. 3. Fast Open: currently for non-TFO, send-queue correctly queues the pure SYN packet. For TFO which queues a pure SYN and then a data packet, send-queue only queues the data packet but not the pure SYN due to the structure of TFO code. This is okay because the SYN receiver would never respond with a SACK on a missing SYN (i.e. SYN is never fast-retransmitted by SACK/RACK). In order to not grow sk_buff, we use an union for the new list and _skb_refdst/destructor fields. This is a bit complicated because we need to make sure _skb_refdst and destructor are properly zeroed before skb is cloned/copied at transmit, and before being freed. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-05 03:59:58 +08:00
INIT_LIST_HEAD(&skb->tcp_tsorted_anchor);
return skb;
}
__kfree_skb(skb);
} else {
sk->sk_prot->enter_memory_pressure(sk);
sk_stream_moderate_sndbuf(sk);
}
return NULL;
}
static unsigned int tcp_xmit_size_goal(struct sock *sk, u32 mss_now,
int large_allowed)
{
struct tcp_sock *tp = tcp_sk(sk);
u32 new_size_goal, size_goal;
tcp: refine TSO autosizing Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to control TSO size, but did this at the wrong place (sendmsg() time) At sendmsg() time, we might have a pessimistic view of flow rate, and we end up building very small skbs (with 2 MSS per skb). This is bad because : - It sends small TSO packets even in Slow Start where rate quickly increases. - It tends to make socket write queue very big, increasing tcp_ack() processing time, but also increasing memory needs, not necessarily accounted for, as fast clones overhead is currently ignored. - Lower GRO efficiency and more ACK packets. Servers with a lot of small lived connections suffer from this. Lets instead fill skbs as much as possible (64KB of payload), but split them at xmit time, when we have a precise idea of the flow rate. skb split is actually quite efficient. Patch looks bigger than necessary, because TCP Small Queue decision now has to take place after the eventual split. As Neal suggested, introduce a new tcp_tso_autosize() helper, so that tcp_tso_should_defer() can be synchronized on same goal. Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable contains number of mss that we can put in GSO packet, and is not related to the autosizing goal anymore. Tested: 40 ms rtt link nstat >/dev/null netperf -H remote -l -2000000 -- -s 1000000 nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets" Before patch : Recv Send Send Socket Socket Message Elapsed Size Size Size Time Throughput bytes bytes bytes secs. 10^6bits/s 87380 2000000 2000000 0.36 44.22 IpInReceives 600 0.0 IpOutRequests 599 0.0 TcpOutSegs 1397 0.0 IpExtOutOctets 2033249 0.0 After patch : Recv Send Send Socket Socket Message Elapsed Size Size Size Time Throughput bytes bytes bytes secs. 10^6bits/sec 87380 2000000 2000000 0.36 44.27 IpInReceives 221 0.0 IpOutRequests 232 0.0 TcpOutSegs 1397 0.0 IpExtOutOctets 2013953 0.0 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
if (!large_allowed)
tcp: refine TSO autosizing Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to control TSO size, but did this at the wrong place (sendmsg() time) At sendmsg() time, we might have a pessimistic view of flow rate, and we end up building very small skbs (with 2 MSS per skb). This is bad because : - It sends small TSO packets even in Slow Start where rate quickly increases. - It tends to make socket write queue very big, increasing tcp_ack() processing time, but also increasing memory needs, not necessarily accounted for, as fast clones overhead is currently ignored. - Lower GRO efficiency and more ACK packets. Servers with a lot of small lived connections suffer from this. Lets instead fill skbs as much as possible (64KB of payload), but split them at xmit time, when we have a precise idea of the flow rate. skb split is actually quite efficient. Patch looks bigger than necessary, because TCP Small Queue decision now has to take place after the eventual split. As Neal suggested, introduce a new tcp_tso_autosize() helper, so that tcp_tso_should_defer() can be synchronized on same goal. Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable contains number of mss that we can put in GSO packet, and is not related to the autosizing goal anymore. Tested: 40 ms rtt link nstat >/dev/null netperf -H remote -l -2000000 -- -s 1000000 nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets" Before patch : Recv Send Send Socket Socket Message Elapsed Size Size Size Time Throughput bytes bytes bytes secs. 10^6bits/s 87380 2000000 2000000 0.36 44.22 IpInReceives 600 0.0 IpOutRequests 599 0.0 TcpOutSegs 1397 0.0 IpExtOutOctets 2033249 0.0 After patch : Recv Send Send Socket Socket Message Elapsed Size Size Size Time Throughput bytes bytes bytes secs. 10^6bits/sec 87380 2000000 2000000 0.36 44.27 IpInReceives 221 0.0 IpOutRequests 232 0.0 TcpOutSegs 1397 0.0 IpExtOutOctets 2013953 0.0 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
return mss_now;
/* Note : tcp_tso_autosize() will eventually split this later */
new_size_goal = tcp_bound_to_half_wnd(tp, sk->sk_gso_max_size);
tcp: refine TSO autosizing Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to control TSO size, but did this at the wrong place (sendmsg() time) At sendmsg() time, we might have a pessimistic view of flow rate, and we end up building very small skbs (with 2 MSS per skb). This is bad because : - It sends small TSO packets even in Slow Start where rate quickly increases. - It tends to make socket write queue very big, increasing tcp_ack() processing time, but also increasing memory needs, not necessarily accounted for, as fast clones overhead is currently ignored. - Lower GRO efficiency and more ACK packets. Servers with a lot of small lived connections suffer from this. Lets instead fill skbs as much as possible (64KB of payload), but split them at xmit time, when we have a precise idea of the flow rate. skb split is actually quite efficient. Patch looks bigger than necessary, because TCP Small Queue decision now has to take place after the eventual split. As Neal suggested, introduce a new tcp_tso_autosize() helper, so that tcp_tso_should_defer() can be synchronized on same goal. Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable contains number of mss that we can put in GSO packet, and is not related to the autosizing goal anymore. Tested: 40 ms rtt link nstat >/dev/null netperf -H remote -l -2000000 -- -s 1000000 nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets" Before patch : Recv Send Send Socket Socket Message Elapsed Size Size Size Time Throughput bytes bytes bytes secs. 10^6bits/s 87380 2000000 2000000 0.36 44.22 IpInReceives 600 0.0 IpOutRequests 599 0.0 TcpOutSegs 1397 0.0 IpExtOutOctets 2033249 0.0 After patch : Recv Send Send Socket Socket Message Elapsed Size Size Size Time Throughput bytes bytes bytes secs. 10^6bits/sec 87380 2000000 2000000 0.36 44.27 IpInReceives 221 0.0 IpOutRequests 232 0.0 TcpOutSegs 1397 0.0 IpExtOutOctets 2013953 0.0 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
/* We try hard to avoid divides here */
size_goal = tp->gso_segs * mss_now;
if (unlikely(new_size_goal < size_goal ||
new_size_goal >= size_goal + mss_now)) {
tp->gso_segs = min_t(u16, new_size_goal / mss_now,
sk->sk_gso_max_segs);
size_goal = tp->gso_segs * mss_now;
}
tcp: refine TSO autosizing Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to control TSO size, but did this at the wrong place (sendmsg() time) At sendmsg() time, we might have a pessimistic view of flow rate, and we end up building very small skbs (with 2 MSS per skb). This is bad because : - It sends small TSO packets even in Slow Start where rate quickly increases. - It tends to make socket write queue very big, increasing tcp_ack() processing time, but also increasing memory needs, not necessarily accounted for, as fast clones overhead is currently ignored. - Lower GRO efficiency and more ACK packets. Servers with a lot of small lived connections suffer from this. Lets instead fill skbs as much as possible (64KB of payload), but split them at xmit time, when we have a precise idea of the flow rate. skb split is actually quite efficient. Patch looks bigger than necessary, because TCP Small Queue decision now has to take place after the eventual split. As Neal suggested, introduce a new tcp_tso_autosize() helper, so that tcp_tso_should_defer() can be synchronized on same goal. Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable contains number of mss that we can put in GSO packet, and is not related to the autosizing goal anymore. Tested: 40 ms rtt link nstat >/dev/null netperf -H remote -l -2000000 -- -s 1000000 nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets" Before patch : Recv Send Send Socket Socket Message Elapsed Size Size Size Time Throughput bytes bytes bytes secs. 10^6bits/s 87380 2000000 2000000 0.36 44.22 IpInReceives 600 0.0 IpOutRequests 599 0.0 TcpOutSegs 1397 0.0 IpExtOutOctets 2033249 0.0 After patch : Recv Send Send Socket Socket Message Elapsed Size Size Size Time Throughput bytes bytes bytes secs. 10^6bits/sec 87380 2000000 2000000 0.36 44.27 IpInReceives 221 0.0 IpOutRequests 232 0.0 TcpOutSegs 1397 0.0 IpExtOutOctets 2013953 0.0 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
return max(size_goal, mss_now);
}
int tcp_send_mss(struct sock *sk, int *size_goal, int flags)
{
int mss_now;
mss_now = tcp_current_mss(sk);
*size_goal = tcp_xmit_size_goal(sk, mss_now, !(flags & MSG_OOB));
return mss_now;
}
/* In some cases, sendmsg() could have added an skb to the write queue,
* but failed adding payload on it. We need to remove it to consume less
* memory, but more importantly be able to generate EPOLLOUT for Edge Trigger
* epoll() users. Another reason is that tcp_write_xmit() does not like
* finding an empty skb in the write queue.
*/
void tcp_remove_empty_skb(struct sock *sk)
{
struct sk_buff *skb = tcp_write_queue_tail(sk);
tcp: don't free a FIN sk_buff in tcp_remove_empty_skb() v1: Implement a more general statement as recommended by Eric Dumazet. The sequence number will be advanced, so this check will fix the FIN case and other cases. A customer reported sockets stuck in the CLOSING state. A Vmcore revealed that the write_queue was not empty as determined by tcp_write_queue_empty() but the sk_buff containing the FIN flag had been freed and the socket was zombied in that state. Corresponding pcaps show no FIN from the Linux kernel on the wire. Some instrumentation was added to the kernel and it was found that there is a timing window where tcp_sendmsg() can run after tcp_send_fin(). tcp_sendmsg() will hit an error, for example: 1269 ▹ if (sk->sk_err || (sk->sk_shutdown & SEND_SHUTDOWN))↩ 1270 ▹ ▹ goto do_error;↩ tcp_remove_empty_skb() will then free the FIN sk_buff as "skb->len == 0". The TCP socket is now wedged in the FIN-WAIT-1 state because the FIN is never sent. If the other side sends a FIN packet the socket will transition to CLOSING and remain that way until the system is rebooted. Fix this by checking for the FIN flag in the sk_buff and don't free it if that is the case. Testing confirmed that fixed the issue. Fixes: fdfc5c8594c2 ("tcp: remove empty skb from write queue in error cases") Signed-off-by: Jon Maxwell <jmaxwell37@gmail.com> Reported-by: Monir Zouaoui <Monir.Zouaoui@mail.schwarz> Reported-by: Simon Stier <simon.stier@mail.schwarz> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-10-25 07:59:03 +08:00
if (skb && TCP_SKB_CB(skb)->seq == TCP_SKB_CB(skb)->end_seq) {
tcp_unlink_write_queue(skb, sk);
if (tcp_write_queue_empty(sk))
tcp_chrono_stop(sk, TCP_CHRONO_BUSY);
tcp_wmem_free_skb(sk, skb);
}
}
/* skb changing from pure zc to mixed, must charge zc */
static int tcp_downgrade_zcopy_pure(struct sock *sk, struct sk_buff *skb)
{
if (unlikely(skb_zcopy_pure(skb))) {
u32 extra = skb->truesize -
SKB_TRUESIZE(skb_end_offset(skb));
if (!sk_wmem_schedule(sk, extra))
return -ENOMEM;
sk_mem_charge(sk, extra);
skb_shinfo(skb)->flags &= ~SKBFL_PURE_ZEROCOPY;
}
return 0;
}
int tcp_wmem_schedule(struct sock *sk, int copy)
{
int left;
if (likely(sk_wmem_schedule(sk, copy)))
return copy;
/* We could be in trouble if we have nothing queued.
* Use whatever is left in sk->sk_forward_alloc and tcp_wmem[0]
* to guarantee some progress.
*/
left = sock_net(sk)->ipv4.sysctl_tcp_wmem[0] - sk->sk_wmem_queued;
if (left > 0)
sk_forced_mem_schedule(sk, min(left, copy));
return min(copy, sk->sk_forward_alloc);
}
void tcp_free_fastopen_req(struct tcp_sock *tp)
{
if (tp->fastopen_req) {
kfree(tp->fastopen_req);
tp->fastopen_req = NULL;
}
}
int tcp_sendmsg_fastopen(struct sock *sk, struct msghdr *msg, int *copied,
size_t size, struct ubuf_info *uarg)
{
struct tcp_sock *tp = tcp_sk(sk);
net/tcp-fastopen: Add new API support This patch adds a new socket option, TCP_FASTOPEN_CONNECT, as an alternative way to perform Fast Open on the active side (client). Prior to this patch, a client needs to replace the connect() call with sendto(MSG_FASTOPEN). This can be cumbersome for applications who want to use Fast Open: these socket operations are often done in lower layer libraries used by many other applications. Changing these libraries and/or the socket call sequences are not trivial. A more convenient approach is to perform Fast Open by simply enabling a socket option when the socket is created w/o changing other socket calls sequence: s = socket() create a new socket setsockopt(s, IPPROTO_TCP, TCP_FASTOPEN_CONNECT …); newly introduced sockopt If set, new functionality described below will be used. Return ENOTSUPP if TFO is not supported or not enabled in the kernel. connect() With cookie present, return 0 immediately. With no cookie, initiate 3WHS with TFO cookie-request option and return -1 with errno = EINPROGRESS. write()/sendmsg() With cookie present, send out SYN with data and return the number of bytes buffered. With no cookie, and 3WHS not yet completed, return -1 with errno = EINPROGRESS. No MSG_FASTOPEN flag is needed. read() Return -1 with errno = EWOULDBLOCK/EAGAIN if connect() is called but write() is not called yet. Return -1 with errno = EWOULDBLOCK/EAGAIN if connection is established but no msg is received yet. Return number of bytes read if socket is established and there is msg received. The new API simplifies life for applications that always perform a write() immediately after a successful connect(). Such applications can now take advantage of Fast Open by merely making one new setsockopt() call at the time of creating the socket. Nothing else about the application's socket call sequence needs to change. Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-24 02:59:22 +08:00
struct inet_sock *inet = inet_sk(sk);
tcp: avoid fastopen API to be used on AF_UNSPEC Fastopen API should be used to perform fastopen operations on the TCP socket. It does not make sense to use fastopen API to perform disconnect by calling it with AF_UNSPEC. The fastopen data path is also prone to race conditions and bugs when using with AF_UNSPEC. One issue reported and analyzed by Vegard Nossum is as follows: +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Thread A: Thread B: ------------------------------------------------------------------------ sendto() - tcp_sendmsg() - sk_stream_memory_free() = 0 - goto wait_for_sndbuf - sk_stream_wait_memory() - sk_wait_event() // sleep | sendto(flags=MSG_FASTOPEN, dest_addr=AF_UNSPEC) | - tcp_sendmsg() | - tcp_sendmsg_fastopen() | - __inet_stream_connect() | - tcp_disconnect() //because of AF_UNSPEC | - tcp_transmit_skb()// send RST | - return 0; // no reconnect! | - sk_stream_wait_connect() | - sock_error() | - xchg(&sk->sk_err, 0) | - return -ECONNRESET - ... // wake up, see sk->sk_err == 0 - skb_entail() on TCP_CLOSE socket If the connection is reopened then we will send a brand new SYN packet after thread A has already queued a buffer. At this point I think the socket internal state (sequence numbers etc.) becomes messed up. When the new connection is closed, the FIN-ACK is rejected because the sequence number is outside the window. The other side tries to retransmit, but __tcp_retransmit_skb() calls tcp_trim_head() on an empty skb which corrupts the skb data length and hits a BUG() in copy_and_csum_bits(). +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Hence, this patch adds a check for AF_UNSPEC in the fastopen data path and return EOPNOTSUPP to user if such case happens. Fixes: cf60af03ca4e7 ("tcp: Fast Open client - sendmsg(MSG_FASTOPEN)") Reported-by: Vegard Nossum <vegard.nossum@oracle.com> Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-05-25 00:59:31 +08:00
struct sockaddr *uaddr = msg->msg_name;
int err, flags;
if (!(READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_fastopen) &
TFO_CLIENT_ENABLE) ||
tcp: avoid fastopen API to be used on AF_UNSPEC Fastopen API should be used to perform fastopen operations on the TCP socket. It does not make sense to use fastopen API to perform disconnect by calling it with AF_UNSPEC. The fastopen data path is also prone to race conditions and bugs when using with AF_UNSPEC. One issue reported and analyzed by Vegard Nossum is as follows: +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Thread A: Thread B: ------------------------------------------------------------------------ sendto() - tcp_sendmsg() - sk_stream_memory_free() = 0 - goto wait_for_sndbuf - sk_stream_wait_memory() - sk_wait_event() // sleep | sendto(flags=MSG_FASTOPEN, dest_addr=AF_UNSPEC) | - tcp_sendmsg() | - tcp_sendmsg_fastopen() | - __inet_stream_connect() | - tcp_disconnect() //because of AF_UNSPEC | - tcp_transmit_skb()// send RST | - return 0; // no reconnect! | - sk_stream_wait_connect() | - sock_error() | - xchg(&sk->sk_err, 0) | - return -ECONNRESET - ... // wake up, see sk->sk_err == 0 - skb_entail() on TCP_CLOSE socket If the connection is reopened then we will send a brand new SYN packet after thread A has already queued a buffer. At this point I think the socket internal state (sequence numbers etc.) becomes messed up. When the new connection is closed, the FIN-ACK is rejected because the sequence number is outside the window. The other side tries to retransmit, but __tcp_retransmit_skb() calls tcp_trim_head() on an empty skb which corrupts the skb data length and hits a BUG() in copy_and_csum_bits(). +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Hence, this patch adds a check for AF_UNSPEC in the fastopen data path and return EOPNOTSUPP to user if such case happens. Fixes: cf60af03ca4e7 ("tcp: Fast Open client - sendmsg(MSG_FASTOPEN)") Reported-by: Vegard Nossum <vegard.nossum@oracle.com> Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-05-25 00:59:31 +08:00
(uaddr && msg->msg_namelen >= sizeof(uaddr->sa_family) &&
uaddr->sa_family == AF_UNSPEC))
return -EOPNOTSUPP;
if (tp->fastopen_req)
return -EALREADY; /* Another Fast Open is in progress */
tp->fastopen_req = kzalloc(sizeof(struct tcp_fastopen_request),
sk->sk_allocation);
if (unlikely(!tp->fastopen_req))
return -ENOBUFS;
tp->fastopen_req->data = msg;
tp->fastopen_req->size = size;
tp->fastopen_req->uarg = uarg;
if (inet_test_bit(DEFER_CONNECT, sk)) {
net/tcp-fastopen: Add new API support This patch adds a new socket option, TCP_FASTOPEN_CONNECT, as an alternative way to perform Fast Open on the active side (client). Prior to this patch, a client needs to replace the connect() call with sendto(MSG_FASTOPEN). This can be cumbersome for applications who want to use Fast Open: these socket operations are often done in lower layer libraries used by many other applications. Changing these libraries and/or the socket call sequences are not trivial. A more convenient approach is to perform Fast Open by simply enabling a socket option when the socket is created w/o changing other socket calls sequence: s = socket() create a new socket setsockopt(s, IPPROTO_TCP, TCP_FASTOPEN_CONNECT …); newly introduced sockopt If set, new functionality described below will be used. Return ENOTSUPP if TFO is not supported or not enabled in the kernel. connect() With cookie present, return 0 immediately. With no cookie, initiate 3WHS with TFO cookie-request option and return -1 with errno = EINPROGRESS. write()/sendmsg() With cookie present, send out SYN with data and return the number of bytes buffered. With no cookie, and 3WHS not yet completed, return -1 with errno = EINPROGRESS. No MSG_FASTOPEN flag is needed. read() Return -1 with errno = EWOULDBLOCK/EAGAIN if connect() is called but write() is not called yet. Return -1 with errno = EWOULDBLOCK/EAGAIN if connection is established but no msg is received yet. Return number of bytes read if socket is established and there is msg received. The new API simplifies life for applications that always perform a write() immediately after a successful connect(). Such applications can now take advantage of Fast Open by merely making one new setsockopt() call at the time of creating the socket. Nothing else about the application's socket call sequence needs to change. Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-24 02:59:22 +08:00
err = tcp_connect(sk);
/* Same failure procedure as in tcp_v4/6_connect */
if (err) {
tcp_set_state(sk, TCP_CLOSE);
inet->inet_dport = 0;
sk->sk_route_caps = 0;
}
}
flags = (msg->msg_flags & MSG_DONTWAIT) ? O_NONBLOCK : 0;
tcp: avoid fastopen API to be used on AF_UNSPEC Fastopen API should be used to perform fastopen operations on the TCP socket. It does not make sense to use fastopen API to perform disconnect by calling it with AF_UNSPEC. The fastopen data path is also prone to race conditions and bugs when using with AF_UNSPEC. One issue reported and analyzed by Vegard Nossum is as follows: +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Thread A: Thread B: ------------------------------------------------------------------------ sendto() - tcp_sendmsg() - sk_stream_memory_free() = 0 - goto wait_for_sndbuf - sk_stream_wait_memory() - sk_wait_event() // sleep | sendto(flags=MSG_FASTOPEN, dest_addr=AF_UNSPEC) | - tcp_sendmsg() | - tcp_sendmsg_fastopen() | - __inet_stream_connect() | - tcp_disconnect() //because of AF_UNSPEC | - tcp_transmit_skb()// send RST | - return 0; // no reconnect! | - sk_stream_wait_connect() | - sock_error() | - xchg(&sk->sk_err, 0) | - return -ECONNRESET - ... // wake up, see sk->sk_err == 0 - skb_entail() on TCP_CLOSE socket If the connection is reopened then we will send a brand new SYN packet after thread A has already queued a buffer. At this point I think the socket internal state (sequence numbers etc.) becomes messed up. When the new connection is closed, the FIN-ACK is rejected because the sequence number is outside the window. The other side tries to retransmit, but __tcp_retransmit_skb() calls tcp_trim_head() on an empty skb which corrupts the skb data length and hits a BUG() in copy_and_csum_bits(). +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Hence, this patch adds a check for AF_UNSPEC in the fastopen data path and return EOPNOTSUPP to user if such case happens. Fixes: cf60af03ca4e7 ("tcp: Fast Open client - sendmsg(MSG_FASTOPEN)") Reported-by: Vegard Nossum <vegard.nossum@oracle.com> Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-05-25 00:59:31 +08:00
err = __inet_stream_connect(sk->sk_socket, uaddr,
net/tcp-fastopen: make connect()'s return case more consistent with non-TFO Without TFO, any subsequent connect() call after a successful one returns -1 EISCONN. The last API update ensured that __inet_stream_connect() can return -1 EINPROGRESS in response to sendmsg() when TFO is in use to indicate that the connection is now in progress. Unfortunately since this function is used both for connect() and sendmsg(), it has the undesired side effect of making connect() now return -1 EINPROGRESS as well after a successful call, while at the same time poll() returns POLLOUT. This can confuse some applications which happen to call connect() and to check for -1 EISCONN to ensure the connection is usable, and for which EINPROGRESS indicates a need to poll, causing a loop. This problem was encountered in haproxy where a call to connect() is precisely used in certain cases to confirm a connection's readiness. While arguably haproxy's behaviour should be improved here, it seems important to aim at a more robust behaviour when the goal of the new API is to make it easier to implement TFO in existing applications. This patch simply ensures that we preserve the same semantics as in the non-TFO case on the connect() syscall when using TFO, while still returning -1 EINPROGRESS on sendmsg(). For this we simply tell __inet_stream_connect() whether we're doing a regular connect() or in fact connecting for a sendmsg() call. Cc: Wei Wang <weiwan@google.com> Cc: Yuchung Cheng <ycheng@google.com> Cc: Eric Dumazet <edumazet@google.com> Signed-off-by: Willy Tarreau <w@1wt.eu> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-25 21:42:46 +08:00
msg->msg_namelen, flags, 1);
/* fastopen_req could already be freed in __inet_stream_connect
* if the connection times out or gets rst
*/
if (tp->fastopen_req) {
*copied = tp->fastopen_req->copied;
tcp_free_fastopen_req(tp);
inet_clear_bit(DEFER_CONNECT, sk);
}
return err;
}
int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
{
struct tcp_sock *tp = tcp_sk(sk);
struct ubuf_info *uarg = NULL;
struct sk_buff *skb;
struct sockcm_cookie sockc;
int flags, err, copied = 0;
int mss_now = 0, size_goal, copied_syn = 0;
int process_backlog = 0;
int zc = 0;
long timeo;
flags = msg->msg_flags;
if ((flags & MSG_ZEROCOPY) && size) {
if (msg->msg_ubuf) {
uarg = msg->msg_ubuf;
if (sk->sk_route_caps & NETIF_F_SG)
zc = MSG_ZEROCOPY;
} else if (sock_flag(sk, SOCK_ZEROCOPY)) {
skb = tcp_write_queue_tail(sk);
uarg = msg_zerocopy_realloc(sk, size, skb_zcopy(skb));
if (!uarg) {
err = -ENOBUFS;
goto out_err;
}
if (sk->sk_route_caps & NETIF_F_SG)
zc = MSG_ZEROCOPY;
else
uarg_to_msgzc(uarg)->zerocopy = 0;
}
} else if (unlikely(msg->msg_flags & MSG_SPLICE_PAGES) && size) {
if (sk->sk_route_caps & NETIF_F_SG)
zc = MSG_SPLICE_PAGES;
}
if (unlikely(flags & MSG_FASTOPEN ||
inet_test_bit(DEFER_CONNECT, sk)) &&
!tp->repair) {
err = tcp_sendmsg_fastopen(sk, msg, &copied_syn, size, uarg);
if (err == -EINPROGRESS && copied_syn > 0)
goto out;
else if (err)
goto out_err;
}
timeo = sock_sndtimeo(sk, flags & MSG_DONTWAIT);
tcp: track application-limited rate samples This commit adds code to track whether the delivery rate represented by each rate_sample was limited by the application. Upon each transmit, we store in the is_app_limited field in the skb a boolean bit indicating whether there is a known "bubble in the pipe": a point in the rate sample interval where the sender was application-limited, and did not transmit even though the cwnd and pacing rate allowed it. This logic marks the flow app-limited on a write if *all* of the following are true: 1) There is less than 1 MSS of unsent data in the write queue available to transmit. 2) There is no packet in the sender's queues (e.g. in fq or the NIC tx queue). 3) The connection is not limited by cwnd. 4) There are no lost packets to retransmit. The tcp_rate_check_app_limited() code in tcp_rate.c determines whether the connection is application-limited at the moment. If the flow is application-limited, it sets the tp->app_limited field. If the flow is application-limited then that means there is effectively a "bubble" of silence in the pipe now, and this silence will be reflected in a lower bandwidth sample for any rate samples from now until we get an ACK indicating this bubble has exited the pipe: specifically, until we get an ACK for the next packet we transmit. When we send every skb we record in scb->tx.is_app_limited whether the resulting rate sample will be application-limited. The code in tcp_rate_gen() checks to see when it is safe to mark all known application-limited bubbles of silence as having exited the pipe. It does this by checking to see when the delivered count moves past the tp->app_limited marker. At this point it zeroes the tp->app_limited marker, as all known bubbles are out of the pipe. We make room for the tx.is_app_limited bit in the skb by borrowing a bit from the in_flight field used by NV to record the number of bytes in flight. The receive window in the TCP header is 16 bits, and the max receive window scaling shift factor is 14 (RFC 1323). So the max receive window offered by the TCP protocol is 2^(16+14) = 2^30. So we only need 30 bits for the tx.in_flight used by NV. Signed-off-by: Van Jacobson <vanj@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Nandita Dukkipati <nanditad@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-09-20 11:39:15 +08:00
tcp_rate_check_app_limited(sk); /* is sending application-limited? */
/* Wait for a connection to finish. One exception is TCP Fast Open
* (passive side) where data is allowed to be sent before a connection
* is fully established.
*/
if (((1 << sk->sk_state) & ~(TCPF_ESTABLISHED | TCPF_CLOSE_WAIT)) &&
!tcp_passive_fastopen(sk)) {
err = sk_stream_wait_connect(sk, &timeo);
if (err != 0)
goto do_error;
}
tcp: Repair socket queues Reading queues under repair mode is done with recvmsg call. The queue-under-repair set by TCP_REPAIR_QUEUE option is used to determine which queue should be read. Thus both send and receive queue can be read with this. Caller must pass the MSG_PEEK flag. Writing to queues is done with sendmsg call and yet again -- the repair-queue option can be used to push data into the receive queue. When putting an skb into receive queue a zero tcp header is appented to its head to address the tcp_hdr(skb)->syn and the ->fin checks by the (after repair) tcp_recvmsg. These flags flags are both set to zero and that's why. The fin cannot be met in the queue while reading the source socket, since the repair only works for closed/established sockets and queueing fin packet always changes its state. The syn in the queue denotes that the respective skb's seq is "off-by-one" as compared to the actual payload lenght. Thus, at the rcv queue refill we can just drop this flag and set the skb's sequences to precice values. When the repair mode is turned off, the write queue seqs are updated so that the whole queue is considered to be 'already sent, waiting for ACKs' (write_seq = snd_nxt <= snd_una). From the protocol POV the send queue looks like it was sent, but the data between the write_seq and snd_nxt is lost in the network. This helps to avoid another sockoption for setting the snd_nxt sequence. Leaving the whole queue in a 'not yet sent' state (as it will be after sendmsg-s) will not allow to receive any acks from the peer since the ack_seq will be after the snd_nxt. Thus even the ack for the window probe will be dropped and the connection will be 'locked' with the zero peer window. Signed-off-by: Pavel Emelyanov <xemul@parallels.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-19 11:41:01 +08:00
if (unlikely(tp->repair)) {
if (tp->repair_queue == TCP_RECV_QUEUE) {
copied = tcp_send_rcvq(sk, msg, size);
tcp: Fix divide by zero when pushing during tcp-repair When in repair-mode and TCP_RECV_QUEUE is set, we end up calling tcp_push with mss_now being 0. If data is in the send-queue and tcp_set_skb_tso_segs gets called, we crash because it will divide by mss_now: [ 347.151939] divide error: 0000 [#1] SMP [ 347.152907] Modules linked in: [ 347.152907] CPU: 1 PID: 1123 Comm: packetdrill Not tainted 3.16.0-rc2 #4 [ 347.152907] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007 [ 347.152907] task: f5b88540 ti: f3c82000 task.ti: f3c82000 [ 347.152907] EIP: 0060:[<c1601359>] EFLAGS: 00210246 CPU: 1 [ 347.152907] EIP is at tcp_set_skb_tso_segs+0x49/0xa0 [ 347.152907] EAX: 00000b67 EBX: f5acd080 ECX: 00000000 EDX: 00000000 [ 347.152907] ESI: f5a28f40 EDI: f3c88f00 EBP: f3c83d10 ESP: f3c83d00 [ 347.152907] DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068 [ 347.152907] CR0: 80050033 CR2: 083158b0 CR3: 35146000 CR4: 000006b0 [ 347.152907] Stack: [ 347.152907] c167f9d9 f5acd080 000005b4 00000002 f3c83d20 c16013e6 f3c88f00 f5acd080 [ 347.152907] f3c83da0 c1603b5a f3c83d38 c10a0188 00000000 00000000 f3c83d84 c10acc85 [ 347.152907] c1ad5ec0 00000000 00000000 c1ad679c 010003e0 00000000 00000000 f3c88fc8 [ 347.152907] Call Trace: [ 347.152907] [<c167f9d9>] ? apic_timer_interrupt+0x2d/0x34 [ 347.152907] [<c16013e6>] tcp_init_tso_segs+0x36/0x50 [ 347.152907] [<c1603b5a>] tcp_write_xmit+0x7a/0xbf0 [ 347.152907] [<c10a0188>] ? up+0x28/0x40 [ 347.152907] [<c10acc85>] ? console_unlock+0x295/0x480 [ 347.152907] [<c10ad24f>] ? vprintk_emit+0x1ef/0x4b0 [ 347.152907] [<c1605716>] __tcp_push_pending_frames+0x36/0xd0 [ 347.152907] [<c15f4860>] tcp_push+0xf0/0x120 [ 347.152907] [<c15f7641>] tcp_sendmsg+0xf1/0xbf0 [ 347.152907] [<c116d920>] ? kmem_cache_free+0xf0/0x120 [ 347.152907] [<c106a682>] ? __sigqueue_free+0x32/0x40 [ 347.152907] [<c106a682>] ? __sigqueue_free+0x32/0x40 [ 347.152907] [<c114f0f0>] ? do_wp_page+0x3e0/0x850 [ 347.152907] [<c161c36a>] inet_sendmsg+0x4a/0xb0 [ 347.152907] [<c1150269>] ? handle_mm_fault+0x709/0xfb0 [ 347.152907] [<c15a006b>] sock_aio_write+0xbb/0xd0 [ 347.152907] [<c1180b79>] do_sync_write+0x69/0xa0 [ 347.152907] [<c1181023>] vfs_write+0x123/0x160 [ 347.152907] [<c1181d55>] SyS_write+0x55/0xb0 [ 347.152907] [<c167f0d8>] sysenter_do_call+0x12/0x28 This can easily be reproduced with the following packetdrill-script (the "magic" with netem, sk_pacing and limit_output_bytes is done to prevent the kernel from pushing all segments, because hitting the limit without doing this is not so easy with packetdrill): 0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3 +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0 +0 bind(3, ..., ...) = 0 +0 listen(3, 1) = 0 +0 < S 0:0(0) win 32792 <mss 1460> +0 > S. 0:0(0) ack 1 <mss 1460> +0.1 < . 1:1(0) ack 1 win 65000 +0 accept(3, ..., ...) = 4 // This forces that not all segments of the snd-queue will be pushed +0 `tc qdisc add dev tun0 root netem delay 10ms` +0 `sysctl -w net.ipv4.tcp_limit_output_bytes=2` +0 setsockopt(4, SOL_SOCKET, 47, [2], 4) = 0 +0 write(4,...,10000) = 10000 +0 write(4,...,10000) = 10000 // Set tcp-repair stuff, particularly TCP_RECV_QUEUE +0 setsockopt(4, SOL_TCP, 19, [1], 4) = 0 +0 setsockopt(4, SOL_TCP, 20, [1], 4) = 0 // This now will make the write push the remaining segments +0 setsockopt(4, SOL_SOCKET, 47, [20000], 4) = 0 +0 `sysctl -w net.ipv4.tcp_limit_output_bytes=130000` // Now we will crash +0 write(4,...,1000) = 1000 This happens since ec3423257508 (tcp: fix retransmission in repair mode). Prior to that, the call to tcp_push was prevented by a check for tp->repair. The patch fixes it, by adding the new goto-label out_nopush. When exiting tcp_sendmsg and a push is not required, which is the case for tp->repair, we go to this label. When repairing and calling send() with TCP_RECV_QUEUE, the data is actually put in the receive-queue. So, no push is required because no data has been added to the send-queue. Cc: Andrew Vagin <avagin@openvz.org> Cc: Pavel Emelyanov <xemul@parallels.com> Fixes: ec3423257508 (tcp: fix retransmission in repair mode) Signed-off-by: Christoph Paasch <christoph.paasch@uclouvain.be> Acked-by: Andrew Vagin <avagin@openvz.org> Acked-by: Pavel Emelyanov <xemul@parallels.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-29 00:26:37 +08:00
goto out_nopush;
tcp: Repair socket queues Reading queues under repair mode is done with recvmsg call. The queue-under-repair set by TCP_REPAIR_QUEUE option is used to determine which queue should be read. Thus both send and receive queue can be read with this. Caller must pass the MSG_PEEK flag. Writing to queues is done with sendmsg call and yet again -- the repair-queue option can be used to push data into the receive queue. When putting an skb into receive queue a zero tcp header is appented to its head to address the tcp_hdr(skb)->syn and the ->fin checks by the (after repair) tcp_recvmsg. These flags flags are both set to zero and that's why. The fin cannot be met in the queue while reading the source socket, since the repair only works for closed/established sockets and queueing fin packet always changes its state. The syn in the queue denotes that the respective skb's seq is "off-by-one" as compared to the actual payload lenght. Thus, at the rcv queue refill we can just drop this flag and set the skb's sequences to precice values. When the repair mode is turned off, the write queue seqs are updated so that the whole queue is considered to be 'already sent, waiting for ACKs' (write_seq = snd_nxt <= snd_una). From the protocol POV the send queue looks like it was sent, but the data between the write_seq and snd_nxt is lost in the network. This helps to avoid another sockoption for setting the snd_nxt sequence. Leaving the whole queue in a 'not yet sent' state (as it will be after sendmsg-s) will not allow to receive any acks from the peer since the ack_seq will be after the snd_nxt. Thus even the ack for the window probe will be dropped and the connection will be 'locked' with the zero peer window. Signed-off-by: Pavel Emelyanov <xemul@parallels.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-19 11:41:01 +08:00
}
err = -EINVAL;
if (tp->repair_queue == TCP_NO_QUEUE)
goto out_err;
/* 'common' sending to sendq */
}
sockcm_init(&sockc, sk);
if (msg->msg_controllen) {
err = sock_cmsg_send(sk, msg, &sockc);
if (unlikely(err)) {
err = -EINVAL;
goto out_err;
}
}
/* This should be in poll */
sk_clear_bit(SOCKWQ_ASYNC_NOSPACE, sk);
/* Ok commence sending. */
copied = 0;
tcp: make tcp_sendmsg() aware of socket backlog Large sendmsg()/write() hold socket lock for the duration of the call, unless sk->sk_sndbuf limit is hit. This is bad because incoming packets are parked into socket backlog for a long time. Critical decisions like fast retransmit might be delayed. Receivers have to maintain a big out of order queue with additional cpu overhead, and also possible stalls in TX once windows are full. Bidirectional flows are particularly hurt since the backlog can become quite big if the copy from user space triggers IO (page faults) Some applications learnt to use sendmsg() (or sendmmsg()) with small chunks to avoid this issue. Kernel should know better, right ? Add a generic sk_flush_backlog() helper and use it right before a new skb is allocated. Typically we put 64KB of payload per skb (unless MSG_EOR is requested) and checking socket backlog every 64KB gives good results. As a matter of fact, tests with TSO/GSO disabled give very nice results, as we manage to keep a small write queue and smaller perceived rtt. Note that sk_flush_backlog() maintains socket ownership, so is not equivalent to a {release_sock(sk); lock_sock(sk);}, to ensure implicit atomicity rules that sendmsg() was giving to (possibly buggy) applications. In this simple implementation, I chose to not call tcp_release_cb(), but we might consider this later. Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Alexei Starovoitov <ast@fb.com> Cc: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-30 05:16:53 +08:00
restart:
mss_now = tcp_send_mss(sk, &size_goal, flags);
err = -EPIPE;
if (sk->sk_err || (sk->sk_shutdown & SEND_SHUTDOWN))
goto do_error;
while (msg_data_left(msg)) {
ssize_t copy = 0;
skb = tcp_write_queue_tail(sk);
if (skb)
copy = size_goal - skb->len;
tcp: Make use of MSG_EOR in tcp_sendmsg This patch adds an eor bit to the TCP_SKB_CB. When MSG_EOR is passed to tcp_sendmsg, the eor bit will be set at the skb containing the last byte of the userland's msg. The eor bit will prevent data from appending to that skb in the future. The change in do_tcp_sendpages is to honor the eor set during the previous tcp_sendmsg(MSG_EOR) call. This patch handles the tcp_sendmsg case. The followup patches will handle other skb coalescing and fragment cases. One potential use case is to use MSG_EOR with SOF_TIMESTAMPING_TX_ACK to get a more accurate TCP ack timestamping on application protocol with multiple outgoing response messages (e.g. HTTP2). Packetdrill script for testing: ~~~~~~ +0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10` +0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1` +0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3 +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0 +0 bind(3, ..., ...) = 0 +0 listen(3, 1) = 0 0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7> 0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7> 0.200 < . 1:1(0) ack 1 win 257 0.200 accept(3, ..., ...) = 4 +0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0 0.200 write(4, ..., 14600) = 14600 0.200 sendto(4, ..., 730, MSG_EOR, ..., ...) = 730 0.200 sendto(4, ..., 730, MSG_EOR, ..., ...) = 730 0.200 > . 1:7301(7300) ack 1 0.200 > P. 7301:14601(7300) ack 1 0.300 < . 1:1(0) ack 14601 win 257 0.300 > P. 14601:15331(730) ack 1 0.300 > P. 15331:16061(730) ack 1 0.400 < . 1:1(0) ack 16061 win 257 0.400 close(4) = 0 0.400 > F. 16061:16061(0) ack 1 0.400 < F. 1:1(0) ack 16062 win 257 0.400 > . 16062:16062(0) ack 2 Signed-off-by: Martin KaFai Lau <kafai@fb.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Soheil Hassas Yeganeh <soheil@google.com> Cc: Willem de Bruijn <willemb@google.com> Cc: Yuchung Cheng <ycheng@google.com> Suggested-by: Eric Dumazet <edumazet@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-26 05:44:48 +08:00
if (copy <= 0 || !tcp_skb_can_collapse_to(skb)) {
bool first_skb;
new_segment:
if (!sk_stream_memory_free(sk))
goto wait_for_space;
if (unlikely(process_backlog >= 16)) {
process_backlog = 0;
if (sk_flush_backlog(sk))
goto restart;
}
tcp: implement rb-tree based retransmit queue Using a linear list to store all skbs in write queue has been okay for quite a while : O(N) is not too bad when N < 500. Things get messy when N is the order of 100,000 : Modern TCP stacks want 10Gbit+ of throughput even with 200 ms RTT flows. 40 ns per cache line miss means a full scan can use 4 ms, blowing away CPU caches. SACK processing often can use various hints to avoid parsing whole retransmit queue. But with high packet losses and/or high reordering, hints no longer work. Sender has to process thousands of unfriendly SACK, accumulating a huge socket backlog, burning a cpu and massively dropping packets. Using an rb-tree for retransmit queue has been avoided for years because it added complexity and overhead, but now is the time to be more resistant and say no to quadratic behavior. 1) RTX queue is no longer part of the write queue : already sent skbs are stored in one rb-tree. 2) Since reaching the head of write queue no longer needs sk->sk_send_head, we added an union of sk_send_head and tcp_rtx_queue Tested: On receiver : netem on ingress : delay 150ms 200us loss 1 GRO disabled to force stress and SACK storms. for f in `seq 1 10` do ./netperf -H lpaa6 -l30 -- -K bbr -o THROUGHPUT|tail -1 done | awk '{print $0} {sum += $0} END {printf "%7u\n",sum}' Before patch : 323.87 351.48 339.59 338.62 306.72 204.07 304.93 291.88 202.47 176.88 2840 After patch: 1700.83 2207.98 2070.17 1544.26 2114.76 2124.89 1693.14 1080.91 2216.82 1299.94 18053 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-06 13:21:27 +08:00
first_skb = tcp_rtx_and_write_queues_empty(sk);
skb = tcp_stream_alloc_skb(sk, sk->sk_allocation,
first_skb);
if (!skb)
goto wait_for_space;
process_backlog++;
tcp_skb_entail(sk, skb);
copy = size_goal;
tcp: don't use timestamp from repaired skb-s to calculate RTT (v2) We don't know right timestamp for repaired skb-s. Wrong RTT estimations isn't good, because some congestion modules heavily depends on it. This patch adds the TCPCB_REPAIRED flag, which is included in TCPCB_RETRANS. Thanks to Eric for the advice how to fix this issue. This patch fixes the warning: [ 879.562947] WARNING: CPU: 0 PID: 2825 at net/ipv4/tcp_input.c:3078 tcp_ack+0x11f5/0x1380() [ 879.567253] CPU: 0 PID: 2825 Comm: socket-tcpbuf-l Not tainted 3.16.0-next-20140811 #1 [ 879.567829] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 879.568177] 0000000000000000 00000000c532680c ffff880039643d00 ffffffff817aa2d2 [ 879.568776] 0000000000000000 ffff880039643d38 ffffffff8109afbd ffff880039d6ba80 [ 879.569386] ffff88003a449800 000000002983d6bd 0000000000000000 000000002983d6bc [ 879.569982] Call Trace: [ 879.570264] [<ffffffff817aa2d2>] dump_stack+0x4d/0x66 [ 879.570599] [<ffffffff8109afbd>] warn_slowpath_common+0x7d/0xa0 [ 879.570935] [<ffffffff8109b0ea>] warn_slowpath_null+0x1a/0x20 [ 879.571292] [<ffffffff816d0a05>] tcp_ack+0x11f5/0x1380 [ 879.571614] [<ffffffff816d10bd>] tcp_rcv_established+0x1ed/0x710 [ 879.571958] [<ffffffff816dc9da>] tcp_v4_do_rcv+0x10a/0x370 [ 879.572315] [<ffffffff81657459>] release_sock+0x89/0x1d0 [ 879.572642] [<ffffffff816c81a0>] do_tcp_setsockopt.isra.36+0x120/0x860 [ 879.573000] [<ffffffff8110a52e>] ? rcu_read_lock_held+0x6e/0x80 [ 879.573352] [<ffffffff816c8912>] tcp_setsockopt+0x32/0x40 [ 879.573678] [<ffffffff81654ac4>] sock_common_setsockopt+0x14/0x20 [ 879.574031] [<ffffffff816537b0>] SyS_setsockopt+0x80/0xf0 [ 879.574393] [<ffffffff817b40a9>] system_call_fastpath+0x16/0x1b [ 879.574730] ---[ end trace a17cbc38eb8c5c00 ]--- v2: moving setting of skb->when for repaired skb-s in tcp_write_xmit, where it's set for other skb-s. Fixes: 431a91242d8d ("tcp: timestamp SYN+DATA messages") Fixes: 740b0f1841f6 ("tcp: switch rtt estimations to usec resolution") Cc: Eric Dumazet <edumazet@google.com> Cc: Pavel Emelyanov <xemul@parallels.com> Cc: "David S. Miller" <davem@davemloft.net> Signed-off-by: Andrey Vagin <avagin@openvz.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-08-13 20:03:10 +08:00
/* All packets are restored as if they have
* already been sent. skb_mstamp_ns isn't set to
* avoid wrong rtt estimation.
*/
if (tp->repair)
TCP_SKB_CB(skb)->sacked |= TCPCB_REPAIRED;
}
/* Try to append data to the end of skb. */
if (copy > msg_data_left(msg))
copy = msg_data_left(msg);
if (zc == 0) {
bool merge = true;
int i = skb_shinfo(skb)->nr_frags;
struct page_frag *pfrag = sk_page_frag(sk);
if (!sk_page_frag_refill(sk, pfrag))
goto wait_for_space;
if (!skb_can_coalesce(skb, i, pfrag->page,
pfrag->offset)) {
if (i >= READ_ONCE(sysctl_max_skb_frags)) {
tcp_mark_push(tp, skb);
goto new_segment;
}
merge = false;
}
copy = min_t(int, copy, pfrag->size - pfrag->offset);
if (unlikely(skb_zcopy_pure(skb) || skb_zcopy_managed(skb))) {
if (tcp_downgrade_zcopy_pure(sk, skb))
goto wait_for_space;
skb_zcopy_downgrade_managed(skb);
}
copy = tcp_wmem_schedule(sk, copy);
if (!copy)
goto wait_for_space;
err = skb_copy_to_page_nocache(sk, &msg->msg_iter, skb,
pfrag->page,
pfrag->offset,
copy);
if (err)
goto do_error;
/* Update the skb. */
if (merge) {
skb_frag_size_add(&skb_shinfo(skb)->frags[i - 1], copy);
} else {
skb_fill_page_desc(skb, i, pfrag->page,
pfrag->offset, copy);
page_ref_inc(pfrag->page);
net-timestamp: TCP timestamping TCP timestamping extends SO_TIMESTAMPING to bytestreams. Bytestreams do not have a 1:1 relationship between send() buffers and network packets. The feature interprets a send call on a bytestream as a request for a timestamp for the last byte in that send() buffer. The choice corresponds to a request for a timestamp when all bytes in the buffer have been sent. That assumption depends on in-order kernel transmission. This is the common case. That said, it is possible to construct a traffic shaping tree that would result in reordering. The guarantee is strong, then, but not ironclad. This implementation supports send and sendpages (splice). GSO replaces one large packet with multiple smaller packets. This patch also copies the option into the correct smaller packet. This patch does not yet support timestamping on data in an initial TCP Fast Open SYN, because that takes a very different data path. If ID generation in ee_data is enabled, bytestream timestamps return a byte offset, instead of the packet counter for datagrams. The implementation supports a single timestamp per packet. It silenty replaces requests for previous timestamps. To avoid missing tstamps, flush the tcp queue by disabling Nagle, cork and autocork. Missing tstamps can be detected by offset when the ee_data ID is enabled. Implementation details: - On GSO, the timestamping code can be included in the main loop. I moved it into its own loop to reduce the impact on the common case to a single branch. - To avoid leaking the absolute seqno to userspace, the offset returned in ee_data must always be relative. It is an offset between an skb and sk field. The first is always set (also for GSO & ACK). The second must also never be uninitialized. Only allow the ID option on sockets in the ESTABLISHED state, for which the seqno is available. Never reset it to zero (instead, move it to the current seqno when reenabling the option). Signed-off-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-08-05 10:11:49 +08:00
}
pfrag->offset += copy;
} else if (zc == MSG_ZEROCOPY) {
net: avoid double accounting for pure zerocopy skbs Track skbs containing only zerocopy data and avoid charging them to kernel memory to correctly account the memory utilization for msg_zerocopy. All of the data in such skbs is held in user pages which are already accounted to user. Before this change, they are charged again in kernel in __zerocopy_sg_from_iter. The charging in kernel is excessive because data is not being copied into skb frags. This excessive charging can lead to kernel going into memory pressure state which impacts all sockets in the system adversely. Mark pure zerocopy skbs with a SKBFL_PURE_ZEROCOPY flag and remove charge/uncharge for data in such skbs. Initially, an skb is marked pure zerocopy when it is empty and in zerocopy path. skb can then change from a pure zerocopy skb to mixed data skb (zerocopy and copy data) if it is at tail of write queue and there is room available in it and non-zerocopy data is being sent in the next sendmsg call. At this time sk_mem_charge is done for the pure zerocopied data and the pure zerocopy flag is unmarked. We found that this happens very rarely on workloads that pass MSG_ZEROCOPY. A pure zerocopy skb can later be coalesced into normal skb if they are next to each other in queue but this patch prevents coalescing from happening. This avoids complexity of charging when skb downgrades from pure zerocopy to mixed. This is also rare. In sk_wmem_free_skb, if it is a pure zerocopy skb, an sk_mem_uncharge for SKB_TRUESIZE(skb_end_offset(skb)) is done for sk_mem_charge in tcp_skb_entail for an skb without data. Testing with the msg_zerocopy.c benchmark between two hosts(100G nics) with zerocopy showed that before this patch the 'sock' variable in memory.stat for cgroup2 that tracks sum of sk_forward_alloc, sk_rmem_alloc and sk_wmem_queued is around 1822720 and with this change it is 0. This is due to no charge to sk_forward_alloc for zerocopy data and shows memory utilization for kernel is lowered. With this commit we don't see the warning we saw in previous commit which resulted in commit 84882cf72cd774cf16fd338bdbf00f69ac9f9194. Signed-off-by: Talal Ahmad <talalahmad@google.com> Acked-by: Arjun Roy <arjunroy@google.com> Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-11-03 10:58:44 +08:00
/* First append to a fragless skb builds initial
* pure zerocopy skb
*/
if (!skb->len)
skb_shinfo(skb)->flags |= SKBFL_PURE_ZEROCOPY;
if (!skb_zcopy_pure(skb)) {
copy = tcp_wmem_schedule(sk, copy);
if (!copy)
net: avoid double accounting for pure zerocopy skbs Track skbs containing only zerocopy data and avoid charging them to kernel memory to correctly account the memory utilization for msg_zerocopy. All of the data in such skbs is held in user pages which are already accounted to user. Before this change, they are charged again in kernel in __zerocopy_sg_from_iter. The charging in kernel is excessive because data is not being copied into skb frags. This excessive charging can lead to kernel going into memory pressure state which impacts all sockets in the system adversely. Mark pure zerocopy skbs with a SKBFL_PURE_ZEROCOPY flag and remove charge/uncharge for data in such skbs. Initially, an skb is marked pure zerocopy when it is empty and in zerocopy path. skb can then change from a pure zerocopy skb to mixed data skb (zerocopy and copy data) if it is at tail of write queue and there is room available in it and non-zerocopy data is being sent in the next sendmsg call. At this time sk_mem_charge is done for the pure zerocopied data and the pure zerocopy flag is unmarked. We found that this happens very rarely on workloads that pass MSG_ZEROCOPY. A pure zerocopy skb can later be coalesced into normal skb if they are next to each other in queue but this patch prevents coalescing from happening. This avoids complexity of charging when skb downgrades from pure zerocopy to mixed. This is also rare. In sk_wmem_free_skb, if it is a pure zerocopy skb, an sk_mem_uncharge for SKB_TRUESIZE(skb_end_offset(skb)) is done for sk_mem_charge in tcp_skb_entail for an skb without data. Testing with the msg_zerocopy.c benchmark between two hosts(100G nics) with zerocopy showed that before this patch the 'sock' variable in memory.stat for cgroup2 that tracks sum of sk_forward_alloc, sk_rmem_alloc and sk_wmem_queued is around 1822720 and with this change it is 0. This is due to no charge to sk_forward_alloc for zerocopy data and shows memory utilization for kernel is lowered. With this commit we don't see the warning we saw in previous commit which resulted in commit 84882cf72cd774cf16fd338bdbf00f69ac9f9194. Signed-off-by: Talal Ahmad <talalahmad@google.com> Acked-by: Arjun Roy <arjunroy@google.com> Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-11-03 10:58:44 +08:00
goto wait_for_space;
}
err = skb_zerocopy_iter_stream(sk, skb, msg, copy, uarg);
if (err == -EMSGSIZE || err == -EEXIST) {
tcp_mark_push(tp, skb);
goto new_segment;
}
if (err < 0)
goto do_error;
copy = err;
} else if (zc == MSG_SPLICE_PAGES) {
/* Splice in data if we can; copy if we can't. */
if (tcp_downgrade_zcopy_pure(sk, skb))
goto wait_for_space;
copy = tcp_wmem_schedule(sk, copy);
if (!copy)
goto wait_for_space;
err = skb_splice_from_iter(skb, &msg->msg_iter, copy,
sk->sk_allocation);
if (err < 0) {
if (err == -EMSGSIZE) {
tcp_mark_push(tp, skb);
goto new_segment;
}
goto do_error;
}
copy = err;
if (!(flags & MSG_NO_SHARED_FRAGS))
skb_shinfo(skb)->flags |= SKBFL_SHARED_FRAG;
sk_wmem_queued_add(sk, copy);
sk_mem_charge(sk, copy);
}
if (!copied)
TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_PSH;
WRITE_ONCE(tp->write_seq, tp->write_seq + copy);
TCP_SKB_CB(skb)->end_seq += copy;
tcp_skb_pcount_set(skb, 0);
copied += copy;
if (!msg_data_left(msg)) {
tcp: Make use of MSG_EOR in tcp_sendmsg This patch adds an eor bit to the TCP_SKB_CB. When MSG_EOR is passed to tcp_sendmsg, the eor bit will be set at the skb containing the last byte of the userland's msg. The eor bit will prevent data from appending to that skb in the future. The change in do_tcp_sendpages is to honor the eor set during the previous tcp_sendmsg(MSG_EOR) call. This patch handles the tcp_sendmsg case. The followup patches will handle other skb coalescing and fragment cases. One potential use case is to use MSG_EOR with SOF_TIMESTAMPING_TX_ACK to get a more accurate TCP ack timestamping on application protocol with multiple outgoing response messages (e.g. HTTP2). Packetdrill script for testing: ~~~~~~ +0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10` +0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1` +0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3 +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0 +0 bind(3, ..., ...) = 0 +0 listen(3, 1) = 0 0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7> 0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7> 0.200 < . 1:1(0) ack 1 win 257 0.200 accept(3, ..., ...) = 4 +0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0 0.200 write(4, ..., 14600) = 14600 0.200 sendto(4, ..., 730, MSG_EOR, ..., ...) = 730 0.200 sendto(4, ..., 730, MSG_EOR, ..., ...) = 730 0.200 > . 1:7301(7300) ack 1 0.200 > P. 7301:14601(7300) ack 1 0.300 < . 1:1(0) ack 14601 win 257 0.300 > P. 14601:15331(730) ack 1 0.300 > P. 15331:16061(730) ack 1 0.400 < . 1:1(0) ack 16061 win 257 0.400 close(4) = 0 0.400 > F. 16061:16061(0) ack 1 0.400 < F. 1:1(0) ack 16062 win 257 0.400 > . 16062:16062(0) ack 2 Signed-off-by: Martin KaFai Lau <kafai@fb.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Soheil Hassas Yeganeh <soheil@google.com> Cc: Willem de Bruijn <willemb@google.com> Cc: Yuchung Cheng <ycheng@google.com> Suggested-by: Eric Dumazet <edumazet@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-26 05:44:48 +08:00
if (unlikely(flags & MSG_EOR))
TCP_SKB_CB(skb)->eor = 1;
goto out;
}
if (skb->len < size_goal || (flags & MSG_OOB) || unlikely(tp->repair))
continue;
if (forced_push(tp)) {
tcp_mark_push(tp, skb);
__tcp_push_pending_frames(sk, mss_now, TCP_NAGLE_PUSH);
} else if (skb == tcp_send_head(sk))
tcp_push_one(sk, mss_now);
continue;
wait_for_space:
set_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
tcp_remove_empty_skb(sk);
if (copied)
tcp_push(sk, flags & ~MSG_MORE, mss_now,
TCP_NAGLE_PUSH, size_goal);
err = sk_stream_wait_memory(sk, &timeo);
if (err != 0)
goto do_error;
mss_now = tcp_send_mss(sk, &size_goal, flags);
}
out:
if (copied) {
tcp_tx_timestamp(sk, sockc.tsflags);
tcp: auto corking With the introduction of TCP Small Queues, TSO auto sizing, and TCP pacing, we can implement Automatic Corking in the kernel, to help applications doing small write()/sendmsg() to TCP sockets. Idea is to change tcp_push() to check if the current skb payload is under skb optimal size (a multiple of MSS bytes) If under 'size_goal', and at least one packet is still in Qdisc or NIC TX queues, set the TCP Small Queue Throttled bit, so that the push will be delayed up to TX completion time. This delay might allow the application to coalesce more bytes in the skb in following write()/sendmsg()/sendfile() system calls. The exact duration of the delay is depending on the dynamics of the system, and might be zero if no packet for this flow is actually held in Qdisc or NIC TX ring. Using FQ/pacing is a way to increase the probability of autocorking being triggered. Add a new sysctl (/proc/sys/net/ipv4/tcp_autocorking) to control this feature and default it to 1 (enabled) Add a new SNMP counter : nstat -a | grep TcpExtTCPAutoCorking This counter is incremented every time we detected skb was under used and its flush was deferred. Tested: Interesting effects when using line buffered commands under ssh. Excellent performance results in term of cpu usage and total throughput. lpq83:~# echo 1 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 9410.39 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 35209.439626 task-clock # 2.901 CPUs utilized 2,294 context-switches # 0.065 K/sec 101 CPU-migrations # 0.003 K/sec 4,079 page-faults # 0.116 K/sec 97,923,241,298 cycles # 2.781 GHz [83.31%] 51,832,908,236 stalled-cycles-frontend # 52.93% frontend cycles idle [83.30%] 25,697,986,603 stalled-cycles-backend # 26.24% backend cycles idle [66.70%] 102,225,978,536 instructions # 1.04 insns per cycle # 0.51 stalled cycles per insn [83.38%] 18,657,696,819 branches # 529.906 M/sec [83.29%] 91,679,646 branch-misses # 0.49% of all branches [83.40%] 12.136204899 seconds time elapsed lpq83:~# echo 0 >/proc/sys/net/ipv4/tcp_autocorking lpq83:~# perf stat ./super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128 6624.89 Performance counter stats for './super_netperf 4 -t TCP_STREAM -H lpq84 -- -m 128': 40045.864494 task-clock # 3.301 CPUs utilized 171 context-switches # 0.004 K/sec 53 CPU-migrations # 0.001 K/sec 4,080 page-faults # 0.102 K/sec 111,340,458,645 cycles # 2.780 GHz [83.34%] 61,778,039,277 stalled-cycles-frontend # 55.49% frontend cycles idle [83.31%] 29,295,522,759 stalled-cycles-backend # 26.31% backend cycles idle [66.67%] 108,654,349,355 instructions # 0.98 insns per cycle # 0.57 stalled cycles per insn [83.34%] 19,552,170,748 branches # 488.244 M/sec [83.34%] 157,875,417 branch-misses # 0.81% of all branches [83.34%] 12.130267788 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 14:36:05 +08:00
tcp_push(sk, flags, mss_now, tp->nonagle, size_goal);
}
tcp: Fix divide by zero when pushing during tcp-repair When in repair-mode and TCP_RECV_QUEUE is set, we end up calling tcp_push with mss_now being 0. If data is in the send-queue and tcp_set_skb_tso_segs gets called, we crash because it will divide by mss_now: [ 347.151939] divide error: 0000 [#1] SMP [ 347.152907] Modules linked in: [ 347.152907] CPU: 1 PID: 1123 Comm: packetdrill Not tainted 3.16.0-rc2 #4 [ 347.152907] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007 [ 347.152907] task: f5b88540 ti: f3c82000 task.ti: f3c82000 [ 347.152907] EIP: 0060:[<c1601359>] EFLAGS: 00210246 CPU: 1 [ 347.152907] EIP is at tcp_set_skb_tso_segs+0x49/0xa0 [ 347.152907] EAX: 00000b67 EBX: f5acd080 ECX: 00000000 EDX: 00000000 [ 347.152907] ESI: f5a28f40 EDI: f3c88f00 EBP: f3c83d10 ESP: f3c83d00 [ 347.152907] DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068 [ 347.152907] CR0: 80050033 CR2: 083158b0 CR3: 35146000 CR4: 000006b0 [ 347.152907] Stack: [ 347.152907] c167f9d9 f5acd080 000005b4 00000002 f3c83d20 c16013e6 f3c88f00 f5acd080 [ 347.152907] f3c83da0 c1603b5a f3c83d38 c10a0188 00000000 00000000 f3c83d84 c10acc85 [ 347.152907] c1ad5ec0 00000000 00000000 c1ad679c 010003e0 00000000 00000000 f3c88fc8 [ 347.152907] Call Trace: [ 347.152907] [<c167f9d9>] ? apic_timer_interrupt+0x2d/0x34 [ 347.152907] [<c16013e6>] tcp_init_tso_segs+0x36/0x50 [ 347.152907] [<c1603b5a>] tcp_write_xmit+0x7a/0xbf0 [ 347.152907] [<c10a0188>] ? up+0x28/0x40 [ 347.152907] [<c10acc85>] ? console_unlock+0x295/0x480 [ 347.152907] [<c10ad24f>] ? vprintk_emit+0x1ef/0x4b0 [ 347.152907] [<c1605716>] __tcp_push_pending_frames+0x36/0xd0 [ 347.152907] [<c15f4860>] tcp_push+0xf0/0x120 [ 347.152907] [<c15f7641>] tcp_sendmsg+0xf1/0xbf0 [ 347.152907] [<c116d920>] ? kmem_cache_free+0xf0/0x120 [ 347.152907] [<c106a682>] ? __sigqueue_free+0x32/0x40 [ 347.152907] [<c106a682>] ? __sigqueue_free+0x32/0x40 [ 347.152907] [<c114f0f0>] ? do_wp_page+0x3e0/0x850 [ 347.152907] [<c161c36a>] inet_sendmsg+0x4a/0xb0 [ 347.152907] [<c1150269>] ? handle_mm_fault+0x709/0xfb0 [ 347.152907] [<c15a006b>] sock_aio_write+0xbb/0xd0 [ 347.152907] [<c1180b79>] do_sync_write+0x69/0xa0 [ 347.152907] [<c1181023>] vfs_write+0x123/0x160 [ 347.152907] [<c1181d55>] SyS_write+0x55/0xb0 [ 347.152907] [<c167f0d8>] sysenter_do_call+0x12/0x28 This can easily be reproduced with the following packetdrill-script (the "magic" with netem, sk_pacing and limit_output_bytes is done to prevent the kernel from pushing all segments, because hitting the limit without doing this is not so easy with packetdrill): 0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3 +0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0 +0 bind(3, ..., ...) = 0 +0 listen(3, 1) = 0 +0 < S 0:0(0) win 32792 <mss 1460> +0 > S. 0:0(0) ack 1 <mss 1460> +0.1 < . 1:1(0) ack 1 win 65000 +0 accept(3, ..., ...) = 4 // This forces that not all segments of the snd-queue will be pushed +0 `tc qdisc add dev tun0 root netem delay 10ms` +0 `sysctl -w net.ipv4.tcp_limit_output_bytes=2` +0 setsockopt(4, SOL_SOCKET, 47, [2], 4) = 0 +0 write(4,...,10000) = 10000 +0 write(4,...,10000) = 10000 // Set tcp-repair stuff, particularly TCP_RECV_QUEUE +0 setsockopt(4, SOL_TCP, 19, [1], 4) = 0 +0 setsockopt(4, SOL_TCP, 20, [1], 4) = 0 // This now will make the write push the remaining segments +0 setsockopt(4, SOL_SOCKET, 47, [20000], 4) = 0 +0 `sysctl -w net.ipv4.tcp_limit_output_bytes=130000` // Now we will crash +0 write(4,...,1000) = 1000 This happens since ec3423257508 (tcp: fix retransmission in repair mode). Prior to that, the call to tcp_push was prevented by a check for tp->repair. The patch fixes it, by adding the new goto-label out_nopush. When exiting tcp_sendmsg and a push is not required, which is the case for tp->repair, we go to this label. When repairing and calling send() with TCP_RECV_QUEUE, the data is actually put in the receive-queue. So, no push is required because no data has been added to the send-queue. Cc: Andrew Vagin <avagin@openvz.org> Cc: Pavel Emelyanov <xemul@parallels.com> Fixes: ec3423257508 (tcp: fix retransmission in repair mode) Signed-off-by: Christoph Paasch <christoph.paasch@uclouvain.be> Acked-by: Andrew Vagin <avagin@openvz.org> Acked-by: Pavel Emelyanov <xemul@parallels.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-29 00:26:37 +08:00
out_nopush:
/* msg->msg_ubuf is pinned by the caller so we don't take extra refs */
if (uarg && !msg->msg_ubuf)
net_zcopy_put(uarg);
return copied + copied_syn;
do_error:
tcp_remove_empty_skb(sk);
if (copied + copied_syn)
goto out;
out_err:
/* msg->msg_ubuf is pinned by the caller so we don't take extra refs */
if (uarg && !msg->msg_ubuf)
net_zcopy_put_abort(uarg, true);
err = sk_stream_error(sk, flags, err);
/* make sure we wake any epoll edge trigger waiter */
if (unlikely(tcp_rtx_and_write_queues_empty(sk) && err == -EAGAIN)) {
sk->sk_write_space(sk);
tcp_chrono_stop(sk, TCP_CHRONO_SNDBUF_LIMITED);
}
return err;
}
EXPORT_SYMBOL_GPL(tcp_sendmsg_locked);
int tcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t size)
{
int ret;
lock_sock(sk);
ret = tcp_sendmsg_locked(sk, msg, size);
release_sock(sk);
return ret;
}
EXPORT_SYMBOL(tcp_sendmsg);
void tcp_splice_eof(struct socket *sock)
{
struct sock *sk = sock->sk;
struct tcp_sock *tp = tcp_sk(sk);
int mss_now, size_goal;
if (!tcp_write_queue_tail(sk))
return;
lock_sock(sk);
mss_now = tcp_send_mss(sk, &size_goal, 0);
tcp_push(sk, 0, mss_now, tp->nonagle, size_goal);
release_sock(sk);
}
EXPORT_SYMBOL_GPL(tcp_splice_eof);
/*
* Handle reading urgent data. BSD has very simple semantics for
* this, no blocking and very strange errors 8)
*/
static int tcp_recv_urg(struct sock *sk, struct msghdr *msg, int len, int flags)
{
struct tcp_sock *tp = tcp_sk(sk);
/* No URG data to read. */
if (sock_flag(sk, SOCK_URGINLINE) || !tp->urg_data ||
tp->urg_data == TCP_URG_READ)
return -EINVAL; /* Yes this is right ! */
if (sk->sk_state == TCP_CLOSE && !sock_flag(sk, SOCK_DONE))
return -ENOTCONN;
if (tp->urg_data & TCP_URG_VALID) {
int err = 0;
char c = tp->urg_data;
if (!(flags & MSG_PEEK))
WRITE_ONCE(tp->urg_data, TCP_URG_READ);
/* Read urgent data. */
msg->msg_flags |= MSG_OOB;
if (len > 0) {
if (!(flags & MSG_TRUNC))
err = memcpy_to_msg(msg, &c, 1);
len = 1;
} else
msg->msg_flags |= MSG_TRUNC;
return err ? -EFAULT : len;
}
if (sk->sk_state == TCP_CLOSE || (sk->sk_shutdown & RCV_SHUTDOWN))
return 0;
/* Fixed the recv(..., MSG_OOB) behaviour. BSD docs and
* the available implementations agree in this case:
* this call should never block, independent of the
* blocking state of the socket.
* Mike <pall@rz.uni-karlsruhe.de>
*/
return -EAGAIN;
}
tcp: Repair socket queues Reading queues under repair mode is done with recvmsg call. The queue-under-repair set by TCP_REPAIR_QUEUE option is used to determine which queue should be read. Thus both send and receive queue can be read with this. Caller must pass the MSG_PEEK flag. Writing to queues is done with sendmsg call and yet again -- the repair-queue option can be used to push data into the receive queue. When putting an skb into receive queue a zero tcp header is appented to its head to address the tcp_hdr(skb)->syn and the ->fin checks by the (after repair) tcp_recvmsg. These flags flags are both set to zero and that's why. The fin cannot be met in the queue while reading the source socket, since the repair only works for closed/established sockets and queueing fin packet always changes its state. The syn in the queue denotes that the respective skb's seq is "off-by-one" as compared to the actual payload lenght. Thus, at the rcv queue refill we can just drop this flag and set the skb's sequences to precice values. When the repair mode is turned off, the write queue seqs are updated so that the whole queue is considered to be 'already sent, waiting for ACKs' (write_seq = snd_nxt <= snd_una). From the protocol POV the send queue looks like it was sent, but the data between the write_seq and snd_nxt is lost in the network. This helps to avoid another sockoption for setting the snd_nxt sequence. Leaving the whole queue in a 'not yet sent' state (as it will be after sendmsg-s) will not allow to receive any acks from the peer since the ack_seq will be after the snd_nxt. Thus even the ack for the window probe will be dropped and the connection will be 'locked' with the zero peer window. Signed-off-by: Pavel Emelyanov <xemul@parallels.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-19 11:41:01 +08:00
static int tcp_peek_sndq(struct sock *sk, struct msghdr *msg, int len)
{
struct sk_buff *skb;
int copied = 0, err = 0;
/* XXX -- need to support SO_PEEK_OFF */
tcp: implement rb-tree based retransmit queue Using a linear list to store all skbs in write queue has been okay for quite a while : O(N) is not too bad when N < 500. Things get messy when N is the order of 100,000 : Modern TCP stacks want 10Gbit+ of throughput even with 200 ms RTT flows. 40 ns per cache line miss means a full scan can use 4 ms, blowing away CPU caches. SACK processing often can use various hints to avoid parsing whole retransmit queue. But with high packet losses and/or high reordering, hints no longer work. Sender has to process thousands of unfriendly SACK, accumulating a huge socket backlog, burning a cpu and massively dropping packets. Using an rb-tree for retransmit queue has been avoided for years because it added complexity and overhead, but now is the time to be more resistant and say no to quadratic behavior. 1) RTX queue is no longer part of the write queue : already sent skbs are stored in one rb-tree. 2) Since reaching the head of write queue no longer needs sk->sk_send_head, we added an union of sk_send_head and tcp_rtx_queue Tested: On receiver : netem on ingress : delay 150ms 200us loss 1 GRO disabled to force stress and SACK storms. for f in `seq 1 10` do ./netperf -H lpaa6 -l30 -- -K bbr -o THROUGHPUT|tail -1 done | awk '{print $0} {sum += $0} END {printf "%7u\n",sum}' Before patch : 323.87 351.48 339.59 338.62 306.72 204.07 304.93 291.88 202.47 176.88 2840 After patch: 1700.83 2207.98 2070.17 1544.26 2114.76 2124.89 1693.14 1080.91 2216.82 1299.94 18053 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-06 13:21:27 +08:00
skb_rbtree_walk(skb, &sk->tcp_rtx_queue) {
err = skb_copy_datagram_msg(skb, 0, msg, skb->len);
if (err)
return err;
copied += skb->len;
}
tcp: Repair socket queues Reading queues under repair mode is done with recvmsg call. The queue-under-repair set by TCP_REPAIR_QUEUE option is used to determine which queue should be read. Thus both send and receive queue can be read with this. Caller must pass the MSG_PEEK flag. Writing to queues is done with sendmsg call and yet again -- the repair-queue option can be used to push data into the receive queue. When putting an skb into receive queue a zero tcp header is appented to its head to address the tcp_hdr(skb)->syn and the ->fin checks by the (after repair) tcp_recvmsg. These flags flags are both set to zero and that's why. The fin cannot be met in the queue while reading the source socket, since the repair only works for closed/established sockets and queueing fin packet always changes its state. The syn in the queue denotes that the respective skb's seq is "off-by-one" as compared to the actual payload lenght. Thus, at the rcv queue refill we can just drop this flag and set the skb's sequences to precice values. When the repair mode is turned off, the write queue seqs are updated so that the whole queue is considered to be 'already sent, waiting for ACKs' (write_seq = snd_nxt <= snd_una). From the protocol POV the send queue looks like it was sent, but the data between the write_seq and snd_nxt is lost in the network. This helps to avoid another sockoption for setting the snd_nxt sequence. Leaving the whole queue in a 'not yet sent' state (as it will be after sendmsg-s) will not allow to receive any acks from the peer since the ack_seq will be after the snd_nxt. Thus even the ack for the window probe will be dropped and the connection will be 'locked' with the zero peer window. Signed-off-by: Pavel Emelyanov <xemul@parallels.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-19 11:41:01 +08:00
skb_queue_walk(&sk->sk_write_queue, skb) {
err = skb_copy_datagram_msg(skb, 0, msg, skb->len);
tcp: Repair socket queues Reading queues under repair mode is done with recvmsg call. The queue-under-repair set by TCP_REPAIR_QUEUE option is used to determine which queue should be read. Thus both send and receive queue can be read with this. Caller must pass the MSG_PEEK flag. Writing to queues is done with sendmsg call and yet again -- the repair-queue option can be used to push data into the receive queue. When putting an skb into receive queue a zero tcp header is appented to its head to address the tcp_hdr(skb)->syn and the ->fin checks by the (after repair) tcp_recvmsg. These flags flags are both set to zero and that's why. The fin cannot be met in the queue while reading the source socket, since the repair only works for closed/established sockets and queueing fin packet always changes its state. The syn in the queue denotes that the respective skb's seq is "off-by-one" as compared to the actual payload lenght. Thus, at the rcv queue refill we can just drop this flag and set the skb's sequences to precice values. When the repair mode is turned off, the write queue seqs are updated so that the whole queue is considered to be 'already sent, waiting for ACKs' (write_seq = snd_nxt <= snd_una). From the protocol POV the send queue looks like it was sent, but the data between the write_seq and snd_nxt is lost in the network. This helps to avoid another sockoption for setting the snd_nxt sequence. Leaving the whole queue in a 'not yet sent' state (as it will be after sendmsg-s) will not allow to receive any acks from the peer since the ack_seq will be after the snd_nxt. Thus even the ack for the window probe will be dropped and the connection will be 'locked' with the zero peer window. Signed-off-by: Pavel Emelyanov <xemul@parallels.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-19 11:41:01 +08:00
if (err)
break;
copied += skb->len;
}
return err ?: copied;
}
/* Clean up the receive buffer for full frames taken by the user,
* then send an ACK if necessary. COPIED is the number of bytes
* tcp_recvmsg has given to the user so far, it speeds up the
* calculation of whether or not we must ACK for the sake of
* a window update.
*/
bpf, sockmap: Incorrectly handling copied_seq The read_skb() logic is incrementing the tcp->copied_seq which is used for among other things calculating how many outstanding bytes can be read by the application. This results in application errors, if the application does an ioctl(FIONREAD) we return zero because this is calculated from the copied_seq value. To fix this we move tcp->copied_seq accounting into the recv handler so that we update these when the recvmsg() hook is called and data is in fact copied into user buffers. This gives an accurate FIONREAD value as expected and improves ACK handling. Before we were calling the tcp_rcv_space_adjust() which would update 'number of bytes copied to user in last RTT' which is wrong for programs returning SK_PASS. The bytes are only copied to the user when recvmsg is handled. Doing the fix for recvmsg is straightforward, but fixing redirect and SK_DROP pkts is a bit tricker. Build a tcp_psock_eat() helper and then call this from skmsg handlers. This fixes another issue where a broken socket with a BPF program doing a resubmit could hang the receiver. This happened because although read_skb() consumed the skb through sock_drop() it did not update the copied_seq. Now if a single reccv socket is redirecting to many sockets (for example for lb) the receiver sk will be hung even though we might expect it to continue. The hang comes from not updating the copied_seq numbers and memory pressure resulting from that. We have a slight layer problem of calling tcp_eat_skb even if its not a TCP socket. To fix we could refactor and create per type receiver handlers. I decided this is more work than we want in the fix and we already have some small tweaks depending on caller that use the helper skb_bpf_strparser(). So we extend that a bit and always set the strparser bit when it is in use and then we can gate the seq_copied updates on this. Fixes: 04919bed948dc ("tcp: Introduce tcp_read_skb()") Signed-off-by: John Fastabend <john.fastabend@gmail.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com> Link: https://lore.kernel.org/bpf/20230523025618.113937-9-john.fastabend@gmail.com
2023-05-23 10:56:12 +08:00
void __tcp_cleanup_rbuf(struct sock *sk, int copied)
{
struct tcp_sock *tp = tcp_sk(sk);
bool time_to_ack = false;
if (inet_csk_ack_scheduled(sk)) {
const struct inet_connection_sock *icsk = inet_csk(sk);
if (/* Once-per-two-segments ACK was not sent by tcp_input.c */
tp->rcv_nxt - tp->rcv_wup > icsk->icsk_ack.rcv_mss ||
/*
* If this read emptied read buffer, we send ACK, if
* connection is not bidirectional, user drained
* receive buffer and there was a small segment
* in queue.
*/
(copied > 0 &&
((icsk->icsk_ack.pending & ICSK_ACK_PUSHED2) ||
((icsk->icsk_ack.pending & ICSK_ACK_PUSHED) &&
!inet_csk_in_pingpong_mode(sk))) &&
!atomic_read(&sk->sk_rmem_alloc)))
time_to_ack = true;
}
/* We send an ACK if we can now advertise a non-zero window
* which has been raised "significantly".
*
* Even if window raised up to infinity, do not send window open ACK
* in states, where we will not receive more. It is useless.
*/
if (copied > 0 && !time_to_ack && !(sk->sk_shutdown & RCV_SHUTDOWN)) {
__u32 rcv_window_now = tcp_receive_window(tp);
/* Optimize, __tcp_select_window() is not cheap. */
if (2*rcv_window_now <= tp->window_clamp) {
__u32 new_window = __tcp_select_window(sk);
/* Send ACK now, if this read freed lots of space
* in our buffer. Certainly, new_window is new window.
* We can advertise it now, if it is not less than current one.
* "Lots" means "at least twice" here.
*/
if (new_window && new_window >= 2 * rcv_window_now)
time_to_ack = true;
}
}
if (time_to_ack)
tcp_send_ack(sk);
}
void tcp_cleanup_rbuf(struct sock *sk, int copied)
{
struct sk_buff *skb = skb_peek(&sk->sk_receive_queue);
struct tcp_sock *tp = tcp_sk(sk);
WARN(skb && !before(tp->copied_seq, TCP_SKB_CB(skb)->end_seq),
"cleanup rbuf bug: copied %X seq %X rcvnxt %X\n",
tp->copied_seq, TCP_SKB_CB(skb)->end_seq, tp->rcv_nxt);
__tcp_cleanup_rbuf(sk, copied);
}
static void tcp_eat_recv_skb(struct sock *sk, struct sk_buff *skb)
{
__skb_unlink(skb, &sk->sk_receive_queue);
if (likely(skb->destructor == sock_rfree)) {
sock_rfree(skb);
skb->destructor = NULL;
skb->sk = NULL;
net: generalize skb freeing deferral to per-cpu lists Logic added in commit f35f821935d8 ("tcp: defer skb freeing after socket lock is released") helped bulk TCP flows to move the cost of skbs frees outside of critical section where socket lock was held. But for RPC traffic, or hosts with RFS enabled, the solution is far from being ideal. For RPC traffic, recvmsg() has to return to user space right after skb payload has been consumed, meaning that BH handler has no chance to pick the skb before recvmsg() thread. This issue is more visible with BIG TCP, as more RPC fit one skb. For RFS, even if BH handler picks the skbs, they are still picked from the cpu on which user thread is running. Ideally, it is better to free the skbs (and associated page frags) on the cpu that originally allocated them. This patch removes the per socket anchor (sk->defer_list) and instead uses a per-cpu list, which will hold more skbs per round. This new per-cpu list is drained at the end of net_action_rx(), after incoming packets have been processed, to lower latencies. In normal conditions, skbs are added to the per-cpu list with no further action. In the (unlikely) cases where the cpu does not run net_action_rx() handler fast enough, we use an IPI to raise NET_RX_SOFTIRQ on the remote cpu. Also, we do not bother draining the per-cpu list from dev_cpu_dead() This is because skbs in this list have no requirement on how fast they should be freed. Note that we can add in the future a small per-cpu cache if we see any contention on sd->defer_lock. Tested on a pair of hosts with 100Gbit NIC, RFS enabled, and /proc/sys/net/ipv4/tcp_rmem[2] tuned to 16MB to work around page recycling strategy used by NIC driver (its page pool capacity being too small compared to number of skbs/pages held in sockets receive queues) Note that this tuning was only done to demonstrate worse conditions for skb freeing for this particular test. These conditions can happen in more general production workload. 10 runs of one TCP_STREAM flow Before: Average throughput: 49685 Mbit. Kernel profiles on cpu running user thread recvmsg() show high cost for skb freeing related functions (*) 57.81% [kernel] [k] copy_user_enhanced_fast_string (*) 12.87% [kernel] [k] skb_release_data (*) 4.25% [kernel] [k] __free_one_page (*) 3.57% [kernel] [k] __list_del_entry_valid 1.85% [kernel] [k] __netif_receive_skb_core 1.60% [kernel] [k] __skb_datagram_iter (*) 1.59% [kernel] [k] free_unref_page_commit (*) 1.16% [kernel] [k] __slab_free 1.16% [kernel] [k] _copy_to_iter (*) 1.01% [kernel] [k] kfree (*) 0.88% [kernel] [k] free_unref_page 0.57% [kernel] [k] ip6_rcv_core 0.55% [kernel] [k] ip6t_do_table 0.54% [kernel] [k] flush_smp_call_function_queue (*) 0.54% [kernel] [k] free_pcppages_bulk 0.51% [kernel] [k] llist_reverse_order 0.38% [kernel] [k] process_backlog (*) 0.38% [kernel] [k] free_pcp_prepare 0.37% [kernel] [k] tcp_recvmsg_locked (*) 0.37% [kernel] [k] __list_add_valid 0.34% [kernel] [k] sock_rfree 0.34% [kernel] [k] _raw_spin_lock_irq (*) 0.33% [kernel] [k] __page_cache_release 0.33% [kernel] [k] tcp_v6_rcv (*) 0.33% [kernel] [k] __put_page (*) 0.29% [kernel] [k] __mod_zone_page_state 0.27% [kernel] [k] _raw_spin_lock After patch: Average throughput: 73076 Mbit. Kernel profiles on cpu running user thread recvmsg() looks better: 81.35% [kernel] [k] copy_user_enhanced_fast_string 1.95% [kernel] [k] _copy_to_iter 1.95% [kernel] [k] __skb_datagram_iter 1.27% [kernel] [k] __netif_receive_skb_core 1.03% [kernel] [k] ip6t_do_table 0.60% [kernel] [k] sock_rfree 0.50% [kernel] [k] tcp_v6_rcv 0.47% [kernel] [k] ip6_rcv_core 0.45% [kernel] [k] read_tsc 0.44% [kernel] [k] _raw_spin_lock_irqsave 0.37% [kernel] [k] _raw_spin_lock 0.37% [kernel] [k] native_irq_return_iret 0.33% [kernel] [k] __inet6_lookup_established 0.31% [kernel] [k] ip6_protocol_deliver_rcu 0.29% [kernel] [k] tcp_rcv_established 0.29% [kernel] [k] llist_reverse_order v2: kdoc issue (kernel bots) do not defer if (alloc_cpu == smp_processor_id()) (Paolo) replace the sk_buff_head with a single-linked list (Jakub) add a READ_ONCE()/WRITE_ONCE() for the lockless read of sd->defer_list Signed-off-by: Eric Dumazet <edumazet@google.com> Acked-by: Paolo Abeni <pabeni@redhat.com> Link: https://lore.kernel.org/r/20220422201237.416238-1-eric.dumazet@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2022-04-23 04:12:37 +08:00
return skb_attempt_defer_free(skb);
}
__kfree_skb(skb);
}
struct sk_buff *tcp_recv_skb(struct sock *sk, u32 seq, u32 *off)
{
struct sk_buff *skb;
u32 offset;
while ((skb = skb_peek(&sk->sk_receive_queue)) != NULL) {
offset = seq - TCP_SKB_CB(skb)->seq;
if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) {
pr_err_once("%s: found a SYN, please report !\n", __func__);
offset--;
}
if (offset < skb->len || (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)) {
*off = offset;
return skb;
}
/* This looks weird, but this can happen if TCP collapsing
* splitted a fat GRO packet, while we released socket lock
* in skb_splice_bits()
*/
tcp_eat_recv_skb(sk, skb);
}
return NULL;
}
EXPORT_SYMBOL(tcp_recv_skb);
/*
* This routine provides an alternative to tcp_recvmsg() for routines
* that would like to handle copying from skbuffs directly in 'sendfile'
* fashion.
* Note:
* - It is assumed that the socket was locked by the caller.
* - The routine does not block.
* - At present, there is no support for reading OOB data
* or for 'peeking' the socket using this routine
* (although both would be easy to implement).
*/
int tcp_read_sock(struct sock *sk, read_descriptor_t *desc,
sk_read_actor_t recv_actor)
{
struct sk_buff *skb;
struct tcp_sock *tp = tcp_sk(sk);
u32 seq = tp->copied_seq;
u32 offset;
int copied = 0;
if (sk->sk_state == TCP_LISTEN)
return -ENOTCONN;
while ((skb = tcp_recv_skb(sk, seq, &offset)) != NULL) {
if (offset < skb->len) {
int used;
size_t len;
len = skb->len - offset;
/* Stop reading if we hit a patch of urgent data */
if (unlikely(tp->urg_data)) {
u32 urg_offset = tp->urg_seq - seq;
if (urg_offset < len)
len = urg_offset;
if (!len)
break;
}
used = recv_actor(desc, skb, offset, len);
tcp: splice: fix an infinite loop in tcp_read_sock() commit 02275a2ee7c0 (tcp: don't abort splice() after small transfers) added a regression. [ 83.843570] INFO: rcu_sched self-detected stall on CPU [ 83.844575] INFO: rcu_sched detected stalls on CPUs/tasks: { 6} (detected by 0, t=21002 jiffies, g=4457, c=4456, q=13132) [ 83.844582] Task dump for CPU 6: [ 83.844584] netperf R running task 0 8966 8952 0x0000000c [ 83.844587] 0000000000000000 0000000000000006 0000000000006c6c 0000000000000000 [ 83.844589] 000000000000006c 0000000000000096 ffffffff819ce2bc ffffffffffffff10 [ 83.844592] ffffffff81088679 0000000000000010 0000000000000246 ffff880c4b9ddcd8 [ 83.844594] Call Trace: [ 83.844596] [<ffffffff81088679>] ? vprintk_emit+0x1c9/0x4c0 [ 83.844601] [<ffffffff815ad449>] ? schedule+0x29/0x70 [ 83.844606] [<ffffffff81537bd2>] ? tcp_splice_data_recv+0x42/0x50 [ 83.844610] [<ffffffff8153beaa>] ? tcp_read_sock+0xda/0x260 [ 83.844613] [<ffffffff81537b90>] ? tcp_prequeue_process+0xb0/0xb0 [ 83.844615] [<ffffffff8153c0f0>] ? tcp_splice_read+0xc0/0x250 [ 83.844618] [<ffffffff814dc0c2>] ? sock_splice_read+0x22/0x30 [ 83.844622] [<ffffffff811b820b>] ? do_splice_to+0x7b/0xa0 [ 83.844627] [<ffffffff811ba4bc>] ? sys_splice+0x59c/0x5d0 [ 83.844630] [<ffffffff8119745b>] ? putname+0x2b/0x40 [ 83.844633] [<ffffffff8118bcb4>] ? do_sys_open+0x174/0x1e0 [ 83.844636] [<ffffffff815b6202>] ? system_call_fastpath+0x16/0x1b if recv_actor() returns 0, we should stop immediately, because looping wont give a chance to drain the pipe. Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Willy Tarreau <w@1wt.eu> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-01-10 15:06:10 +08:00
if (used <= 0) {
if (!copied)
copied = used;
break;
}
if (WARN_ON_ONCE(used > len))
used = len;
seq += used;
copied += used;
offset += used;
/* If recv_actor drops the lock (e.g. TCP splice
* receive) the skb pointer might be invalid when
* getting here: tcp_collapse might have deleted it
* while aggregating skbs from the socket queue.
*/
skb = tcp_recv_skb(sk, seq - 1, &offset);
if (!skb)
break;
/* TCP coalescing might have appended data to the skb.
* Try to splice more frags
*/
if (offset + 1 != skb->len)
continue;
}
if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) {
tcp_eat_recv_skb(sk, skb);
++seq;
break;
}
tcp_eat_recv_skb(sk, skb);
if (!desc->count)
break;
WRITE_ONCE(tp->copied_seq, seq);
}
WRITE_ONCE(tp->copied_seq, seq);
tcp_rcv_space_adjust(sk);
/* Clean up data we have read: This will do ACK frames. */
if (copied > 0) {
tcp_recv_skb(sk, seq, &offset);
tcp_cleanup_rbuf(sk, copied);
}
return copied;
}
EXPORT_SYMBOL(tcp_read_sock);
int tcp_read_skb(struct sock *sk, skb_read_actor_t recv_actor)
{
struct sk_buff *skb;
int copied = 0;
if (sk->sk_state == TCP_LISTEN)
return -ENOTCONN;
bpf: tcp_read_skb needs to pop skb regardless of seq Before fix e5c6de5fa0258 tcp_read_skb() would increment the tp->copied-seq value. This (as described in the commit) would cause an error for apps because once that is incremented the application might believe there is no data to be read. Then some apps would stall or abort believing no data is available. However, the fix is incomplete because it introduces another issue in the skb dequeue. The loop does tcp_recv_skb() in a while loop to consume as many skbs as possible. The problem is the call is ... tcp_recv_skb(sk, seq, &offset) ... where 'seq' is: u32 seq = tp->copied_seq; Now we can hit a case where we've yet incremented copied_seq from BPF side, but then tcp_recv_skb() fails this test ... if (offset < skb->len || (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)) ... so that instead of returning the skb we call tcp_eat_recv_skb() which frees the skb. This is because the routine believes the SKB has been collapsed per comment: /* This looks weird, but this can happen if TCP collapsing * splitted a fat GRO packet, while we released socket lock * in skb_splice_bits() */ This can't happen here we've unlinked the full SKB and orphaned it. Anyways it would confuse any BPF programs if the data were suddenly moved underneath it. To fix this situation do simpler operation and just skb_peek() the data of the queue followed by the unlink. It shouldn't need to check this condition and tcp_read_skb() reads entire skbs so there is no need to handle the 'offset!=0' case as we would see in tcp_read_sock(). Fixes: e5c6de5fa0258 ("bpf, sockmap: Incorrectly handling copied_seq") Fixes: 04919bed948dc ("tcp: Introduce tcp_read_skb()") Signed-off-by: John Fastabend <john.fastabend@gmail.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com> Link: https://lore.kernel.org/bpf/20230926035300.135096-2-john.fastabend@gmail.com
2023-09-26 11:52:58 +08:00
while ((skb = skb_peek(&sk->sk_receive_queue)) != NULL) {
u8 tcp_flags;
int used;
__skb_unlink(skb, &sk->sk_receive_queue);
WARN_ON_ONCE(!skb_set_owner_sk_safe(skb, sk));
tcp_flags = TCP_SKB_CB(skb)->tcp_flags;
used = recv_actor(sk, skb);
if (used < 0) {
if (!copied)
copied = used;
break;
}
copied += used;
bpf: tcp_read_skb needs to pop skb regardless of seq Before fix e5c6de5fa0258 tcp_read_skb() would increment the tp->copied-seq value. This (as described in the commit) would cause an error for apps because once that is incremented the application might believe there is no data to be read. Then some apps would stall or abort believing no data is available. However, the fix is incomplete because it introduces another issue in the skb dequeue. The loop does tcp_recv_skb() in a while loop to consume as many skbs as possible. The problem is the call is ... tcp_recv_skb(sk, seq, &offset) ... where 'seq' is: u32 seq = tp->copied_seq; Now we can hit a case where we've yet incremented copied_seq from BPF side, but then tcp_recv_skb() fails this test ... if (offset < skb->len || (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)) ... so that instead of returning the skb we call tcp_eat_recv_skb() which frees the skb. This is because the routine believes the SKB has been collapsed per comment: /* This looks weird, but this can happen if TCP collapsing * splitted a fat GRO packet, while we released socket lock * in skb_splice_bits() */ This can't happen here we've unlinked the full SKB and orphaned it. Anyways it would confuse any BPF programs if the data were suddenly moved underneath it. To fix this situation do simpler operation and just skb_peek() the data of the queue followed by the unlink. It shouldn't need to check this condition and tcp_read_skb() reads entire skbs so there is no need to handle the 'offset!=0' case as we would see in tcp_read_sock(). Fixes: e5c6de5fa0258 ("bpf, sockmap: Incorrectly handling copied_seq") Fixes: 04919bed948dc ("tcp: Introduce tcp_read_skb()") Signed-off-by: John Fastabend <john.fastabend@gmail.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com> Link: https://lore.kernel.org/bpf/20230926035300.135096-2-john.fastabend@gmail.com
2023-09-26 11:52:58 +08:00
if (tcp_flags & TCPHDR_FIN)
break;
}
return copied;
}
EXPORT_SYMBOL(tcp_read_skb);
void tcp_read_done(struct sock *sk, size_t len)
{
struct tcp_sock *tp = tcp_sk(sk);
u32 seq = tp->copied_seq;
struct sk_buff *skb;
size_t left;
u32 offset;
if (sk->sk_state == TCP_LISTEN)
return;
left = len;
while (left && (skb = tcp_recv_skb(sk, seq, &offset)) != NULL) {
int used;
used = min_t(size_t, skb->len - offset, left);
seq += used;
left -= used;
if (skb->len > offset + used)
break;
if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) {
tcp_eat_recv_skb(sk, skb);
++seq;
break;
}
tcp_eat_recv_skb(sk, skb);
}
WRITE_ONCE(tp->copied_seq, seq);
tcp_rcv_space_adjust(sk);
/* Clean up data we have read: This will do ACK frames. */
if (left != len)
tcp_cleanup_rbuf(sk, len - left);
}
EXPORT_SYMBOL(tcp_read_done);
int tcp_peek_len(struct socket *sock)
{
return tcp_inq(sock->sk);
}
EXPORT_SYMBOL(tcp_peek_len);
/* Make sure sk_rcvbuf is big enough to satisfy SO_RCVLOWAT hint */
int tcp_set_rcvlowat(struct sock *sk, int val)
{
int space, cap;
if (sk->sk_userlocks & SOCK_RCVBUF_LOCK)
cap = sk->sk_rcvbuf >> 1;
else
cap = READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_rmem[2]) >> 1;
val = min(val, cap);
WRITE_ONCE(sk->sk_rcvlowat, val ? : 1);
/* Check if we need to signal EPOLLIN right now */
tcp_data_ready(sk);
if (sk->sk_userlocks & SOCK_RCVBUF_LOCK)
return 0;
space = tcp_space_from_win(sk, val);
if (space > sk->sk_rcvbuf) {
WRITE_ONCE(sk->sk_rcvbuf, space);
tcp_sk(sk)->window_clamp = val;
}
return 0;
}
EXPORT_SYMBOL(tcp_set_rcvlowat);
void tcp_update_recv_tstamps(struct sk_buff *skb,
struct scm_timestamping_internal *tss)
{
if (skb->tstamp)
tss->ts[0] = ktime_to_timespec64(skb->tstamp);
else
tss->ts[0] = (struct timespec64) {0};
if (skb_hwtstamps(skb)->hwtstamp)
tss->ts[2] = ktime_to_timespec64(skb_hwtstamps(skb)->hwtstamp);
else
tss->ts[2] = (struct timespec64) {0};
}
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
#ifdef CONFIG_MMU
static const struct vm_operations_struct tcp_vm_ops = {
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
};
tcp: implement mmap() for zero copy receive Some networks can make sure TCP payload can exactly fit 4KB pages, with well chosen MSS/MTU and architectures. Implement mmap() system call so that applications can avoid copying data without complex splice() games. Note that a successful mmap( X bytes) on TCP socket is consuming bytes, as if recvmsg() has been done. (tp->copied += X) Only PROT_READ mappings are accepted, as skb page frags are fundamentally shared and read only. If tcp_mmap() finds data that is not a full page, or a patch of urgent data, -EINVAL is returned, no bytes are consumed. Application must fallback to recvmsg() to read the problematic sequence. mmap() wont block, regardless of socket being in blocking or non-blocking mode. If not enough bytes are in receive queue, mmap() would return -EAGAIN, or -EIO if socket is in a state where no other bytes can be added into receive queue. An application might use SO_RCVLOWAT, poll() and/or ioctl( FIONREAD) to efficiently use mmap() On the sender side, MSG_EOR might help to clearly separate unaligned headers and 4K-aligned chunks if necessary. Tested: mlx4 (cx-3) 40Gbit NIC, with tcp_mmap program provided in following patch. MTU set to 4168 (4096 TCP payload, 40 bytes IPv6 header, 32 bytes TCP header) Without mmap() (tcp_mmap -s) received 32768 MB (0 % mmap'ed) in 8.13342 s, 33.7961 Gbit, cpu usage user:0.034 sys:3.778, 116.333 usec per MB, 63062 c-switches received 32768 MB (0 % mmap'ed) in 8.14501 s, 33.748 Gbit, cpu usage user:0.029 sys:3.997, 122.864 usec per MB, 61903 c-switches received 32768 MB (0 % mmap'ed) in 8.11723 s, 33.8635 Gbit, cpu usage user:0.048 sys:3.964, 122.437 usec per MB, 62983 c-switches received 32768 MB (0 % mmap'ed) in 8.39189 s, 32.7552 Gbit, cpu usage user:0.038 sys:4.181, 128.754 usec per MB, 55834 c-switches With mmap() on receiver (tcp_mmap -s -z) received 32768 MB (100 % mmap'ed) in 8.03083 s, 34.2278 Gbit, cpu usage user:0.024 sys:1.466, 45.4712 usec per MB, 65479 c-switches received 32768 MB (100 % mmap'ed) in 7.98805 s, 34.4111 Gbit, cpu usage user:0.026 sys:1.401, 43.5486 usec per MB, 65447 c-switches received 32768 MB (100 % mmap'ed) in 7.98377 s, 34.4296 Gbit, cpu usage user:0.028 sys:1.452, 45.166 usec per MB, 65496 c-switches received 32768 MB (99.9969 % mmap'ed) in 8.01838 s, 34.281 Gbit, cpu usage user:0.02 sys:1.446, 44.7388 usec per MB, 65505 c-switches Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-17 01:33:38 +08:00
int tcp_mmap(struct file *file, struct socket *sock,
struct vm_area_struct *vma)
{
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
if (vma->vm_flags & (VM_WRITE | VM_EXEC))
return -EPERM;
mm: replace vma->vm_flags direct modifications with modifier calls Replace direct modifications to vma->vm_flags with calls to modifier functions to be able to track flag changes and to keep vma locking correctness. [akpm@linux-foundation.org: fix drivers/misc/open-dice.c, per Hyeonggon Yoo] Link: https://lkml.kernel.org/r/20230126193752.297968-5-surenb@google.com Signed-off-by: Suren Baghdasaryan <surenb@google.com> Acked-by: Michal Hocko <mhocko@suse.com> Acked-by: Mel Gorman <mgorman@techsingularity.net> Acked-by: Mike Rapoport (IBM) <rppt@kernel.org> Acked-by: Sebastian Reichel <sebastian.reichel@collabora.com> Reviewed-by: Liam R. Howlett <Liam.Howlett@Oracle.com> Reviewed-by: Hyeonggon Yoo <42.hyeyoo@gmail.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arjun Roy <arjunroy@google.com> Cc: Axel Rasmussen <axelrasmussen@google.com> Cc: David Hildenbrand <david@redhat.com> Cc: David Howells <dhowells@redhat.com> Cc: Davidlohr Bueso <dave@stgolabs.net> Cc: David Rientjes <rientjes@google.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Greg Thelen <gthelen@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jann Horn <jannh@google.com> Cc: Joel Fernandes <joelaf@google.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Kent Overstreet <kent.overstreet@linux.dev> Cc: Laurent Dufour <ldufour@linux.ibm.com> Cc: Lorenzo Stoakes <lstoakes@gmail.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Minchan Kim <minchan@google.com> Cc: Paul E. McKenney <paulmck@kernel.org> Cc: Peter Oskolkov <posk@google.com> Cc: Peter Xu <peterx@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Punit Agrawal <punit.agrawal@bytedance.com> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: Shakeel Butt <shakeelb@google.com> Cc: Soheil Hassas Yeganeh <soheil@google.com> Cc: Song Liu <songliubraving@fb.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Will Deacon <will@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2023-01-27 03:37:49 +08:00
vm_flags_clear(vma, VM_MAYWRITE | VM_MAYEXEC);
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
/* Instruct vm_insert_page() to not mmap_read_lock(mm) */
mm: replace vma->vm_flags direct modifications with modifier calls Replace direct modifications to vma->vm_flags with calls to modifier functions to be able to track flag changes and to keep vma locking correctness. [akpm@linux-foundation.org: fix drivers/misc/open-dice.c, per Hyeonggon Yoo] Link: https://lkml.kernel.org/r/20230126193752.297968-5-surenb@google.com Signed-off-by: Suren Baghdasaryan <surenb@google.com> Acked-by: Michal Hocko <mhocko@suse.com> Acked-by: Mel Gorman <mgorman@techsingularity.net> Acked-by: Mike Rapoport (IBM) <rppt@kernel.org> Acked-by: Sebastian Reichel <sebastian.reichel@collabora.com> Reviewed-by: Liam R. Howlett <Liam.Howlett@Oracle.com> Reviewed-by: Hyeonggon Yoo <42.hyeyoo@gmail.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arjun Roy <arjunroy@google.com> Cc: Axel Rasmussen <axelrasmussen@google.com> Cc: David Hildenbrand <david@redhat.com> Cc: David Howells <dhowells@redhat.com> Cc: Davidlohr Bueso <dave@stgolabs.net> Cc: David Rientjes <rientjes@google.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Greg Thelen <gthelen@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jann Horn <jannh@google.com> Cc: Joel Fernandes <joelaf@google.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Kent Overstreet <kent.overstreet@linux.dev> Cc: Laurent Dufour <ldufour@linux.ibm.com> Cc: Lorenzo Stoakes <lstoakes@gmail.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Minchan Kim <minchan@google.com> Cc: Paul E. McKenney <paulmck@kernel.org> Cc: Peter Oskolkov <posk@google.com> Cc: Peter Xu <peterx@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Punit Agrawal <punit.agrawal@bytedance.com> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: Shakeel Butt <shakeelb@google.com> Cc: Soheil Hassas Yeganeh <soheil@google.com> Cc: Song Liu <songliubraving@fb.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Will Deacon <will@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2023-01-27 03:37:49 +08:00
vm_flags_set(vma, VM_MIXEDMAP);
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
vma->vm_ops = &tcp_vm_ops;
return 0;
}
EXPORT_SYMBOL(tcp_mmap);
static skb_frag_t *skb_advance_to_frag(struct sk_buff *skb, u32 offset_skb,
u32 *offset_frag)
{
skb_frag_t *frag;
tcp: Fix uninitialized access in skb frags array for Rx 0cp. TCP Receive zerocopy iterates through the SKB queue via tcp_recv_skb(), acquiring a pointer to an SKB and an offset within that SKB to read from. From there, it iterates the SKB frags array to determine which offset to start remapping pages from. However, this is built on the assumption that the offset read so far within the SKB is smaller than the SKB length. If this assumption is violated, we can attempt to read an invalid frags array element, which would cause a fault. tcp_recv_skb() can cause such an SKB to be returned when the TCP FIN flag is set. Therefore, we must guard against this occurrence inside skb_advance_frag(). One way that we can reproduce this error follows: 1) In a receiver program, call getsockopt(TCP_ZEROCOPY_RECEIVE) with: char some_array[32 * 1024]; struct tcp_zerocopy_receive zc = { .copybuf_address = (__u64) &some_array[0], .copybuf_len = 32 * 1024, }; 2) In a sender program, after a TCP handshake, send the following sequence of packets: i) Seq = [X, X+4000] ii) Seq = [X+4000, X+5000] iii) Seq = [X+4000, X+5000], Flags = FIN | URG, urgptr=1000 (This can happen without URG, if we have a signal pending, but URG is a convenient way to reproduce the behaviour). In this case, the following event sequence will occur on the receiver: tcp_zerocopy_receive(): -> receive_fallback_to_copy() // copybuf_len >= inq -> tcp_recvmsg_locked() // reads 5000 bytes, then breaks due to URG -> tcp_recv_skb() // yields skb with skb->len == offset -> tcp_zerocopy_set_hint_for_skb() -> skb_advance_to_frag() // will returns a frags ptr. >= nr_frags -> find_next_mappable_frag() // will dereference this bad frags ptr. With this patch, skb_advance_to_frag() will no longer return an invalid frags pointer, and will return NULL instead, fixing the issue. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Fixes: 05255b823a61 ("tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive") Link: https://lore.kernel.org/r/20211111235215.2605384-1-arjunroy.kdev@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-11-12 07:52:15 +08:00
if (unlikely(offset_skb >= skb->len))
return NULL;
offset_skb -= skb_headlen(skb);
if ((int)offset_skb < 0 || skb_has_frag_list(skb))
return NULL;
frag = skb_shinfo(skb)->frags;
while (offset_skb) {
if (skb_frag_size(frag) > offset_skb) {
*offset_frag = offset_skb;
return frag;
}
offset_skb -= skb_frag_size(frag);
++frag;
}
*offset_frag = 0;
return frag;
}
static bool can_map_frag(const skb_frag_t *frag)
{
return skb_frag_size(frag) == PAGE_SIZE && !skb_frag_off(frag);
}
static int find_next_mappable_frag(const skb_frag_t *frag,
int remaining_in_skb)
{
int offset = 0;
if (likely(can_map_frag(frag)))
return 0;
while (offset < remaining_in_skb && !can_map_frag(frag)) {
offset += skb_frag_size(frag);
++frag;
}
return offset;
}
static void tcp_zerocopy_set_hint_for_skb(struct sock *sk,
struct tcp_zerocopy_receive *zc,
struct sk_buff *skb, u32 offset)
{
u32 frag_offset, partial_frag_remainder = 0;
int mappable_offset;
skb_frag_t *frag;
/* worst case: skip to next skb. try to improve on this case below */
zc->recv_skip_hint = skb->len - offset;
/* Find the frag containing this offset (and how far into that frag) */
frag = skb_advance_to_frag(skb, offset, &frag_offset);
if (!frag)
return;
if (frag_offset) {
struct skb_shared_info *info = skb_shinfo(skb);
/* We read part of the last frag, must recvmsg() rest of skb. */
if (frag == &info->frags[info->nr_frags - 1])
return;
/* Else, we must at least read the remainder in this frag. */
partial_frag_remainder = skb_frag_size(frag) - frag_offset;
zc->recv_skip_hint -= partial_frag_remainder;
++frag;
}
/* partial_frag_remainder: If part way through a frag, must read rest.
* mappable_offset: Bytes till next mappable frag, *not* counting bytes
* in partial_frag_remainder.
*/
mappable_offset = find_next_mappable_frag(frag, zc->recv_skip_hint);
zc->recv_skip_hint = mappable_offset + partial_frag_remainder;
}
static int tcp_recvmsg_locked(struct sock *sk, struct msghdr *msg, size_t len,
int flags, struct scm_timestamping_internal *tss,
int *cmsg_flags);
static int receive_fallback_to_copy(struct sock *sk,
struct tcp_zerocopy_receive *zc, int inq,
struct scm_timestamping_internal *tss)
{
unsigned long copy_address = (unsigned long)zc->copybuf_address;
struct msghdr msg = {};
int err;
zc->length = 0;
zc->recv_skip_hint = 0;
if (copy_address != zc->copybuf_address)
return -EINVAL;
err = import_ubuf(ITER_DEST, (void __user *)copy_address, inq,
&msg.msg_iter);
if (err)
return err;
err = tcp_recvmsg_locked(sk, &msg, inq, MSG_DONTWAIT,
tss, &zc->msg_flags);
if (err < 0)
return err;
zc->copybuf_len = err;
if (likely(zc->copybuf_len)) {
struct sk_buff *skb;
u32 offset;
skb = tcp_recv_skb(sk, tcp_sk(sk)->copied_seq, &offset);
if (skb)
tcp_zerocopy_set_hint_for_skb(sk, zc, skb, offset);
}
return 0;
}
static int tcp_copy_straggler_data(struct tcp_zerocopy_receive *zc,
struct sk_buff *skb, u32 copylen,
u32 *offset, u32 *seq)
{
unsigned long copy_address = (unsigned long)zc->copybuf_address;
struct msghdr msg = {};
int err;
if (copy_address != zc->copybuf_address)
return -EINVAL;
err = import_ubuf(ITER_DEST, (void __user *)copy_address, copylen,
&msg.msg_iter);
if (err)
return err;
err = skb_copy_datagram_msg(skb, *offset, &msg, copylen);
if (err)
return err;
zc->recv_skip_hint -= copylen;
*offset += copylen;
*seq += copylen;
return (__s32)copylen;
}
static int tcp_zc_handle_leftover(struct tcp_zerocopy_receive *zc,
struct sock *sk,
struct sk_buff *skb,
u32 *seq,
s32 copybuf_len,
struct scm_timestamping_internal *tss)
{
u32 offset, copylen = min_t(u32, copybuf_len, zc->recv_skip_hint);
if (!copylen)
return 0;
/* skb is null if inq < PAGE_SIZE. */
if (skb) {
offset = *seq - TCP_SKB_CB(skb)->seq;
} else {
skb = tcp_recv_skb(sk, *seq, &offset);
if (TCP_SKB_CB(skb)->has_rxtstamp) {
tcp_update_recv_tstamps(skb, tss);
zc->msg_flags |= TCP_CMSG_TS;
}
}
zc->copybuf_len = tcp_copy_straggler_data(zc, skb, copylen, &offset,
seq);
return zc->copybuf_len < 0 ? 0 : copylen;
}
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
static int tcp_zerocopy_vm_insert_batch_error(struct vm_area_struct *vma,
struct page **pending_pages,
unsigned long pages_remaining,
unsigned long *address,
u32 *length,
u32 *seq,
struct tcp_zerocopy_receive *zc,
u32 total_bytes_to_map,
int err)
{
/* At least one page did not map. Try zapping if we skipped earlier. */
if (err == -EBUSY &&
zc->flags & TCP_RECEIVE_ZEROCOPY_FLAG_TLB_CLEAN_HINT) {
u32 maybe_zap_len;
maybe_zap_len = total_bytes_to_map - /* All bytes to map */
*length + /* Mapped or pending */
(pages_remaining * PAGE_SIZE); /* Failed map. */
mm: remove zap_page_range and create zap_vma_pages zap_page_range was originally designed to unmap pages within an address range that could span multiple vmas. While working on [1], it was discovered that all callers of zap_page_range pass a range entirely within a single vma. In addition, the mmu notification call within zap_page range does not correctly handle ranges that span multiple vmas. When crossing a vma boundary, a new mmu_notifier_range_init/end call pair with the new vma should be made. Instead of fixing zap_page_range, do the following: - Create a new routine zap_vma_pages() that will remove all pages within the passed vma. Most users of zap_page_range pass the entire vma and can use this new routine. - For callers of zap_page_range not passing the entire vma, instead call zap_page_range_single(). - Remove zap_page_range. [1] https://lore.kernel.org/linux-mm/20221114235507.294320-2-mike.kravetz@oracle.com/ Link: https://lkml.kernel.org/r/20230104002732.232573-1-mike.kravetz@oracle.com Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com> Suggested-by: Peter Xu <peterx@redhat.com> Acked-by: Michal Hocko <mhocko@suse.com> Acked-by: Peter Xu <peterx@redhat.com> Acked-by: Heiko Carstens <hca@linux.ibm.com> [s390] Reviewed-by: Christoph Hellwig <hch@lst.de> Cc: Christian Borntraeger <borntraeger@linux.ibm.com> Cc: Christian Brauner <brauner@kernel.org> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: David Hildenbrand <david@redhat.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Nadav Amit <nadav.amit@gmail.com> Cc: Palmer Dabbelt <palmer@dabbelt.com> Cc: Rik van Riel <riel@surriel.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Will Deacon <will@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2023-01-04 08:27:32 +08:00
zap_page_range_single(vma, *address, maybe_zap_len, NULL);
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
err = 0;
}
if (!err) {
unsigned long leftover_pages = pages_remaining;
int bytes_mapped;
mm: remove zap_page_range and create zap_vma_pages zap_page_range was originally designed to unmap pages within an address range that could span multiple vmas. While working on [1], it was discovered that all callers of zap_page_range pass a range entirely within a single vma. In addition, the mmu notification call within zap_page range does not correctly handle ranges that span multiple vmas. When crossing a vma boundary, a new mmu_notifier_range_init/end call pair with the new vma should be made. Instead of fixing zap_page_range, do the following: - Create a new routine zap_vma_pages() that will remove all pages within the passed vma. Most users of zap_page_range pass the entire vma and can use this new routine. - For callers of zap_page_range not passing the entire vma, instead call zap_page_range_single(). - Remove zap_page_range. [1] https://lore.kernel.org/linux-mm/20221114235507.294320-2-mike.kravetz@oracle.com/ Link: https://lkml.kernel.org/r/20230104002732.232573-1-mike.kravetz@oracle.com Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com> Suggested-by: Peter Xu <peterx@redhat.com> Acked-by: Michal Hocko <mhocko@suse.com> Acked-by: Peter Xu <peterx@redhat.com> Acked-by: Heiko Carstens <hca@linux.ibm.com> [s390] Reviewed-by: Christoph Hellwig <hch@lst.de> Cc: Christian Borntraeger <borntraeger@linux.ibm.com> Cc: Christian Brauner <brauner@kernel.org> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: David Hildenbrand <david@redhat.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Nadav Amit <nadav.amit@gmail.com> Cc: Palmer Dabbelt <palmer@dabbelt.com> Cc: Rik van Riel <riel@surriel.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Will Deacon <will@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2023-01-04 08:27:32 +08:00
/* We called zap_page_range_single, try to reinsert. */
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
err = vm_insert_pages(vma, *address,
pending_pages,
&pages_remaining);
bytes_mapped = PAGE_SIZE * (leftover_pages - pages_remaining);
*seq += bytes_mapped;
*address += bytes_mapped;
}
if (err) {
/* Either we were unable to zap, OR we zapped, retried an
* insert, and still had an issue. Either ways, pages_remaining
* is the number of pages we were unable to map, and we unroll
* some state we speculatively touched before.
*/
const int bytes_not_mapped = PAGE_SIZE * pages_remaining;
*length -= bytes_not_mapped;
zc->recv_skip_hint += bytes_not_mapped;
}
return err;
}
static int tcp_zerocopy_vm_insert_batch(struct vm_area_struct *vma,
struct page **pages,
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
unsigned int pages_to_map,
unsigned long *address,
u32 *length,
u32 *seq,
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
struct tcp_zerocopy_receive *zc,
u32 total_bytes_to_map)
{
unsigned long pages_remaining = pages_to_map;
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
unsigned int pages_mapped;
unsigned int bytes_mapped;
int err;
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
err = vm_insert_pages(vma, *address, pages, &pages_remaining);
pages_mapped = pages_to_map - (unsigned int)pages_remaining;
bytes_mapped = PAGE_SIZE * pages_mapped;
/* Even if vm_insert_pages fails, it may have partially succeeded in
* mapping (some but not all of the pages).
*/
*seq += bytes_mapped;
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
*address += bytes_mapped;
if (likely(!err))
return 0;
/* Error: maybe zap and retry + rollback state for failed inserts. */
return tcp_zerocopy_vm_insert_batch_error(vma, pages + pages_mapped,
pages_remaining, address, length, seq, zc, total_bytes_to_map,
err);
}
#define TCP_VALID_ZC_MSG_FLAGS (TCP_CMSG_TS)
static void tcp_zc_finalize_rx_tstamp(struct sock *sk,
struct tcp_zerocopy_receive *zc,
struct scm_timestamping_internal *tss)
{
unsigned long msg_control_addr;
struct msghdr cmsg_dummy;
msg_control_addr = (unsigned long)zc->msg_control;
cmsg_dummy.msg_control_user = (void __user *)msg_control_addr;
cmsg_dummy.msg_controllen =
(__kernel_size_t)zc->msg_controllen;
cmsg_dummy.msg_flags = in_compat_syscall()
? MSG_CMSG_COMPAT : 0;
cmsg_dummy.msg_control_is_user = true;
zc->msg_flags = 0;
if (zc->msg_control == msg_control_addr &&
zc->msg_controllen == cmsg_dummy.msg_controllen) {
tcp_recv_timestamp(&cmsg_dummy, sk, tss);
zc->msg_control = (__u64)
((uintptr_t)cmsg_dummy.msg_control_user);
zc->msg_controllen =
(__u64)cmsg_dummy.msg_controllen;
zc->msg_flags = (__u32)cmsg_dummy.msg_flags;
}
}
tcp: Use per-vma locking for receive zerocopy Per-VMA locking allows us to lock a struct vm_area_struct without taking the process-wide mmap lock in read mode. Consider a process workload where the mmap lock is taken constantly in write mode. In this scenario, all zerocopy receives are periodically blocked during that period of time - though in principle, the memory ranges being used by TCP are not touched by the operations that need the mmap write lock. This results in performance degradation. Now consider another workload where the mmap lock is never taken in write mode, but there are many TCP connections using receive zerocopy that are concurrently receiving. These connections all take the mmap lock in read mode, but this does induce a lot of contention and atomic ops for this process-wide lock. This results in additional CPU overhead caused by contending on the cache line for this lock. However, with per-vma locking, both of these problems can be avoided. As a test, I ran an RPC-style request/response workload with 4KB payloads and receive zerocopy enabled, with 100 simultaneous TCP connections. I measured perf cycles within the find_tcp_vma/mmap_read_lock/mmap_read_unlock codepath, with and without per-vma locking enabled. When using process-wide mmap semaphore read locking, about 1% of measured perf cycles were within this path. With per-VMA locking, this value dropped to about 0.45%. Signed-off-by: Arjun Roy <arjunroy@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-06-17 03:34:27 +08:00
static struct vm_area_struct *find_tcp_vma(struct mm_struct *mm,
unsigned long address,
bool *mmap_locked)
{
struct vm_area_struct *vma = lock_vma_under_rcu(mm, address);
tcp: Use per-vma locking for receive zerocopy Per-VMA locking allows us to lock a struct vm_area_struct without taking the process-wide mmap lock in read mode. Consider a process workload where the mmap lock is taken constantly in write mode. In this scenario, all zerocopy receives are periodically blocked during that period of time - though in principle, the memory ranges being used by TCP are not touched by the operations that need the mmap write lock. This results in performance degradation. Now consider another workload where the mmap lock is never taken in write mode, but there are many TCP connections using receive zerocopy that are concurrently receiving. These connections all take the mmap lock in read mode, but this does induce a lot of contention and atomic ops for this process-wide lock. This results in additional CPU overhead caused by contending on the cache line for this lock. However, with per-vma locking, both of these problems can be avoided. As a test, I ran an RPC-style request/response workload with 4KB payloads and receive zerocopy enabled, with 100 simultaneous TCP connections. I measured perf cycles within the find_tcp_vma/mmap_read_lock/mmap_read_unlock codepath, with and without per-vma locking enabled. When using process-wide mmap semaphore read locking, about 1% of measured perf cycles were within this path. With per-VMA locking, this value dropped to about 0.45%. Signed-off-by: Arjun Roy <arjunroy@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-06-17 03:34:27 +08:00
if (vma) {
if (vma->vm_ops != &tcp_vm_ops) {
tcp: Use per-vma locking for receive zerocopy Per-VMA locking allows us to lock a struct vm_area_struct without taking the process-wide mmap lock in read mode. Consider a process workload where the mmap lock is taken constantly in write mode. In this scenario, all zerocopy receives are periodically blocked during that period of time - though in principle, the memory ranges being used by TCP are not touched by the operations that need the mmap write lock. This results in performance degradation. Now consider another workload where the mmap lock is never taken in write mode, but there are many TCP connections using receive zerocopy that are concurrently receiving. These connections all take the mmap lock in read mode, but this does induce a lot of contention and atomic ops for this process-wide lock. This results in additional CPU overhead caused by contending on the cache line for this lock. However, with per-vma locking, both of these problems can be avoided. As a test, I ran an RPC-style request/response workload with 4KB payloads and receive zerocopy enabled, with 100 simultaneous TCP connections. I measured perf cycles within the find_tcp_vma/mmap_read_lock/mmap_read_unlock codepath, with and without per-vma locking enabled. When using process-wide mmap semaphore read locking, about 1% of measured perf cycles were within this path. With per-VMA locking, this value dropped to about 0.45%. Signed-off-by: Arjun Roy <arjunroy@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-06-17 03:34:27 +08:00
vma_end_read(vma);
return NULL;
}
*mmap_locked = false;
return vma;
}
mmap_read_lock(mm);
vma = vma_lookup(mm, address);
if (!vma || vma->vm_ops != &tcp_vm_ops) {
tcp: Use per-vma locking for receive zerocopy Per-VMA locking allows us to lock a struct vm_area_struct without taking the process-wide mmap lock in read mode. Consider a process workload where the mmap lock is taken constantly in write mode. In this scenario, all zerocopy receives are periodically blocked during that period of time - though in principle, the memory ranges being used by TCP are not touched by the operations that need the mmap write lock. This results in performance degradation. Now consider another workload where the mmap lock is never taken in write mode, but there are many TCP connections using receive zerocopy that are concurrently receiving. These connections all take the mmap lock in read mode, but this does induce a lot of contention and atomic ops for this process-wide lock. This results in additional CPU overhead caused by contending on the cache line for this lock. However, with per-vma locking, both of these problems can be avoided. As a test, I ran an RPC-style request/response workload with 4KB payloads and receive zerocopy enabled, with 100 simultaneous TCP connections. I measured perf cycles within the find_tcp_vma/mmap_read_lock/mmap_read_unlock codepath, with and without per-vma locking enabled. When using process-wide mmap semaphore read locking, about 1% of measured perf cycles were within this path. With per-VMA locking, this value dropped to about 0.45%. Signed-off-by: Arjun Roy <arjunroy@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-06-17 03:34:27 +08:00
mmap_read_unlock(mm);
return NULL;
}
*mmap_locked = true;
return vma;
}
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
#define TCP_ZEROCOPY_PAGE_BATCH_SIZE 32
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
static int tcp_zerocopy_receive(struct sock *sk,
struct tcp_zerocopy_receive *zc,
struct scm_timestamping_internal *tss)
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
{
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
u32 length = 0, offset, vma_len, avail_len, copylen = 0;
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
unsigned long address = (unsigned long)zc->address;
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
struct page *pages[TCP_ZEROCOPY_PAGE_BATCH_SIZE];
s32 copybuf_len = zc->copybuf_len;
struct tcp_sock *tp = tcp_sk(sk);
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
const skb_frag_t *frags = NULL;
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
unsigned int pages_to_map = 0;
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
struct vm_area_struct *vma;
struct sk_buff *skb = NULL;
u32 seq = tp->copied_seq;
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
u32 total_bytes_to_map;
int inq = tcp_inq(sk);
tcp: Use per-vma locking for receive zerocopy Per-VMA locking allows us to lock a struct vm_area_struct without taking the process-wide mmap lock in read mode. Consider a process workload where the mmap lock is taken constantly in write mode. In this scenario, all zerocopy receives are periodically blocked during that period of time - though in principle, the memory ranges being used by TCP are not touched by the operations that need the mmap write lock. This results in performance degradation. Now consider another workload where the mmap lock is never taken in write mode, but there are many TCP connections using receive zerocopy that are concurrently receiving. These connections all take the mmap lock in read mode, but this does induce a lot of contention and atomic ops for this process-wide lock. This results in additional CPU overhead caused by contending on the cache line for this lock. However, with per-vma locking, both of these problems can be avoided. As a test, I ran an RPC-style request/response workload with 4KB payloads and receive zerocopy enabled, with 100 simultaneous TCP connections. I measured perf cycles within the find_tcp_vma/mmap_read_lock/mmap_read_unlock codepath, with and without per-vma locking enabled. When using process-wide mmap semaphore read locking, about 1% of measured perf cycles were within this path. With per-VMA locking, this value dropped to about 0.45%. Signed-off-by: Arjun Roy <arjunroy@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-06-17 03:34:27 +08:00
bool mmap_locked;
tcp: implement mmap() for zero copy receive Some networks can make sure TCP payload can exactly fit 4KB pages, with well chosen MSS/MTU and architectures. Implement mmap() system call so that applications can avoid copying data without complex splice() games. Note that a successful mmap( X bytes) on TCP socket is consuming bytes, as if recvmsg() has been done. (tp->copied += X) Only PROT_READ mappings are accepted, as skb page frags are fundamentally shared and read only. If tcp_mmap() finds data that is not a full page, or a patch of urgent data, -EINVAL is returned, no bytes are consumed. Application must fallback to recvmsg() to read the problematic sequence. mmap() wont block, regardless of socket being in blocking or non-blocking mode. If not enough bytes are in receive queue, mmap() would return -EAGAIN, or -EIO if socket is in a state where no other bytes can be added into receive queue. An application might use SO_RCVLOWAT, poll() and/or ioctl( FIONREAD) to efficiently use mmap() On the sender side, MSG_EOR might help to clearly separate unaligned headers and 4K-aligned chunks if necessary. Tested: mlx4 (cx-3) 40Gbit NIC, with tcp_mmap program provided in following patch. MTU set to 4168 (4096 TCP payload, 40 bytes IPv6 header, 32 bytes TCP header) Without mmap() (tcp_mmap -s) received 32768 MB (0 % mmap'ed) in 8.13342 s, 33.7961 Gbit, cpu usage user:0.034 sys:3.778, 116.333 usec per MB, 63062 c-switches received 32768 MB (0 % mmap'ed) in 8.14501 s, 33.748 Gbit, cpu usage user:0.029 sys:3.997, 122.864 usec per MB, 61903 c-switches received 32768 MB (0 % mmap'ed) in 8.11723 s, 33.8635 Gbit, cpu usage user:0.048 sys:3.964, 122.437 usec per MB, 62983 c-switches received 32768 MB (0 % mmap'ed) in 8.39189 s, 32.7552 Gbit, cpu usage user:0.038 sys:4.181, 128.754 usec per MB, 55834 c-switches With mmap() on receiver (tcp_mmap -s -z) received 32768 MB (100 % mmap'ed) in 8.03083 s, 34.2278 Gbit, cpu usage user:0.024 sys:1.466, 45.4712 usec per MB, 65479 c-switches received 32768 MB (100 % mmap'ed) in 7.98805 s, 34.4111 Gbit, cpu usage user:0.026 sys:1.401, 43.5486 usec per MB, 65447 c-switches received 32768 MB (100 % mmap'ed) in 7.98377 s, 34.4296 Gbit, cpu usage user:0.028 sys:1.452, 45.166 usec per MB, 65496 c-switches received 32768 MB (99.9969 % mmap'ed) in 8.01838 s, 34.281 Gbit, cpu usage user:0.02 sys:1.446, 44.7388 usec per MB, 65505 c-switches Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-17 01:33:38 +08:00
int ret;
zc->copybuf_len = 0;
zc->msg_flags = 0;
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
if (address & (PAGE_SIZE - 1) || address != zc->address)
tcp: implement mmap() for zero copy receive Some networks can make sure TCP payload can exactly fit 4KB pages, with well chosen MSS/MTU and architectures. Implement mmap() system call so that applications can avoid copying data without complex splice() games. Note that a successful mmap( X bytes) on TCP socket is consuming bytes, as if recvmsg() has been done. (tp->copied += X) Only PROT_READ mappings are accepted, as skb page frags are fundamentally shared and read only. If tcp_mmap() finds data that is not a full page, or a patch of urgent data, -EINVAL is returned, no bytes are consumed. Application must fallback to recvmsg() to read the problematic sequence. mmap() wont block, regardless of socket being in blocking or non-blocking mode. If not enough bytes are in receive queue, mmap() would return -EAGAIN, or -EIO if socket is in a state where no other bytes can be added into receive queue. An application might use SO_RCVLOWAT, poll() and/or ioctl( FIONREAD) to efficiently use mmap() On the sender side, MSG_EOR might help to clearly separate unaligned headers and 4K-aligned chunks if necessary. Tested: mlx4 (cx-3) 40Gbit NIC, with tcp_mmap program provided in following patch. MTU set to 4168 (4096 TCP payload, 40 bytes IPv6 header, 32 bytes TCP header) Without mmap() (tcp_mmap -s) received 32768 MB (0 % mmap'ed) in 8.13342 s, 33.7961 Gbit, cpu usage user:0.034 sys:3.778, 116.333 usec per MB, 63062 c-switches received 32768 MB (0 % mmap'ed) in 8.14501 s, 33.748 Gbit, cpu usage user:0.029 sys:3.997, 122.864 usec per MB, 61903 c-switches received 32768 MB (0 % mmap'ed) in 8.11723 s, 33.8635 Gbit, cpu usage user:0.048 sys:3.964, 122.437 usec per MB, 62983 c-switches received 32768 MB (0 % mmap'ed) in 8.39189 s, 32.7552 Gbit, cpu usage user:0.038 sys:4.181, 128.754 usec per MB, 55834 c-switches With mmap() on receiver (tcp_mmap -s -z) received 32768 MB (100 % mmap'ed) in 8.03083 s, 34.2278 Gbit, cpu usage user:0.024 sys:1.466, 45.4712 usec per MB, 65479 c-switches received 32768 MB (100 % mmap'ed) in 7.98805 s, 34.4111 Gbit, cpu usage user:0.026 sys:1.401, 43.5486 usec per MB, 65447 c-switches received 32768 MB (100 % mmap'ed) in 7.98377 s, 34.4296 Gbit, cpu usage user:0.028 sys:1.452, 45.166 usec per MB, 65496 c-switches received 32768 MB (99.9969 % mmap'ed) in 8.01838 s, 34.281 Gbit, cpu usage user:0.02 sys:1.446, 44.7388 usec per MB, 65505 c-switches Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-17 01:33:38 +08:00
return -EINVAL;
if (sk->sk_state == TCP_LISTEN)
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
return -ENOTCONN;
tcp: implement mmap() for zero copy receive Some networks can make sure TCP payload can exactly fit 4KB pages, with well chosen MSS/MTU and architectures. Implement mmap() system call so that applications can avoid copying data without complex splice() games. Note that a successful mmap( X bytes) on TCP socket is consuming bytes, as if recvmsg() has been done. (tp->copied += X) Only PROT_READ mappings are accepted, as skb page frags are fundamentally shared and read only. If tcp_mmap() finds data that is not a full page, or a patch of urgent data, -EINVAL is returned, no bytes are consumed. Application must fallback to recvmsg() to read the problematic sequence. mmap() wont block, regardless of socket being in blocking or non-blocking mode. If not enough bytes are in receive queue, mmap() would return -EAGAIN, or -EIO if socket is in a state where no other bytes can be added into receive queue. An application might use SO_RCVLOWAT, poll() and/or ioctl( FIONREAD) to efficiently use mmap() On the sender side, MSG_EOR might help to clearly separate unaligned headers and 4K-aligned chunks if necessary. Tested: mlx4 (cx-3) 40Gbit NIC, with tcp_mmap program provided in following patch. MTU set to 4168 (4096 TCP payload, 40 bytes IPv6 header, 32 bytes TCP header) Without mmap() (tcp_mmap -s) received 32768 MB (0 % mmap'ed) in 8.13342 s, 33.7961 Gbit, cpu usage user:0.034 sys:3.778, 116.333 usec per MB, 63062 c-switches received 32768 MB (0 % mmap'ed) in 8.14501 s, 33.748 Gbit, cpu usage user:0.029 sys:3.997, 122.864 usec per MB, 61903 c-switches received 32768 MB (0 % mmap'ed) in 8.11723 s, 33.8635 Gbit, cpu usage user:0.048 sys:3.964, 122.437 usec per MB, 62983 c-switches received 32768 MB (0 % mmap'ed) in 8.39189 s, 32.7552 Gbit, cpu usage user:0.038 sys:4.181, 128.754 usec per MB, 55834 c-switches With mmap() on receiver (tcp_mmap -s -z) received 32768 MB (100 % mmap'ed) in 8.03083 s, 34.2278 Gbit, cpu usage user:0.024 sys:1.466, 45.4712 usec per MB, 65479 c-switches received 32768 MB (100 % mmap'ed) in 7.98805 s, 34.4111 Gbit, cpu usage user:0.026 sys:1.401, 43.5486 usec per MB, 65447 c-switches received 32768 MB (100 % mmap'ed) in 7.98377 s, 34.4296 Gbit, cpu usage user:0.028 sys:1.452, 45.166 usec per MB, 65496 c-switches received 32768 MB (99.9969 % mmap'ed) in 8.01838 s, 34.281 Gbit, cpu usage user:0.02 sys:1.446, 44.7388 usec per MB, 65505 c-switches Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-17 01:33:38 +08:00
sock_rps_record_flow(sk);
if (inq && inq <= copybuf_len)
return receive_fallback_to_copy(sk, zc, inq, tss);
if (inq < PAGE_SIZE) {
zc->length = 0;
zc->recv_skip_hint = inq;
if (!inq && sock_flag(sk, SOCK_DONE))
return -EIO;
return 0;
}
tcp: Use per-vma locking for receive zerocopy Per-VMA locking allows us to lock a struct vm_area_struct without taking the process-wide mmap lock in read mode. Consider a process workload where the mmap lock is taken constantly in write mode. In this scenario, all zerocopy receives are periodically blocked during that period of time - though in principle, the memory ranges being used by TCP are not touched by the operations that need the mmap write lock. This results in performance degradation. Now consider another workload where the mmap lock is never taken in write mode, but there are many TCP connections using receive zerocopy that are concurrently receiving. These connections all take the mmap lock in read mode, but this does induce a lot of contention and atomic ops for this process-wide lock. This results in additional CPU overhead caused by contending on the cache line for this lock. However, with per-vma locking, both of these problems can be avoided. As a test, I ran an RPC-style request/response workload with 4KB payloads and receive zerocopy enabled, with 100 simultaneous TCP connections. I measured perf cycles within the find_tcp_vma/mmap_read_lock/mmap_read_unlock codepath, with and without per-vma locking enabled. When using process-wide mmap semaphore read locking, about 1% of measured perf cycles were within this path. With per-VMA locking, this value dropped to about 0.45%. Signed-off-by: Arjun Roy <arjunroy@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-06-17 03:34:27 +08:00
vma = find_tcp_vma(current->mm, address, &mmap_locked);
if (!vma)
tcp: fix error recovery in tcp_zerocopy_receive() If user provides wrong virtual address in TCP_ZEROCOPY_RECEIVE operation we want to return -EINVAL error. But depending on zc->recv_skip_hint content, we might return -EIO error if the socket has SOCK_DONE set. Make sure to return -EINVAL in this case. BUG: KMSAN: uninit-value in tcp_zerocopy_receive net/ipv4/tcp.c:1833 [inline] BUG: KMSAN: uninit-value in do_tcp_getsockopt+0x4494/0x6320 net/ipv4/tcp.c:3685 CPU: 1 PID: 625 Comm: syz-executor.0 Not tainted 5.7.0-rc4-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x1c9/0x220 lib/dump_stack.c:118 kmsan_report+0xf7/0x1e0 mm/kmsan/kmsan_report.c:121 __msan_warning+0x58/0xa0 mm/kmsan/kmsan_instr.c:215 tcp_zerocopy_receive net/ipv4/tcp.c:1833 [inline] do_tcp_getsockopt+0x4494/0x6320 net/ipv4/tcp.c:3685 tcp_getsockopt+0xf8/0x1f0 net/ipv4/tcp.c:3728 sock_common_getsockopt+0x13f/0x180 net/core/sock.c:3131 __sys_getsockopt+0x533/0x7b0 net/socket.c:2177 __do_sys_getsockopt net/socket.c:2192 [inline] __se_sys_getsockopt+0xe1/0x100 net/socket.c:2189 __x64_sys_getsockopt+0x62/0x80 net/socket.c:2189 do_syscall_64+0xb8/0x160 arch/x86/entry/common.c:297 entry_SYSCALL_64_after_hwframe+0x44/0xa9 RIP: 0033:0x45c829 Code: 0d b7 fb ff c3 66 2e 0f 1f 84 00 00 00 00 00 66 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 db b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00 RSP: 002b:00007f1deeb72c78 EFLAGS: 00000246 ORIG_RAX: 0000000000000037 RAX: ffffffffffffffda RBX: 00000000004e01e0 RCX: 000000000045c829 RDX: 0000000000000023 RSI: 0000000000000006 RDI: 0000000000000009 RBP: 000000000078bf00 R08: 0000000020000200 R09: 0000000000000000 R10: 00000000200001c0 R11: 0000000000000246 R12: 00000000ffffffff R13: 00000000000001d8 R14: 00000000004d3038 R15: 00007f1deeb736d4 Local variable ----zc@do_tcp_getsockopt created at: do_tcp_getsockopt+0x1a74/0x6320 net/ipv4/tcp.c:3670 do_tcp_getsockopt+0x1a74/0x6320 net/ipv4/tcp.c:3670 Fixes: 05255b823a61 ("tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-05-15 04:58:13 +08:00
return -EINVAL;
tcp: Use per-vma locking for receive zerocopy Per-VMA locking allows us to lock a struct vm_area_struct without taking the process-wide mmap lock in read mode. Consider a process workload where the mmap lock is taken constantly in write mode. In this scenario, all zerocopy receives are periodically blocked during that period of time - though in principle, the memory ranges being used by TCP are not touched by the operations that need the mmap write lock. This results in performance degradation. Now consider another workload where the mmap lock is never taken in write mode, but there are many TCP connections using receive zerocopy that are concurrently receiving. These connections all take the mmap lock in read mode, but this does induce a lot of contention and atomic ops for this process-wide lock. This results in additional CPU overhead caused by contending on the cache line for this lock. However, with per-vma locking, both of these problems can be avoided. As a test, I ran an RPC-style request/response workload with 4KB payloads and receive zerocopy enabled, with 100 simultaneous TCP connections. I measured perf cycles within the find_tcp_vma/mmap_read_lock/mmap_read_unlock codepath, with and without per-vma locking enabled. When using process-wide mmap semaphore read locking, about 1% of measured perf cycles were within this path. With per-VMA locking, this value dropped to about 0.45%. Signed-off-by: Arjun Roy <arjunroy@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-06-17 03:34:27 +08:00
vma_len = min_t(unsigned long, zc->length, vma->vm_end - address);
avail_len = min_t(u32, vma_len, inq);
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
total_bytes_to_map = avail_len & ~(PAGE_SIZE - 1);
if (total_bytes_to_map) {
if (!(zc->flags & TCP_RECEIVE_ZEROCOPY_FLAG_TLB_CLEAN_HINT))
mm: remove zap_page_range and create zap_vma_pages zap_page_range was originally designed to unmap pages within an address range that could span multiple vmas. While working on [1], it was discovered that all callers of zap_page_range pass a range entirely within a single vma. In addition, the mmu notification call within zap_page range does not correctly handle ranges that span multiple vmas. When crossing a vma boundary, a new mmu_notifier_range_init/end call pair with the new vma should be made. Instead of fixing zap_page_range, do the following: - Create a new routine zap_vma_pages() that will remove all pages within the passed vma. Most users of zap_page_range pass the entire vma and can use this new routine. - For callers of zap_page_range not passing the entire vma, instead call zap_page_range_single(). - Remove zap_page_range. [1] https://lore.kernel.org/linux-mm/20221114235507.294320-2-mike.kravetz@oracle.com/ Link: https://lkml.kernel.org/r/20230104002732.232573-1-mike.kravetz@oracle.com Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com> Suggested-by: Peter Xu <peterx@redhat.com> Acked-by: Michal Hocko <mhocko@suse.com> Acked-by: Peter Xu <peterx@redhat.com> Acked-by: Heiko Carstens <hca@linux.ibm.com> [s390] Reviewed-by: Christoph Hellwig <hch@lst.de> Cc: Christian Borntraeger <borntraeger@linux.ibm.com> Cc: Christian Brauner <brauner@kernel.org> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: David Hildenbrand <david@redhat.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Nadav Amit <nadav.amit@gmail.com> Cc: Palmer Dabbelt <palmer@dabbelt.com> Cc: Rik van Riel <riel@surriel.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Will Deacon <will@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2023-01-04 08:27:32 +08:00
zap_page_range_single(vma, address, total_bytes_to_map,
NULL);
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
zc->length = total_bytes_to_map;
zc->recv_skip_hint = 0;
} else {
zc->length = avail_len;
zc->recv_skip_hint = avail_len;
}
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
ret = 0;
while (length + PAGE_SIZE <= zc->length) {
int mappable_offset;
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
struct page *page;
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
if (zc->recv_skip_hint < PAGE_SIZE) {
u32 offset_frag;
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
if (skb) {
if (zc->recv_skip_hint > 0)
break;
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
skb = skb->next;
offset = seq - TCP_SKB_CB(skb)->seq;
} else {
skb = tcp_recv_skb(sk, seq, &offset);
}
if (TCP_SKB_CB(skb)->has_rxtstamp) {
tcp_update_recv_tstamps(skb, tss);
zc->msg_flags |= TCP_CMSG_TS;
}
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
zc->recv_skip_hint = skb->len - offset;
frags = skb_advance_to_frag(skb, offset, &offset_frag);
if (!frags || offset_frag)
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
break;
tcp: implement mmap() for zero copy receive Some networks can make sure TCP payload can exactly fit 4KB pages, with well chosen MSS/MTU and architectures. Implement mmap() system call so that applications can avoid copying data without complex splice() games. Note that a successful mmap( X bytes) on TCP socket is consuming bytes, as if recvmsg() has been done. (tp->copied += X) Only PROT_READ mappings are accepted, as skb page frags are fundamentally shared and read only. If tcp_mmap() finds data that is not a full page, or a patch of urgent data, -EINVAL is returned, no bytes are consumed. Application must fallback to recvmsg() to read the problematic sequence. mmap() wont block, regardless of socket being in blocking or non-blocking mode. If not enough bytes are in receive queue, mmap() would return -EAGAIN, or -EIO if socket is in a state where no other bytes can be added into receive queue. An application might use SO_RCVLOWAT, poll() and/or ioctl( FIONREAD) to efficiently use mmap() On the sender side, MSG_EOR might help to clearly separate unaligned headers and 4K-aligned chunks if necessary. Tested: mlx4 (cx-3) 40Gbit NIC, with tcp_mmap program provided in following patch. MTU set to 4168 (4096 TCP payload, 40 bytes IPv6 header, 32 bytes TCP header) Without mmap() (tcp_mmap -s) received 32768 MB (0 % mmap'ed) in 8.13342 s, 33.7961 Gbit, cpu usage user:0.034 sys:3.778, 116.333 usec per MB, 63062 c-switches received 32768 MB (0 % mmap'ed) in 8.14501 s, 33.748 Gbit, cpu usage user:0.029 sys:3.997, 122.864 usec per MB, 61903 c-switches received 32768 MB (0 % mmap'ed) in 8.11723 s, 33.8635 Gbit, cpu usage user:0.048 sys:3.964, 122.437 usec per MB, 62983 c-switches received 32768 MB (0 % mmap'ed) in 8.39189 s, 32.7552 Gbit, cpu usage user:0.038 sys:4.181, 128.754 usec per MB, 55834 c-switches With mmap() on receiver (tcp_mmap -s -z) received 32768 MB (100 % mmap'ed) in 8.03083 s, 34.2278 Gbit, cpu usage user:0.024 sys:1.466, 45.4712 usec per MB, 65479 c-switches received 32768 MB (100 % mmap'ed) in 7.98805 s, 34.4111 Gbit, cpu usage user:0.026 sys:1.401, 43.5486 usec per MB, 65447 c-switches received 32768 MB (100 % mmap'ed) in 7.98377 s, 34.4296 Gbit, cpu usage user:0.028 sys:1.452, 45.166 usec per MB, 65496 c-switches received 32768 MB (99.9969 % mmap'ed) in 8.01838 s, 34.281 Gbit, cpu usage user:0.02 sys:1.446, 44.7388 usec per MB, 65505 c-switches Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-17 01:33:38 +08:00
}
mappable_offset = find_next_mappable_frag(frags,
zc->recv_skip_hint);
if (mappable_offset) {
zc->recv_skip_hint = mappable_offset;
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
break;
}
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
page = skb_frag_page(frags);
prefetchw(page);
pages[pages_to_map++] = page;
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
length += PAGE_SIZE;
zc->recv_skip_hint -= PAGE_SIZE;
frags++;
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
if (pages_to_map == TCP_ZEROCOPY_PAGE_BATCH_SIZE ||
zc->recv_skip_hint < PAGE_SIZE) {
/* Either full batch, or we're about to go to next skb
* (and we cannot unroll failed ops across skbs).
*/
ret = tcp_zerocopy_vm_insert_batch(vma, pages,
pages_to_map,
&address, &length,
&seq, zc,
total_bytes_to_map);
if (ret)
goto out;
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
pages_to_map = 0;
}
}
net-zerocopy: Defer vm zap unless actually needed. Zapping pages is required only if we are calling vm_insert_page into a region where pages had previously been mapped. Receive zerocopy allows reusing such regions, and hitherto called zap_page_range() before calling vm_insert_page() in that range. zap_page_range() can also be triggered from userspace with madvise(MADV_DONTNEED). If userspace is configured to call this before reusing a segment, or if there was nothing mapped at this virtual address to begin with, we can avoid calling zap_page_range() under the socket lock. That said, if userspace does not do that, then we are still responsible for calling zap_page_range(). This patch adds a flag that the user can use to hint to the kernel that a zap is not required. If the flag is not set, or if an older user application does not have a flags field at all, then the kernel calls zap_page_range as before. Also, if the flag is set but a zap is still required, the kernel performs that zap as necessary. Thus incorrectly indicating that a zap can be avoided does not change the correctness of operation. It also increases the batchsize for vm_insert_pages and prefetches the page struct for the batch since we're about to bump the refcount. An alternative mechanism could be to not have a flag, assume by default a zap is not needed, and fall back to zapping if needed. However, this would harm performance for older applications for which a zap is necessary, and thus we implement it with an explicit flag so newer applications can opt in. When using RPC-style traffic with medium sized (tens of KB) RPCs, this change yields an efficency improvement of about 30% for QPS/CPU usage. Signed-off-by: Arjun Roy <arjunroy@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-03 06:53:49 +08:00
if (pages_to_map) {
ret = tcp_zerocopy_vm_insert_batch(vma, pages, pages_to_map,
&address, &length, &seq,
zc, total_bytes_to_map);
tcp: implement mmap() for zero copy receive Some networks can make sure TCP payload can exactly fit 4KB pages, with well chosen MSS/MTU and architectures. Implement mmap() system call so that applications can avoid copying data without complex splice() games. Note that a successful mmap( X bytes) on TCP socket is consuming bytes, as if recvmsg() has been done. (tp->copied += X) Only PROT_READ mappings are accepted, as skb page frags are fundamentally shared and read only. If tcp_mmap() finds data that is not a full page, or a patch of urgent data, -EINVAL is returned, no bytes are consumed. Application must fallback to recvmsg() to read the problematic sequence. mmap() wont block, regardless of socket being in blocking or non-blocking mode. If not enough bytes are in receive queue, mmap() would return -EAGAIN, or -EIO if socket is in a state where no other bytes can be added into receive queue. An application might use SO_RCVLOWAT, poll() and/or ioctl( FIONREAD) to efficiently use mmap() On the sender side, MSG_EOR might help to clearly separate unaligned headers and 4K-aligned chunks if necessary. Tested: mlx4 (cx-3) 40Gbit NIC, with tcp_mmap program provided in following patch. MTU set to 4168 (4096 TCP payload, 40 bytes IPv6 header, 32 bytes TCP header) Without mmap() (tcp_mmap -s) received 32768 MB (0 % mmap'ed) in 8.13342 s, 33.7961 Gbit, cpu usage user:0.034 sys:3.778, 116.333 usec per MB, 63062 c-switches received 32768 MB (0 % mmap'ed) in 8.14501 s, 33.748 Gbit, cpu usage user:0.029 sys:3.997, 122.864 usec per MB, 61903 c-switches received 32768 MB (0 % mmap'ed) in 8.11723 s, 33.8635 Gbit, cpu usage user:0.048 sys:3.964, 122.437 usec per MB, 62983 c-switches received 32768 MB (0 % mmap'ed) in 8.39189 s, 32.7552 Gbit, cpu usage user:0.038 sys:4.181, 128.754 usec per MB, 55834 c-switches With mmap() on receiver (tcp_mmap -s -z) received 32768 MB (100 % mmap'ed) in 8.03083 s, 34.2278 Gbit, cpu usage user:0.024 sys:1.466, 45.4712 usec per MB, 65479 c-switches received 32768 MB (100 % mmap'ed) in 7.98805 s, 34.4111 Gbit, cpu usage user:0.026 sys:1.401, 43.5486 usec per MB, 65447 c-switches received 32768 MB (100 % mmap'ed) in 7.98377 s, 34.4296 Gbit, cpu usage user:0.028 sys:1.452, 45.166 usec per MB, 65496 c-switches received 32768 MB (99.9969 % mmap'ed) in 8.01838 s, 34.281 Gbit, cpu usage user:0.02 sys:1.446, 44.7388 usec per MB, 65505 c-switches Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-17 01:33:38 +08:00
}
out:
tcp: Use per-vma locking for receive zerocopy Per-VMA locking allows us to lock a struct vm_area_struct without taking the process-wide mmap lock in read mode. Consider a process workload where the mmap lock is taken constantly in write mode. In this scenario, all zerocopy receives are periodically blocked during that period of time - though in principle, the memory ranges being used by TCP are not touched by the operations that need the mmap write lock. This results in performance degradation. Now consider another workload where the mmap lock is never taken in write mode, but there are many TCP connections using receive zerocopy that are concurrently receiving. These connections all take the mmap lock in read mode, but this does induce a lot of contention and atomic ops for this process-wide lock. This results in additional CPU overhead caused by contending on the cache line for this lock. However, with per-vma locking, both of these problems can be avoided. As a test, I ran an RPC-style request/response workload with 4KB payloads and receive zerocopy enabled, with 100 simultaneous TCP connections. I measured perf cycles within the find_tcp_vma/mmap_read_lock/mmap_read_unlock codepath, with and without per-vma locking enabled. When using process-wide mmap semaphore read locking, about 1% of measured perf cycles were within this path. With per-VMA locking, this value dropped to about 0.45%. Signed-off-by: Arjun Roy <arjunroy@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-06-17 03:34:27 +08:00
if (mmap_locked)
mmap_read_unlock(current->mm);
else
vma_end_read(vma);
/* Try to copy straggler data. */
if (!ret)
copylen = tcp_zc_handle_leftover(zc, sk, skb, &seq, copybuf_len, tss);
if (length + copylen) {
WRITE_ONCE(tp->copied_seq, seq);
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
tcp_rcv_space_adjust(sk);
/* Clean up data we have read: This will do ACK frames. */
tcp_recv_skb(sk, seq, &offset);
tcp_cleanup_rbuf(sk, length + copylen);
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
ret = 0;
if (length == zc->length)
zc->recv_skip_hint = 0;
} else {
if (!zc->recv_skip_hint && sock_flag(sk, SOCK_DONE))
ret = -EIO;
}
zc->length = length;
tcp: implement mmap() for zero copy receive Some networks can make sure TCP payload can exactly fit 4KB pages, with well chosen MSS/MTU and architectures. Implement mmap() system call so that applications can avoid copying data without complex splice() games. Note that a successful mmap( X bytes) on TCP socket is consuming bytes, as if recvmsg() has been done. (tp->copied += X) Only PROT_READ mappings are accepted, as skb page frags are fundamentally shared and read only. If tcp_mmap() finds data that is not a full page, or a patch of urgent data, -EINVAL is returned, no bytes are consumed. Application must fallback to recvmsg() to read the problematic sequence. mmap() wont block, regardless of socket being in blocking or non-blocking mode. If not enough bytes are in receive queue, mmap() would return -EAGAIN, or -EIO if socket is in a state where no other bytes can be added into receive queue. An application might use SO_RCVLOWAT, poll() and/or ioctl( FIONREAD) to efficiently use mmap() On the sender side, MSG_EOR might help to clearly separate unaligned headers and 4K-aligned chunks if necessary. Tested: mlx4 (cx-3) 40Gbit NIC, with tcp_mmap program provided in following patch. MTU set to 4168 (4096 TCP payload, 40 bytes IPv6 header, 32 bytes TCP header) Without mmap() (tcp_mmap -s) received 32768 MB (0 % mmap'ed) in 8.13342 s, 33.7961 Gbit, cpu usage user:0.034 sys:3.778, 116.333 usec per MB, 63062 c-switches received 32768 MB (0 % mmap'ed) in 8.14501 s, 33.748 Gbit, cpu usage user:0.029 sys:3.997, 122.864 usec per MB, 61903 c-switches received 32768 MB (0 % mmap'ed) in 8.11723 s, 33.8635 Gbit, cpu usage user:0.048 sys:3.964, 122.437 usec per MB, 62983 c-switches received 32768 MB (0 % mmap'ed) in 8.39189 s, 32.7552 Gbit, cpu usage user:0.038 sys:4.181, 128.754 usec per MB, 55834 c-switches With mmap() on receiver (tcp_mmap -s -z) received 32768 MB (100 % mmap'ed) in 8.03083 s, 34.2278 Gbit, cpu usage user:0.024 sys:1.466, 45.4712 usec per MB, 65479 c-switches received 32768 MB (100 % mmap'ed) in 7.98805 s, 34.4111 Gbit, cpu usage user:0.026 sys:1.401, 43.5486 usec per MB, 65447 c-switches received 32768 MB (100 % mmap'ed) in 7.98377 s, 34.4296 Gbit, cpu usage user:0.028 sys:1.452, 45.166 usec per MB, 65496 c-switches received 32768 MB (99.9969 % mmap'ed) in 8.01838 s, 34.281 Gbit, cpu usage user:0.02 sys:1.446, 44.7388 usec per MB, 65505 c-switches Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-17 01:33:38 +08:00
return ret;
}
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
#endif
tcp: implement mmap() for zero copy receive Some networks can make sure TCP payload can exactly fit 4KB pages, with well chosen MSS/MTU and architectures. Implement mmap() system call so that applications can avoid copying data without complex splice() games. Note that a successful mmap( X bytes) on TCP socket is consuming bytes, as if recvmsg() has been done. (tp->copied += X) Only PROT_READ mappings are accepted, as skb page frags are fundamentally shared and read only. If tcp_mmap() finds data that is not a full page, or a patch of urgent data, -EINVAL is returned, no bytes are consumed. Application must fallback to recvmsg() to read the problematic sequence. mmap() wont block, regardless of socket being in blocking or non-blocking mode. If not enough bytes are in receive queue, mmap() would return -EAGAIN, or -EIO if socket is in a state where no other bytes can be added into receive queue. An application might use SO_RCVLOWAT, poll() and/or ioctl( FIONREAD) to efficiently use mmap() On the sender side, MSG_EOR might help to clearly separate unaligned headers and 4K-aligned chunks if necessary. Tested: mlx4 (cx-3) 40Gbit NIC, with tcp_mmap program provided in following patch. MTU set to 4168 (4096 TCP payload, 40 bytes IPv6 header, 32 bytes TCP header) Without mmap() (tcp_mmap -s) received 32768 MB (0 % mmap'ed) in 8.13342 s, 33.7961 Gbit, cpu usage user:0.034 sys:3.778, 116.333 usec per MB, 63062 c-switches received 32768 MB (0 % mmap'ed) in 8.14501 s, 33.748 Gbit, cpu usage user:0.029 sys:3.997, 122.864 usec per MB, 61903 c-switches received 32768 MB (0 % mmap'ed) in 8.11723 s, 33.8635 Gbit, cpu usage user:0.048 sys:3.964, 122.437 usec per MB, 62983 c-switches received 32768 MB (0 % mmap'ed) in 8.39189 s, 32.7552 Gbit, cpu usage user:0.038 sys:4.181, 128.754 usec per MB, 55834 c-switches With mmap() on receiver (tcp_mmap -s -z) received 32768 MB (100 % mmap'ed) in 8.03083 s, 34.2278 Gbit, cpu usage user:0.024 sys:1.466, 45.4712 usec per MB, 65479 c-switches received 32768 MB (100 % mmap'ed) in 7.98805 s, 34.4111 Gbit, cpu usage user:0.026 sys:1.401, 43.5486 usec per MB, 65447 c-switches received 32768 MB (100 % mmap'ed) in 7.98377 s, 34.4296 Gbit, cpu usage user:0.028 sys:1.452, 45.166 usec per MB, 65496 c-switches received 32768 MB (99.9969 % mmap'ed) in 8.01838 s, 34.281 Gbit, cpu usage user:0.02 sys:1.446, 44.7388 usec per MB, 65505 c-switches Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-17 01:33:38 +08:00
/* Similar to __sock_recv_timestamp, but does not require an skb */
void tcp_recv_timestamp(struct msghdr *msg, const struct sock *sk,
struct scm_timestamping_internal *tss)
{
int new_tstamp = sock_flag(sk, SOCK_TSTAMP_NEW);
bool has_timestamping = false;
if (tss->ts[0].tv_sec || tss->ts[0].tv_nsec) {
if (sock_flag(sk, SOCK_RCVTSTAMP)) {
if (sock_flag(sk, SOCK_RCVTSTAMPNS)) {
if (new_tstamp) {
struct __kernel_timespec kts = {
.tv_sec = tss->ts[0].tv_sec,
.tv_nsec = tss->ts[0].tv_nsec,
};
put_cmsg(msg, SOL_SOCKET, SO_TIMESTAMPNS_NEW,
sizeof(kts), &kts);
} else {
struct __kernel_old_timespec ts_old = {
.tv_sec = tss->ts[0].tv_sec,
.tv_nsec = tss->ts[0].tv_nsec,
};
put_cmsg(msg, SOL_SOCKET, SO_TIMESTAMPNS_OLD,
sizeof(ts_old), &ts_old);
}
} else {
if (new_tstamp) {
struct __kernel_sock_timeval stv = {
.tv_sec = tss->ts[0].tv_sec,
.tv_usec = tss->ts[0].tv_nsec / 1000,
};
put_cmsg(msg, SOL_SOCKET, SO_TIMESTAMP_NEW,
sizeof(stv), &stv);
} else {
struct __kernel_old_timeval tv = {
.tv_sec = tss->ts[0].tv_sec,
.tv_usec = tss->ts[0].tv_nsec / 1000,
};
put_cmsg(msg, SOL_SOCKET, SO_TIMESTAMP_OLD,
sizeof(tv), &tv);
}
}
}
if (READ_ONCE(sk->sk_tsflags) & SOF_TIMESTAMPING_SOFTWARE)
has_timestamping = true;
else
tss->ts[0] = (struct timespec64) {0};
}
if (tss->ts[2].tv_sec || tss->ts[2].tv_nsec) {
if (READ_ONCE(sk->sk_tsflags) & SOF_TIMESTAMPING_RAW_HARDWARE)
has_timestamping = true;
else
tss->ts[2] = (struct timespec64) {0};
}
if (has_timestamping) {
tss->ts[1] = (struct timespec64) {0};
if (sock_flag(sk, SOCK_TSTAMP_NEW))
put_cmsg_scm_timestamping64(msg, tss);
else
put_cmsg_scm_timestamping(msg, tss);
}
}
tcp: send in-queue bytes in cmsg upon read Applications with many concurrent connections, high variance in receive queue length and tight memory bounds cannot allocate worst-case buffer size to drain sockets. Knowing the size of receive queue length, applications can optimize how they allocate buffers to read from the socket. The number of bytes pending on the socket is directly available through ioctl(FIONREAD/SIOCINQ) and can be approximated using getsockopt(MEMINFO) (rmem_alloc includes skb overheads in addition to application data). But, both of these options add an extra syscall per recvmsg. Moreover, ioctl(FIONREAD/SIOCINQ) takes the socket lock. Add the TCP_INQ socket option to TCP. When this socket option is set, recvmsg() relays the number of bytes available on the socket for reading to the application via the TCP_CM_INQ control message. Calculate the number of bytes after releasing the socket lock to include the processed backlog, if any. To avoid an extra branch in the hot path of recvmsg() for this new control message, move all cmsg processing inside an existing branch for processing receive timestamps. Since the socket lock is not held when calculating the size of receive queue, TCP_INQ is a hint. For example, it can overestimate the queue size by one byte, if FIN is received. With this method, applications can start reading from the socket using a small buffer, and then use larger buffers based on the remaining data when needed. V3 change-log: As suggested by David Miller, added loads with barrier to check whether we have multiple threads calling recvmsg in parallel. When that happens we lock the socket to calculate inq. V4 change-log: Removed inline from a static function. Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Willem de Bruijn <willemb@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Neal Cardwell <ncardwell@google.com> Suggested-by: David Miller <davem@davemloft.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-05-02 03:39:15 +08:00
static int tcp_inq_hint(struct sock *sk)
{
const struct tcp_sock *tp = tcp_sk(sk);
u32 copied_seq = READ_ONCE(tp->copied_seq);
u32 rcv_nxt = READ_ONCE(tp->rcv_nxt);
int inq;
inq = rcv_nxt - copied_seq;
if (unlikely(inq < 0 || copied_seq != READ_ONCE(tp->copied_seq))) {
lock_sock(sk);
inq = tp->rcv_nxt - tp->copied_seq;
release_sock(sk);
}
/* After receiving a FIN, tell the user-space to continue reading
* by returning a non-zero inq.
*/
if (inq == 0 && sock_flag(sk, SOCK_DONE))
inq = 1;
tcp: send in-queue bytes in cmsg upon read Applications with many concurrent connections, high variance in receive queue length and tight memory bounds cannot allocate worst-case buffer size to drain sockets. Knowing the size of receive queue length, applications can optimize how they allocate buffers to read from the socket. The number of bytes pending on the socket is directly available through ioctl(FIONREAD/SIOCINQ) and can be approximated using getsockopt(MEMINFO) (rmem_alloc includes skb overheads in addition to application data). But, both of these options add an extra syscall per recvmsg. Moreover, ioctl(FIONREAD/SIOCINQ) takes the socket lock. Add the TCP_INQ socket option to TCP. When this socket option is set, recvmsg() relays the number of bytes available on the socket for reading to the application via the TCP_CM_INQ control message. Calculate the number of bytes after releasing the socket lock to include the processed backlog, if any. To avoid an extra branch in the hot path of recvmsg() for this new control message, move all cmsg processing inside an existing branch for processing receive timestamps. Since the socket lock is not held when calculating the size of receive queue, TCP_INQ is a hint. For example, it can overestimate the queue size by one byte, if FIN is received. With this method, applications can start reading from the socket using a small buffer, and then use larger buffers based on the remaining data when needed. V3 change-log: As suggested by David Miller, added loads with barrier to check whether we have multiple threads calling recvmsg in parallel. When that happens we lock the socket to calculate inq. V4 change-log: Removed inline from a static function. Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Willem de Bruijn <willemb@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Neal Cardwell <ncardwell@google.com> Suggested-by: David Miller <davem@davemloft.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-05-02 03:39:15 +08:00
return inq;
}
/*
* This routine copies from a sock struct into the user buffer.
*
* Technical note: in 2.3 we work on _locked_ socket, so that
* tricks with *seq access order and skb->users are not required.
* Probably, code can be easily improved even more.
*/
static int tcp_recvmsg_locked(struct sock *sk, struct msghdr *msg, size_t len,
int flags, struct scm_timestamping_internal *tss,
int *cmsg_flags)
{
struct tcp_sock *tp = tcp_sk(sk);
int copied = 0;
u32 peek_seq;
u32 *seq;
unsigned long used;
int err;
int target; /* Read at least this many bytes */
long timeo;
struct sk_buff *skb, *last;
u32 urg_hole = 0;
err = -ENOTCONN;
if (sk->sk_state == TCP_LISTEN)
goto out;
if (tp->recvmsg_inq) {
*cmsg_flags = TCP_CMSG_INQ;
msg->msg_get_inq = 1;
}
timeo = sock_rcvtimeo(sk, flags & MSG_DONTWAIT);
/* Urgent data needs to be handled specially. */
if (flags & MSG_OOB)
goto recv_urg;
tcp: Repair socket queues Reading queues under repair mode is done with recvmsg call. The queue-under-repair set by TCP_REPAIR_QUEUE option is used to determine which queue should be read. Thus both send and receive queue can be read with this. Caller must pass the MSG_PEEK flag. Writing to queues is done with sendmsg call and yet again -- the repair-queue option can be used to push data into the receive queue. When putting an skb into receive queue a zero tcp header is appented to its head to address the tcp_hdr(skb)->syn and the ->fin checks by the (after repair) tcp_recvmsg. These flags flags are both set to zero and that's why. The fin cannot be met in the queue while reading the source socket, since the repair only works for closed/established sockets and queueing fin packet always changes its state. The syn in the queue denotes that the respective skb's seq is "off-by-one" as compared to the actual payload lenght. Thus, at the rcv queue refill we can just drop this flag and set the skb's sequences to precice values. When the repair mode is turned off, the write queue seqs are updated so that the whole queue is considered to be 'already sent, waiting for ACKs' (write_seq = snd_nxt <= snd_una). From the protocol POV the send queue looks like it was sent, but the data between the write_seq and snd_nxt is lost in the network. This helps to avoid another sockoption for setting the snd_nxt sequence. Leaving the whole queue in a 'not yet sent' state (as it will be after sendmsg-s) will not allow to receive any acks from the peer since the ack_seq will be after the snd_nxt. Thus even the ack for the window probe will be dropped and the connection will be 'locked' with the zero peer window. Signed-off-by: Pavel Emelyanov <xemul@parallels.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-19 11:41:01 +08:00
if (unlikely(tp->repair)) {
err = -EPERM;
if (!(flags & MSG_PEEK))
goto out;
if (tp->repair_queue == TCP_SEND_QUEUE)
goto recv_sndq;
err = -EINVAL;
if (tp->repair_queue == TCP_NO_QUEUE)
goto out;
/* 'common' recv queue MSG_PEEK-ing */
}
seq = &tp->copied_seq;
if (flags & MSG_PEEK) {
peek_seq = tp->copied_seq;
seq = &peek_seq;
}
target = sock_rcvlowat(sk, flags & MSG_WAITALL, len);
do {
u32 offset;
/* Are we at urgent data? Stop if we have read anything or have SIGURG pending. */
if (unlikely(tp->urg_data) && tp->urg_seq == *seq) {
if (copied)
break;
if (signal_pending(current)) {
copied = timeo ? sock_intr_errno(timeo) : -EAGAIN;
break;
}
}
/* Next get a buffer. */
last = skb_peek_tail(&sk->sk_receive_queue);
skb_queue_walk(&sk->sk_receive_queue, skb) {
last = skb;
/* Now that we have two receive queues this
* shouldn't happen.
*/
if (WARN(before(*seq, TCP_SKB_CB(skb)->seq),
"TCP recvmsg seq # bug: copied %X, seq %X, rcvnxt %X, fl %X\n",
*seq, TCP_SKB_CB(skb)->seq, tp->rcv_nxt,
flags))
break;
offset = *seq - TCP_SKB_CB(skb)->seq;
if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) {
pr_err_once("%s: found a SYN, please report !\n", __func__);
offset--;
}
if (offset < skb->len)
goto found_ok_skb;
if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)
goto found_fin_ok;
WARN(!(flags & MSG_PEEK),
"TCP recvmsg seq # bug 2: copied %X, seq %X, rcvnxt %X, fl %X\n",
*seq, TCP_SKB_CB(skb)->seq, tp->rcv_nxt, flags);
}
/* Well, if we have backlog, try to process it now yet. */
net: silence data-races on sk_backlog.tail sk->sk_backlog.tail might be read without holding the socket spinlock, we need to add proper READ_ONCE()/WRITE_ONCE() to silence the warnings. KCSAN reported : BUG: KCSAN: data-race in tcp_add_backlog / tcp_recvmsg write to 0xffff8881265109f8 of 8 bytes by interrupt on cpu 1: __sk_add_backlog include/net/sock.h:907 [inline] sk_add_backlog include/net/sock.h:938 [inline] tcp_add_backlog+0x476/0xce0 net/ipv4/tcp_ipv4.c:1759 tcp_v4_rcv+0x1a70/0x1bd0 net/ipv4/tcp_ipv4.c:1947 ip_protocol_deliver_rcu+0x4d/0x420 net/ipv4/ip_input.c:204 ip_local_deliver_finish+0x110/0x140 net/ipv4/ip_input.c:231 NF_HOOK include/linux/netfilter.h:305 [inline] NF_HOOK include/linux/netfilter.h:299 [inline] ip_local_deliver+0x133/0x210 net/ipv4/ip_input.c:252 dst_input include/net/dst.h:442 [inline] ip_rcv_finish+0x121/0x160 net/ipv4/ip_input.c:413 NF_HOOK include/linux/netfilter.h:305 [inline] NF_HOOK include/linux/netfilter.h:299 [inline] ip_rcv+0x18f/0x1a0 net/ipv4/ip_input.c:523 __netif_receive_skb_one_core+0xa7/0xe0 net/core/dev.c:4929 __netif_receive_skb+0x37/0xf0 net/core/dev.c:5043 netif_receive_skb_internal+0x59/0x190 net/core/dev.c:5133 napi_skb_finish net/core/dev.c:5596 [inline] napi_gro_receive+0x28f/0x330 net/core/dev.c:5629 receive_buf+0x284/0x30b0 drivers/net/virtio_net.c:1061 virtnet_receive drivers/net/virtio_net.c:1323 [inline] virtnet_poll+0x436/0x7d0 drivers/net/virtio_net.c:1428 napi_poll net/core/dev.c:6311 [inline] net_rx_action+0x3ae/0xa90 net/core/dev.c:6379 __do_softirq+0x115/0x33f kernel/softirq.c:292 invoke_softirq kernel/softirq.c:373 [inline] irq_exit+0xbb/0xe0 kernel/softirq.c:413 exiting_irq arch/x86/include/asm/apic.h:536 [inline] do_IRQ+0xa6/0x180 arch/x86/kernel/irq.c:263 ret_from_intr+0x0/0x19 native_safe_halt+0xe/0x10 arch/x86/kernel/paravirt.c:71 arch_cpu_idle+0x1f/0x30 arch/x86/kernel/process.c:571 default_idle_call+0x1e/0x40 kernel/sched/idle.c:94 cpuidle_idle_call kernel/sched/idle.c:154 [inline] do_idle+0x1af/0x280 kernel/sched/idle.c:263 cpu_startup_entry+0x1b/0x20 kernel/sched/idle.c:355 start_secondary+0x208/0x260 arch/x86/kernel/smpboot.c:264 secondary_startup_64+0xa4/0xb0 arch/x86/kernel/head_64.S:241 read to 0xffff8881265109f8 of 8 bytes by task 8057 on cpu 0: tcp_recvmsg+0x46e/0x1b40 net/ipv4/tcp.c:2050 inet_recvmsg+0xbb/0x250 net/ipv4/af_inet.c:838 sock_recvmsg_nosec net/socket.c:871 [inline] sock_recvmsg net/socket.c:889 [inline] sock_recvmsg+0x92/0xb0 net/socket.c:885 sock_read_iter+0x15f/0x1e0 net/socket.c:967 call_read_iter include/linux/fs.h:1889 [inline] new_sync_read+0x389/0x4f0 fs/read_write.c:414 __vfs_read+0xb1/0xc0 fs/read_write.c:427 vfs_read fs/read_write.c:461 [inline] vfs_read+0x143/0x2c0 fs/read_write.c:446 ksys_read+0xd5/0x1b0 fs/read_write.c:587 __do_sys_read fs/read_write.c:597 [inline] __se_sys_read fs/read_write.c:595 [inline] __x64_sys_read+0x4c/0x60 fs/read_write.c:595 do_syscall_64+0xcc/0x370 arch/x86/entry/common.c:290 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Reported by Kernel Concurrency Sanitizer on: CPU: 0 PID: 8057 Comm: syz-fuzzer Not tainted 5.4.0-rc6+ #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2019-11-07 02:04:11 +08:00
if (copied >= target && !READ_ONCE(sk->sk_backlog.tail))
break;
if (copied) {
if (!timeo ||
sk->sk_err ||
sk->sk_state == TCP_CLOSE ||
(sk->sk_shutdown & RCV_SHUTDOWN) ||
signal_pending(current))
break;
} else {
if (sock_flag(sk, SOCK_DONE))
break;
if (sk->sk_err) {
copied = sock_error(sk);
break;
}
if (sk->sk_shutdown & RCV_SHUTDOWN)
break;
if (sk->sk_state == TCP_CLOSE) {
/* This occurs when user tries to read
* from never connected socket.
*/
copied = -ENOTCONN;
break;
}
if (!timeo) {
copied = -EAGAIN;
break;
}
if (signal_pending(current)) {
copied = sock_intr_errno(timeo);
break;
}
}
if (copied >= target) {
/* Do not sleep, just process backlog. */
__sk_flush_backlog(sk);
} else {
tcp_cleanup_rbuf(sk, copied);
err = sk_wait_data(sk, &timeo, last);
if (err < 0) {
err = copied ? : err;
goto out;
}
}
if ((flags & MSG_PEEK) &&
(peek_seq - copied - urg_hole != tp->copied_seq)) {
net_dbg_ratelimited("TCP(%s:%d): Application bug, race in MSG_PEEK\n",
current->comm,
task_pid_nr(current));
peek_seq = tp->copied_seq;
}
continue;
found_ok_skb:
/* Ok so how much can we use? */
used = skb->len - offset;
if (len < used)
used = len;
/* Do we have urgent data here? */
if (unlikely(tp->urg_data)) {
u32 urg_offset = tp->urg_seq - *seq;
if (urg_offset < used) {
if (!urg_offset) {
if (!sock_flag(sk, SOCK_URGINLINE)) {
WRITE_ONCE(*seq, *seq + 1);
urg_hole++;
offset++;
used--;
if (!used)
goto skip_copy;
}
} else
used = urg_offset;
}
}
if (!(flags & MSG_TRUNC)) {
err = skb_copy_datagram_msg(skb, offset, msg, used);
if (err) {
/* Exception. Bailout! */
if (!copied)
copied = -EFAULT;
break;
}
}
WRITE_ONCE(*seq, *seq + used);
copied += used;
len -= used;
tcp_rcv_space_adjust(sk);
skip_copy:
if (unlikely(tp->urg_data) && after(tp->copied_seq, tp->urg_seq)) {
WRITE_ONCE(tp->urg_data, 0);
tcp_fast_path_check(sk);
}
if (TCP_SKB_CB(skb)->has_rxtstamp) {
tcp_update_recv_tstamps(skb, tss);
*cmsg_flags |= TCP_CMSG_TS;
}
if (used + offset < skb->len)
continue;
if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)
goto found_fin_ok;
if (!(flags & MSG_PEEK))
tcp_eat_recv_skb(sk, skb);
continue;
found_fin_ok:
/* Process the FIN. */
WRITE_ONCE(*seq, *seq + 1);
if (!(flags & MSG_PEEK))
tcp_eat_recv_skb(sk, skb);
break;
} while (len > 0);
/* According to UNIX98, msg_name/msg_namelen are ignored
* on connected socket. I was just happy when found this 8) --ANK
*/
/* Clean up data we have read: This will do ACK frames. */
tcp_cleanup_rbuf(sk, copied);
return copied;
out:
return err;
recv_urg:
err = tcp_recv_urg(sk, msg, len, flags);
goto out;
tcp: Repair socket queues Reading queues under repair mode is done with recvmsg call. The queue-under-repair set by TCP_REPAIR_QUEUE option is used to determine which queue should be read. Thus both send and receive queue can be read with this. Caller must pass the MSG_PEEK flag. Writing to queues is done with sendmsg call and yet again -- the repair-queue option can be used to push data into the receive queue. When putting an skb into receive queue a zero tcp header is appented to its head to address the tcp_hdr(skb)->syn and the ->fin checks by the (after repair) tcp_recvmsg. These flags flags are both set to zero and that's why. The fin cannot be met in the queue while reading the source socket, since the repair only works for closed/established sockets and queueing fin packet always changes its state. The syn in the queue denotes that the respective skb's seq is "off-by-one" as compared to the actual payload lenght. Thus, at the rcv queue refill we can just drop this flag and set the skb's sequences to precice values. When the repair mode is turned off, the write queue seqs are updated so that the whole queue is considered to be 'already sent, waiting for ACKs' (write_seq = snd_nxt <= snd_una). From the protocol POV the send queue looks like it was sent, but the data between the write_seq and snd_nxt is lost in the network. This helps to avoid another sockoption for setting the snd_nxt sequence. Leaving the whole queue in a 'not yet sent' state (as it will be after sendmsg-s) will not allow to receive any acks from the peer since the ack_seq will be after the snd_nxt. Thus even the ack for the window probe will be dropped and the connection will be 'locked' with the zero peer window. Signed-off-by: Pavel Emelyanov <xemul@parallels.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-19 11:41:01 +08:00
recv_sndq:
err = tcp_peek_sndq(sk, msg, len);
goto out;
}
int tcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, int flags,
int *addr_len)
{
int cmsg_flags = 0, ret;
struct scm_timestamping_internal tss;
if (unlikely(flags & MSG_ERRQUEUE))
return inet_recv_error(sk, msg, len, addr_len);
if (sk_can_busy_loop(sk) &&
skb_queue_empty_lockless(&sk->sk_receive_queue) &&
sk->sk_state == TCP_ESTABLISHED)
sk_busy_loop(sk, flags & MSG_DONTWAIT);
lock_sock(sk);
ret = tcp_recvmsg_locked(sk, msg, len, flags, &tss, &cmsg_flags);
release_sock(sk);
if ((cmsg_flags || msg->msg_get_inq) && ret >= 0) {
if (cmsg_flags & TCP_CMSG_TS)
tcp_recv_timestamp(msg, sk, &tss);
if (msg->msg_get_inq) {
msg->msg_inq = tcp_inq_hint(sk);
if (cmsg_flags & TCP_CMSG_INQ)
put_cmsg(msg, SOL_TCP, TCP_CM_INQ,
sizeof(msg->msg_inq), &msg->msg_inq);
}
}
return ret;
}
EXPORT_SYMBOL(tcp_recvmsg);
void tcp_set_state(struct sock *sk, int state)
{
int oldstate = sk->sk_state;
/* We defined a new enum for TCP states that are exported in BPF
* so as not force the internal TCP states to be frozen. The
* following checks will detect if an internal state value ever
* differs from the BPF value. If this ever happens, then we will
* need to remap the internal value to the BPF value before calling
* tcp_call_bpf_2arg.
*/
BUILD_BUG_ON((int)BPF_TCP_ESTABLISHED != (int)TCP_ESTABLISHED);
BUILD_BUG_ON((int)BPF_TCP_SYN_SENT != (int)TCP_SYN_SENT);
BUILD_BUG_ON((int)BPF_TCP_SYN_RECV != (int)TCP_SYN_RECV);
BUILD_BUG_ON((int)BPF_TCP_FIN_WAIT1 != (int)TCP_FIN_WAIT1);
BUILD_BUG_ON((int)BPF_TCP_FIN_WAIT2 != (int)TCP_FIN_WAIT2);
BUILD_BUG_ON((int)BPF_TCP_TIME_WAIT != (int)TCP_TIME_WAIT);
BUILD_BUG_ON((int)BPF_TCP_CLOSE != (int)TCP_CLOSE);
BUILD_BUG_ON((int)BPF_TCP_CLOSE_WAIT != (int)TCP_CLOSE_WAIT);
BUILD_BUG_ON((int)BPF_TCP_LAST_ACK != (int)TCP_LAST_ACK);
BUILD_BUG_ON((int)BPF_TCP_LISTEN != (int)TCP_LISTEN);
BUILD_BUG_ON((int)BPF_TCP_CLOSING != (int)TCP_CLOSING);
BUILD_BUG_ON((int)BPF_TCP_NEW_SYN_RECV != (int)TCP_NEW_SYN_RECV);
BUILD_BUG_ON((int)BPF_TCP_BOUND_INACTIVE != (int)TCP_BOUND_INACTIVE);
BUILD_BUG_ON((int)BPF_TCP_MAX_STATES != (int)TCP_MAX_STATES);
bpf: net: Emit anonymous enum with BPF_TCP_CLOSE value explicitly The selftest failed to compile with clang-built bpf-next. Adding LLVM=1 to your vmlinux and selftest build will use clang. The error message is: progs/test_sk_storage_tracing.c:38:18: error: use of undeclared identifier 'BPF_TCP_CLOSE' if (newstate == BPF_TCP_CLOSE) ^ 1 error generated. make: *** [Makefile:423: /bpf-next/tools/testing/selftests/bpf/test_sk_storage_tracing.o] Error 1 The reason for the failure is that BPF_TCP_CLOSE, a value of an anonymous enum defined in uapi bpf.h, is not defined in vmlinux.h. gcc does not have this problem. Since vmlinux.h is derived from BTF which is derived from vmlinux DWARF, that means gcc-produced vmlinux DWARF has BPF_TCP_CLOSE while llvm-produced vmlinux DWARF does not have. BPF_TCP_CLOSE is referenced in net/ipv4/tcp.c as BUILD_BUG_ON((int)BPF_TCP_CLOSE != (int)TCP_CLOSE); The following test mimics the above BUILD_BUG_ON, preprocessed with clang compiler, and shows gcc DWARF contains BPF_TCP_CLOSE while llvm DWARF does not. $ cat t.c enum { BPF_TCP_ESTABLISHED = 1, BPF_TCP_CLOSE = 7, }; enum { TCP_ESTABLISHED = 1, TCP_CLOSE = 7, }; int test() { do { extern void __compiletime_assert_767(void) ; if ((int)BPF_TCP_CLOSE != (int)TCP_CLOSE) __compiletime_assert_767(); } while (0); return 0; } $ clang t.c -O2 -c -g && llvm-dwarfdump t.o | grep BPF_TCP_CLOSE $ gcc t.c -O2 -c -g && llvm-dwarfdump t.o | grep BPF_TCP_CLOSE DW_AT_name ("BPF_TCP_CLOSE") Further checking clang code find clang actually tried to evaluate condition at compile time. If it is definitely true/false, it will perform optimization and the whole if condition will be removed before generating IR/debuginfo. This patch explicited add an expression after the above mentioned BUILD_BUG_ON in net/ipv4/tcp.c like (void)BPF_TCP_ESTABLISHED to enable generation of debuginfo for the anonymous enum which also includes BPF_TCP_CLOSE. Signed-off-by: Yonghong Song <yhs@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20210317174132.589276-1-yhs@fb.com
2021-03-18 01:41:32 +08:00
/* bpf uapi header bpf.h defines an anonymous enum with values
* BPF_TCP_* used by bpf programs. Currently gcc built vmlinux
* is able to emit this enum in DWARF due to the above BUILD_BUG_ON.
* But clang built vmlinux does not have this enum in DWARF
* since clang removes the above code before generating IR/debuginfo.
* Let us explicitly emit the type debuginfo to ensure the
* above-mentioned anonymous enum in the vmlinux DWARF and hence BTF
* regardless of which compiler is used.
*/
BTF_TYPE_EMIT_ENUM(BPF_TCP_ESTABLISHED);
if (BPF_SOCK_OPS_TEST_FLAG(tcp_sk(sk), BPF_SOCK_OPS_STATE_CB_FLAG))
tcp_call_bpf_2arg(sk, BPF_SOCK_OPS_STATE_CB, oldstate, state);
switch (state) {
case TCP_ESTABLISHED:
if (oldstate != TCP_ESTABLISHED)
TCP_INC_STATS(sock_net(sk), TCP_MIB_CURRESTAB);
break;
case TCP_CLOSE:
if (oldstate == TCP_CLOSE_WAIT || oldstate == TCP_ESTABLISHED)
TCP_INC_STATS(sock_net(sk), TCP_MIB_ESTABRESETS);
sk->sk_prot->unhash(sk);
if (inet_csk(sk)->icsk_bind_hash &&
!(sk->sk_userlocks & SOCK_BINDPORT_LOCK))
[SOCK] proto: Add hashinfo member to struct proto This way we can remove TCP and DCCP specific versions of sk->sk_prot->get_port: both v4 and v6 use inet_csk_get_port sk->sk_prot->hash: inet_hash is directly used, only v6 need a specific version to deal with mapped sockets sk->sk_prot->unhash: both v4 and v6 use inet_hash directly struct inet_connection_sock_af_ops also gets a new member, bind_conflict, so that inet_csk_get_port can find the per family routine. Now only the lookup routines receive as a parameter a struct inet_hashtable. With this we further reuse code, reducing the difference among INET transport protocols. Eventually work has to be done on UDP and SCTP to make them share this infrastructure and get as a bonus inet_diag interfaces so that iproute can be used with these protocols. net-2.6/net/ipv4/inet_hashtables.c: struct proto | +8 struct inet_connection_sock_af_ops | +8 2 structs changed __inet_hash_nolisten | +18 __inet_hash | -210 inet_put_port | +8 inet_bind_bucket_create | +1 __inet_hash_connect | -8 5 functions changed, 27 bytes added, 218 bytes removed, diff: -191 net-2.6/net/core/sock.c: proto_seq_show | +3 1 function changed, 3 bytes added, diff: +3 net-2.6/net/ipv4/inet_connection_sock.c: inet_csk_get_port | +15 1 function changed, 15 bytes added, diff: +15 net-2.6/net/ipv4/tcp.c: tcp_set_state | -7 1 function changed, 7 bytes removed, diff: -7 net-2.6/net/ipv4/tcp_ipv4.c: tcp_v4_get_port | -31 tcp_v4_hash | -48 tcp_v4_destroy_sock | -7 tcp_v4_syn_recv_sock | -2 tcp_unhash | -179 5 functions changed, 267 bytes removed, diff: -267 net-2.6/net/ipv6/inet6_hashtables.c: __inet6_hash | +8 1 function changed, 8 bytes added, diff: +8 net-2.6/net/ipv4/inet_hashtables.c: inet_unhash | +190 inet_hash | +242 2 functions changed, 432 bytes added, diff: +432 vmlinux: 16 functions changed, 485 bytes added, 492 bytes removed, diff: -7 /home/acme/git/net-2.6/net/ipv6/tcp_ipv6.c: tcp_v6_get_port | -31 tcp_v6_hash | -7 tcp_v6_syn_recv_sock | -9 3 functions changed, 47 bytes removed, diff: -47 /home/acme/git/net-2.6/net/dccp/proto.c: dccp_destroy_sock | -7 dccp_unhash | -179 dccp_hash | -49 dccp_set_state | -7 dccp_done | +1 5 functions changed, 1 bytes added, 242 bytes removed, diff: -241 /home/acme/git/net-2.6/net/dccp/ipv4.c: dccp_v4_get_port | -31 dccp_v4_request_recv_sock | -2 2 functions changed, 33 bytes removed, diff: -33 /home/acme/git/net-2.6/net/dccp/ipv6.c: dccp_v6_get_port | -31 dccp_v6_hash | -7 dccp_v6_request_recv_sock | +5 3 functions changed, 5 bytes added, 38 bytes removed, diff: -33 Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-03 20:06:04 +08:00
inet_put_port(sk);
fallthrough;
default:
if (oldstate == TCP_ESTABLISHED)
TCP_DEC_STATS(sock_net(sk), TCP_MIB_CURRESTAB);
}
/* Change state AFTER socket is unhashed to avoid closed
* socket sitting in hash tables.
*/
inet_sk_state_store(sk, state);
}
EXPORT_SYMBOL_GPL(tcp_set_state);
/*
* State processing on a close. This implements the state shift for
* sending our FIN frame. Note that we only send a FIN for some
* states. A shutdown() may have already sent the FIN, or we may be
* closed.
*/
static const unsigned char new_state[16] = {
/* current state: new state: action: */
[0 /* (Invalid) */] = TCP_CLOSE,
[TCP_ESTABLISHED] = TCP_FIN_WAIT1 | TCP_ACTION_FIN,
[TCP_SYN_SENT] = TCP_CLOSE,
[TCP_SYN_RECV] = TCP_FIN_WAIT1 | TCP_ACTION_FIN,
[TCP_FIN_WAIT1] = TCP_FIN_WAIT1,
[TCP_FIN_WAIT2] = TCP_FIN_WAIT2,
[TCP_TIME_WAIT] = TCP_CLOSE,
[TCP_CLOSE] = TCP_CLOSE,
[TCP_CLOSE_WAIT] = TCP_LAST_ACK | TCP_ACTION_FIN,
[TCP_LAST_ACK] = TCP_LAST_ACK,
[TCP_LISTEN] = TCP_CLOSE,
[TCP_CLOSING] = TCP_CLOSING,
[TCP_NEW_SYN_RECV] = TCP_CLOSE, /* should not happen ! */
};
static int tcp_close_state(struct sock *sk)
{
int next = (int)new_state[sk->sk_state];
int ns = next & TCP_STATE_MASK;
tcp_set_state(sk, ns);
return next & TCP_ACTION_FIN;
}
/*
* Shutdown the sending side of a connection. Much like close except
* that we don't receive shut down or sock_set_flag(sk, SOCK_DEAD).
*/
void tcp_shutdown(struct sock *sk, int how)
{
/* We need to grab some memory, and put together a FIN,
* and then put it into the queue to be sent.
* Tim MacKenzie(tym@dibbler.cs.monash.edu.au) 4 Dec '92.
*/
if (!(how & SEND_SHUTDOWN))
return;
/* If we've already sent a FIN, or it's a closed state, skip this. */
if ((1 << sk->sk_state) &
(TCPF_ESTABLISHED | TCPF_SYN_SENT |
TCPF_SYN_RECV | TCPF_CLOSE_WAIT)) {
/* Clear out any half completed packets. FIN if needed. */
if (tcp_close_state(sk))
tcp_send_fin(sk);
}
}
EXPORT_SYMBOL(tcp_shutdown);
tcp: switch orphan_count to bare per-cpu counters Use of percpu_counter structure to track count of orphaned sockets is causing problems on modern hosts with 256 cpus or more. Stefan Bach reported a serious spinlock contention in real workloads, that I was able to reproduce with a netfilter rule dropping incoming FIN packets. 53.56% server [kernel.kallsyms] [k] queued_spin_lock_slowpath | ---queued_spin_lock_slowpath | --53.51%--_raw_spin_lock_irqsave | --53.51%--__percpu_counter_sum tcp_check_oom | |--39.03%--__tcp_close | tcp_close | inet_release | inet6_release | sock_close | __fput | ____fput | task_work_run | exit_to_usermode_loop | do_syscall_64 | entry_SYSCALL_64_after_hwframe | __GI___libc_close | --14.48%--tcp_out_of_resources tcp_write_timeout tcp_retransmit_timer tcp_write_timer_handler tcp_write_timer call_timer_fn expire_timers __run_timers run_timer_softirq __softirqentry_text_start As explained in commit cf86a086a180 ("net/dst: use a smaller percpu_counter batch for dst entries accounting"), default batch size is too big for the default value of tcp_max_orphans (262144). But even if we reduce batch sizes, there would still be cases where the estimated count of orphans is beyond the limit, and where tcp_too_many_orphans() has to call the expensive percpu_counter_sum_positive(). One solution is to use plain per-cpu counters, and have a timer to periodically refresh this cache. Updating this cache every 100ms seems about right, tcp pressure state is not radically changing over shorter periods. percpu_counter was nice 15 years ago while hosts had less than 16 cpus, not anymore by current standards. v2: Fix the build issue for CONFIG_CRYPTO_DEV_CHELSIO_TLS=m, reported by kernel test robot <lkp@intel.com> Remove unused socket argument from tcp_too_many_orphans() Fixes: dd24c00191d5 ("net: Use a percpu_counter for orphan_count") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Stefan Bach <sfb@google.com> Cc: Neal Cardwell <ncardwell@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-10-14 21:41:26 +08:00
int tcp_orphan_count_sum(void)
{
int i, total = 0;
for_each_possible_cpu(i)
total += per_cpu(tcp_orphan_count, i);
return max(total, 0);
}
static int tcp_orphan_cache;
static struct timer_list tcp_orphan_timer;
#define TCP_ORPHAN_TIMER_PERIOD msecs_to_jiffies(100)
static void tcp_orphan_update(struct timer_list *unused)
{
WRITE_ONCE(tcp_orphan_cache, tcp_orphan_count_sum());
mod_timer(&tcp_orphan_timer, jiffies + TCP_ORPHAN_TIMER_PERIOD);
}
static bool tcp_too_many_orphans(int shift)
{
return READ_ONCE(tcp_orphan_cache) << shift >
READ_ONCE(sysctl_tcp_max_orphans);
tcp: switch orphan_count to bare per-cpu counters Use of percpu_counter structure to track count of orphaned sockets is causing problems on modern hosts with 256 cpus or more. Stefan Bach reported a serious spinlock contention in real workloads, that I was able to reproduce with a netfilter rule dropping incoming FIN packets. 53.56% server [kernel.kallsyms] [k] queued_spin_lock_slowpath | ---queued_spin_lock_slowpath | --53.51%--_raw_spin_lock_irqsave | --53.51%--__percpu_counter_sum tcp_check_oom | |--39.03%--__tcp_close | tcp_close | inet_release | inet6_release | sock_close | __fput | ____fput | task_work_run | exit_to_usermode_loop | do_syscall_64 | entry_SYSCALL_64_after_hwframe | __GI___libc_close | --14.48%--tcp_out_of_resources tcp_write_timeout tcp_retransmit_timer tcp_write_timer_handler tcp_write_timer call_timer_fn expire_timers __run_timers run_timer_softirq __softirqentry_text_start As explained in commit cf86a086a180 ("net/dst: use a smaller percpu_counter batch for dst entries accounting"), default batch size is too big for the default value of tcp_max_orphans (262144). But even if we reduce batch sizes, there would still be cases where the estimated count of orphans is beyond the limit, and where tcp_too_many_orphans() has to call the expensive percpu_counter_sum_positive(). One solution is to use plain per-cpu counters, and have a timer to periodically refresh this cache. Updating this cache every 100ms seems about right, tcp pressure state is not radically changing over shorter periods. percpu_counter was nice 15 years ago while hosts had less than 16 cpus, not anymore by current standards. v2: Fix the build issue for CONFIG_CRYPTO_DEV_CHELSIO_TLS=m, reported by kernel test robot <lkp@intel.com> Remove unused socket argument from tcp_too_many_orphans() Fixes: dd24c00191d5 ("net: Use a percpu_counter for orphan_count") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Stefan Bach <sfb@google.com> Cc: Neal Cardwell <ncardwell@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-10-14 21:41:26 +08:00
}
bool tcp_check_oom(struct sock *sk, int shift)
{
bool too_many_orphans, out_of_socket_memory;
tcp: switch orphan_count to bare per-cpu counters Use of percpu_counter structure to track count of orphaned sockets is causing problems on modern hosts with 256 cpus or more. Stefan Bach reported a serious spinlock contention in real workloads, that I was able to reproduce with a netfilter rule dropping incoming FIN packets. 53.56% server [kernel.kallsyms] [k] queued_spin_lock_slowpath | ---queued_spin_lock_slowpath | --53.51%--_raw_spin_lock_irqsave | --53.51%--__percpu_counter_sum tcp_check_oom | |--39.03%--__tcp_close | tcp_close | inet_release | inet6_release | sock_close | __fput | ____fput | task_work_run | exit_to_usermode_loop | do_syscall_64 | entry_SYSCALL_64_after_hwframe | __GI___libc_close | --14.48%--tcp_out_of_resources tcp_write_timeout tcp_retransmit_timer tcp_write_timer_handler tcp_write_timer call_timer_fn expire_timers __run_timers run_timer_softirq __softirqentry_text_start As explained in commit cf86a086a180 ("net/dst: use a smaller percpu_counter batch for dst entries accounting"), default batch size is too big for the default value of tcp_max_orphans (262144). But even if we reduce batch sizes, there would still be cases where the estimated count of orphans is beyond the limit, and where tcp_too_many_orphans() has to call the expensive percpu_counter_sum_positive(). One solution is to use plain per-cpu counters, and have a timer to periodically refresh this cache. Updating this cache every 100ms seems about right, tcp pressure state is not radically changing over shorter periods. percpu_counter was nice 15 years ago while hosts had less than 16 cpus, not anymore by current standards. v2: Fix the build issue for CONFIG_CRYPTO_DEV_CHELSIO_TLS=m, reported by kernel test robot <lkp@intel.com> Remove unused socket argument from tcp_too_many_orphans() Fixes: dd24c00191d5 ("net: Use a percpu_counter for orphan_count") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Stefan Bach <sfb@google.com> Cc: Neal Cardwell <ncardwell@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-10-14 21:41:26 +08:00
too_many_orphans = tcp_too_many_orphans(shift);
out_of_socket_memory = tcp_out_of_memory(sk);
if (too_many_orphans)
net_info_ratelimited("too many orphaned sockets\n");
if (out_of_socket_memory)
net_info_ratelimited("out of memory -- consider tuning tcp_mem\n");
return too_many_orphans || out_of_socket_memory;
}
void __tcp_close(struct sock *sk, long timeout)
{
struct sk_buff *skb;
int data_was_unread = 0;
[TCP]: Fix sock_orphan dead lock Calling sock_orphan inside bh_lock_sock in tcp_close can lead to dead locks. For example, the inet_diag code holds sk_callback_lock without disabling BH. If an inbound packet arrives during that admittedly tiny window, it will cause a dead lock on bh_lock_sock. Another possible path would be through sock_wfree if the network device driver frees the tx skb in process context with BH enabled. We can fix this by moving sock_orphan out of bh_lock_sock. The tricky bit is to work out when we need to destroy the socket ourselves and when it has already been destroyed by someone else. By moving sock_orphan before the release_sock we can solve this problem. This is because as long as we own the socket lock its state cannot change. So we simply record the socket state before the release_sock and then check the state again after we regain the socket lock. If the socket state has transitioned to TCP_CLOSE in the time being, we know that the socket has been destroyed. Otherwise the socket is still ours to keep. Note that I've also moved the increment on the orphan count forward. This may look like a problem as we're increasing it even if the socket is just about to be destroyed where it'll be decreased again. However, this simply enlarges a window that already exists. This also changes the orphan count test by one. Considering what the orphan count is meant to do this is no big deal. This problem was discoverd by Ingo Molnar using his lock validator. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-05-04 14:31:35 +08:00
int state;
tcp: add annotations around sk->sk_shutdown accesses Now sk->sk_shutdown is no longer a bitfield, we can add standard READ_ONCE()/WRITE_ONCE() annotations to silence KCSAN reports like the following: BUG: KCSAN: data-race in tcp_disconnect / tcp_poll write to 0xffff88814588582c of 1 bytes by task 3404 on cpu 1: tcp_disconnect+0x4d6/0xdb0 net/ipv4/tcp.c:3121 __inet_stream_connect+0x5dd/0x6e0 net/ipv4/af_inet.c:715 inet_stream_connect+0x48/0x70 net/ipv4/af_inet.c:727 __sys_connect_file net/socket.c:2001 [inline] __sys_connect+0x19b/0x1b0 net/socket.c:2018 __do_sys_connect net/socket.c:2028 [inline] __se_sys_connect net/socket.c:2025 [inline] __x64_sys_connect+0x41/0x50 net/socket.c:2025 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd read to 0xffff88814588582c of 1 bytes by task 3374 on cpu 0: tcp_poll+0x2e6/0x7d0 net/ipv4/tcp.c:562 sock_poll+0x253/0x270 net/socket.c:1383 vfs_poll include/linux/poll.h:88 [inline] io_poll_check_events io_uring/poll.c:281 [inline] io_poll_task_func+0x15a/0x820 io_uring/poll.c:333 handle_tw_list io_uring/io_uring.c:1184 [inline] tctx_task_work+0x1fe/0x4d0 io_uring/io_uring.c:1246 task_work_run+0x123/0x160 kernel/task_work.c:179 get_signal+0xe64/0xff0 kernel/signal.c:2635 arch_do_signal_or_restart+0x89/0x2a0 arch/x86/kernel/signal.c:306 exit_to_user_mode_loop+0x6f/0xe0 kernel/entry/common.c:168 exit_to_user_mode_prepare+0x6c/0xb0 kernel/entry/common.c:204 __syscall_exit_to_user_mode_work kernel/entry/common.c:286 [inline] syscall_exit_to_user_mode+0x26/0x140 kernel/entry/common.c:297 do_syscall_64+0x4d/0xc0 arch/x86/entry/common.c:86 entry_SYSCALL_64_after_hwframe+0x63/0xcd value changed: 0x03 -> 0x00 Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-05-10 04:36:56 +08:00
WRITE_ONCE(sk->sk_shutdown, SHUTDOWN_MASK);
if (sk->sk_state == TCP_LISTEN) {
tcp_set_state(sk, TCP_CLOSE);
/* Special case. */
inet_csk_listen_stop(sk);
goto adjudge_to_death;
}
/* We need to flush the recv. buffs. We do this only on the
* descriptor close, not protocol-sourced closes, because the
* reader process may not have drained the data yet!
*/
while ((skb = __skb_dequeue(&sk->sk_receive_queue)) != NULL) {
u32 len = TCP_SKB_CB(skb)->end_seq - TCP_SKB_CB(skb)->seq;
if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)
len--;
data_was_unread += len;
__kfree_skb(skb);
}
tcp: do not send reset to already closed sockets i've found that tcp_close() can be called for an already closed socket, but still sends reset in this case (tcp_send_active_reset()) which seems to be incorrect. Moreover, a packet with reset is sent with different source port as original port number has been already cleared on socket. Besides that incrementing stat counter for LINUX_MIB_TCPABORTONCLOSE also does not look correct in this case. Initially this issue was found on 2.6.18-x RHEL5 kernel, but the same seems to be true for the current mainstream kernel (checked on 2.6.35-rc3). Please, correct me if i missed something. How that happens: 1) the server receives a packet for socket in TCP_CLOSE_WAIT state that triggers a tcp_reset(): Call Trace: <IRQ> [<ffffffff8025b9b9>] tcp_reset+0x12f/0x1e8 [<ffffffff80046125>] tcp_rcv_state_process+0x1c0/0xa08 [<ffffffff8003eb22>] tcp_v4_do_rcv+0x310/0x37a [<ffffffff80028bea>] tcp_v4_rcv+0x74d/0xb43 [<ffffffff8024ef4c>] ip_local_deliver_finish+0x0/0x259 [<ffffffff80037131>] ip_local_deliver+0x200/0x2f4 [<ffffffff8003843c>] ip_rcv+0x64c/0x69f [<ffffffff80021d89>] netif_receive_skb+0x4c4/0x4fa [<ffffffff80032eca>] process_backlog+0x90/0xec [<ffffffff8000cc50>] net_rx_action+0xbb/0x1f1 [<ffffffff80012d3a>] __do_softirq+0xf5/0x1ce [<ffffffff8001147a>] handle_IRQ_event+0x56/0xb0 [<ffffffff8006334c>] call_softirq+0x1c/0x28 [<ffffffff80070476>] do_softirq+0x2c/0x85 [<ffffffff80070441>] do_IRQ+0x149/0x152 [<ffffffff80062665>] ret_from_intr+0x0/0xa <EOI> [<ffffffff80008a2e>] __handle_mm_fault+0x6cd/0x1303 [<ffffffff80008903>] __handle_mm_fault+0x5a2/0x1303 [<ffffffff80033a9d>] cache_free_debugcheck+0x21f/0x22e [<ffffffff8006a263>] do_page_fault+0x49a/0x7dc [<ffffffff80066487>] thread_return+0x89/0x174 [<ffffffff800c5aee>] audit_syscall_exit+0x341/0x35c [<ffffffff80062e39>] error_exit+0x0/0x84 tcp_rcv_state_process() ... // (sk_state == TCP_CLOSE_WAIT here) ... /* step 2: check RST bit */ if(th->rst) { tcp_reset(sk); goto discard; } ... --------------------------------- tcp_rcv_state_process tcp_reset tcp_done tcp_set_state(sk, TCP_CLOSE); inet_put_port __inet_put_port inet_sk(sk)->num = 0; sk->sk_shutdown = SHUTDOWN_MASK; 2) After that the process (socket owner) tries to write something to that socket and "inet_autobind" sets a _new_ (which differs from the original!) port number for the socket: Call Trace: [<ffffffff80255a12>] inet_bind_hash+0x33/0x5f [<ffffffff80257180>] inet_csk_get_port+0x216/0x268 [<ffffffff8026bcc9>] inet_autobind+0x22/0x8f [<ffffffff80049140>] inet_sendmsg+0x27/0x57 [<ffffffff8003a9d9>] do_sock_write+0xae/0xea [<ffffffff80226ac7>] sock_writev+0xdc/0xf6 [<ffffffff800680c7>] _spin_lock_irqsave+0x9/0xe [<ffffffff8001fb49>] __pollwait+0x0/0xdd [<ffffffff8008d533>] default_wake_function+0x0/0xe [<ffffffff800a4f10>] autoremove_wake_function+0x0/0x2e [<ffffffff800f0b49>] do_readv_writev+0x163/0x274 [<ffffffff80066538>] thread_return+0x13a/0x174 [<ffffffff800145d8>] tcp_poll+0x0/0x1c9 [<ffffffff800c56d3>] audit_syscall_entry+0x180/0x1b3 [<ffffffff800f0dd0>] sys_writev+0x49/0xe4 [<ffffffff800622dd>] tracesys+0xd5/0xe0 3) sendmsg fails at last with -EPIPE (=> 'write' returns -EPIPE in userspace): F: tcp_sendmsg1 -EPIPE: sk=ffff81000bda00d0, sport=49847, old_state=7, new_state=7, sk_err=0, sk_shutdown=3 Call Trace: [<ffffffff80027557>] tcp_sendmsg+0xcb/0xe87 [<ffffffff80033300>] release_sock+0x10/0xae [<ffffffff8016f20f>] vgacon_cursor+0x0/0x1a7 [<ffffffff8026bd32>] inet_autobind+0x8b/0x8f [<ffffffff8003a9d9>] do_sock_write+0xae/0xea [<ffffffff80226ac7>] sock_writev+0xdc/0xf6 [<ffffffff800680c7>] _spin_lock_irqsave+0x9/0xe [<ffffffff8001fb49>] __pollwait+0x0/0xdd [<ffffffff8008d533>] default_wake_function+0x0/0xe [<ffffffff800a4f10>] autoremove_wake_function+0x0/0x2e [<ffffffff800f0b49>] do_readv_writev+0x163/0x274 [<ffffffff80066538>] thread_return+0x13a/0x174 [<ffffffff800145d8>] tcp_poll+0x0/0x1c9 [<ffffffff800c56d3>] audit_syscall_entry+0x180/0x1b3 [<ffffffff800f0dd0>] sys_writev+0x49/0xe4 [<ffffffff800622dd>] tracesys+0xd5/0xe0 tcp_sendmsg() ... /* Wait for a connection to finish. */ if ((1 << sk->sk_state) & ~(TCPF_ESTABLISHED | TCPF_CLOSE_WAIT)) { int old_state = sk->sk_state; if ((err = sk_stream_wait_connect(sk, &timeo)) != 0) { if (f_d && (err == -EPIPE)) { printk("F: tcp_sendmsg1 -EPIPE: sk=%p, sport=%u, old_state=%d, new_state=%d, " "sk_err=%d, sk_shutdown=%d\n", sk, ntohs(inet_sk(sk)->sport), old_state, sk->sk_state, sk->sk_err, sk->sk_shutdown); dump_stack(); } goto out_err; } } ... 4) Then the process (socket owner) understands that it's time to close that socket and does that (and thus triggers sending reset packet): Call Trace: ... [<ffffffff80032077>] dev_queue_xmit+0x343/0x3d6 [<ffffffff80034698>] ip_output+0x351/0x384 [<ffffffff80251ae9>] dst_output+0x0/0xe [<ffffffff80036ec6>] ip_queue_xmit+0x567/0x5d2 [<ffffffff80095700>] vprintk+0x21/0x33 [<ffffffff800070f0>] check_poison_obj+0x2e/0x206 [<ffffffff80013587>] poison_obj+0x36/0x45 [<ffffffff8025dea6>] tcp_send_active_reset+0x15/0x14d [<ffffffff80023481>] dbg_redzone1+0x1c/0x25 [<ffffffff8025dea6>] tcp_send_active_reset+0x15/0x14d [<ffffffff8000ca94>] cache_alloc_debugcheck_after+0x189/0x1c8 [<ffffffff80023405>] tcp_transmit_skb+0x764/0x786 [<ffffffff8025df8a>] tcp_send_active_reset+0xf9/0x14d [<ffffffff80258ff1>] tcp_close+0x39a/0x960 [<ffffffff8026be12>] inet_release+0x69/0x80 [<ffffffff80059b31>] sock_release+0x4f/0xcf [<ffffffff80059d4c>] sock_close+0x2c/0x30 [<ffffffff800133c9>] __fput+0xac/0x197 [<ffffffff800252bc>] filp_close+0x59/0x61 [<ffffffff8001eff6>] sys_close+0x85/0xc7 [<ffffffff800622dd>] tracesys+0xd5/0xe0 So, in brief: * a received packet for socket in TCP_CLOSE_WAIT state triggers tcp_reset() which clears inet_sk(sk)->num and put socket into TCP_CLOSE state * an attempt to write to that socket forces inet_autobind() to get a new port (but the write itself fails with -EPIPE) * tcp_close() called for socket in TCP_CLOSE state sends an active reset via socket with newly allocated port This adds an additional check in tcp_close() for already closed sockets. We do not want to send anything to closed sockets. Signed-off-by: Konstantin Khorenko <khorenko@openvz.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-06-25 12:54:58 +08:00
/* If socket has been already reset (e.g. in tcp_reset()) - kill it. */
if (sk->sk_state == TCP_CLOSE)
goto adjudge_to_death;
/* As outlined in RFC 2525, section 2.17, we send a RST here because
* data was lost. To witness the awful effects of the old behavior of
* always doing a FIN, run an older 2.1.x kernel or 2.0.x, start a bulk
* GET in an FTP client, suspend the process, wait for the client to
* advertise a zero window, then kill -9 the FTP client, wheee...
* Note: timeout is always zero in such a case.
*/
if (unlikely(tcp_sk(sk)->repair)) {
sk->sk_prot->disconnect(sk, 0);
} else if (data_was_unread) {
/* Unread data was tossed, zap the connection. */
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTONCLOSE);
tcp_set_state(sk, TCP_CLOSE);
tcp_send_active_reset(sk, sk->sk_allocation);
} else if (sock_flag(sk, SOCK_LINGER) && !sk->sk_lingertime) {
/* Check zero linger _after_ checking for unread data. */
sk->sk_prot->disconnect(sk, 0);
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTONDATA);
} else if (tcp_close_state(sk)) {
/* We FIN if the application ate all the data before
* zapping the connection.
*/
/* RED-PEN. Formally speaking, we have broken TCP state
* machine. State transitions:
*
* TCP_ESTABLISHED -> TCP_FIN_WAIT1
* TCP_SYN_RECV -> TCP_FIN_WAIT1 (forget it, it's impossible)
* TCP_CLOSE_WAIT -> TCP_LAST_ACK
*
* are legal only when FIN has been sent (i.e. in window),
* rather than queued out of window. Purists blame.
*
* F.e. "RFC state" is ESTABLISHED,
* if Linux state is FIN-WAIT-1, but FIN is still not sent.
*
* The visible declinations are that sometimes
* we enter time-wait state, when it is not required really
* (harmless), do not send active resets, when they are
* required by specs (TCP_ESTABLISHED, TCP_CLOSE_WAIT, when
* they look as CLOSING or LAST_ACK for Linux)
* Probably, I missed some more holelets.
* --ANK
* XXX (TFO) - To start off we don't support SYN+ACK+FIN
* in a single packet! (May consider it later but will
* probably need API support or TCP_CORK SYN-ACK until
* data is written and socket is closed.)
*/
tcp_send_fin(sk);
}
sk_stream_wait_close(sk, timeout);
adjudge_to_death:
[TCP]: Fix sock_orphan dead lock Calling sock_orphan inside bh_lock_sock in tcp_close can lead to dead locks. For example, the inet_diag code holds sk_callback_lock without disabling BH. If an inbound packet arrives during that admittedly tiny window, it will cause a dead lock on bh_lock_sock. Another possible path would be through sock_wfree if the network device driver frees the tx skb in process context with BH enabled. We can fix this by moving sock_orphan out of bh_lock_sock. The tricky bit is to work out when we need to destroy the socket ourselves and when it has already been destroyed by someone else. By moving sock_orphan before the release_sock we can solve this problem. This is because as long as we own the socket lock its state cannot change. So we simply record the socket state before the release_sock and then check the state again after we regain the socket lock. If the socket state has transitioned to TCP_CLOSE in the time being, we know that the socket has been destroyed. Otherwise the socket is still ours to keep. Note that I've also moved the increment on the orphan count forward. This may look like a problem as we're increasing it even if the socket is just about to be destroyed where it'll be decreased again. However, this simply enlarges a window that already exists. This also changes the orphan count test by one. Considering what the orphan count is meant to do this is no big deal. This problem was discoverd by Ingo Molnar using his lock validator. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-05-04 14:31:35 +08:00
state = sk->sk_state;
sock_hold(sk);
sock_orphan(sk);
local_bh_disable();
bh_lock_sock(sk);
/* remove backlog if any, without releasing ownership. */
__release_sock(sk);
tcp: switch orphan_count to bare per-cpu counters Use of percpu_counter structure to track count of orphaned sockets is causing problems on modern hosts with 256 cpus or more. Stefan Bach reported a serious spinlock contention in real workloads, that I was able to reproduce with a netfilter rule dropping incoming FIN packets. 53.56% server [kernel.kallsyms] [k] queued_spin_lock_slowpath | ---queued_spin_lock_slowpath | --53.51%--_raw_spin_lock_irqsave | --53.51%--__percpu_counter_sum tcp_check_oom | |--39.03%--__tcp_close | tcp_close | inet_release | inet6_release | sock_close | __fput | ____fput | task_work_run | exit_to_usermode_loop | do_syscall_64 | entry_SYSCALL_64_after_hwframe | __GI___libc_close | --14.48%--tcp_out_of_resources tcp_write_timeout tcp_retransmit_timer tcp_write_timer_handler tcp_write_timer call_timer_fn expire_timers __run_timers run_timer_softirq __softirqentry_text_start As explained in commit cf86a086a180 ("net/dst: use a smaller percpu_counter batch for dst entries accounting"), default batch size is too big for the default value of tcp_max_orphans (262144). But even if we reduce batch sizes, there would still be cases where the estimated count of orphans is beyond the limit, and where tcp_too_many_orphans() has to call the expensive percpu_counter_sum_positive(). One solution is to use plain per-cpu counters, and have a timer to periodically refresh this cache. Updating this cache every 100ms seems about right, tcp pressure state is not radically changing over shorter periods. percpu_counter was nice 15 years ago while hosts had less than 16 cpus, not anymore by current standards. v2: Fix the build issue for CONFIG_CRYPTO_DEV_CHELSIO_TLS=m, reported by kernel test robot <lkp@intel.com> Remove unused socket argument from tcp_too_many_orphans() Fixes: dd24c00191d5 ("net: Use a percpu_counter for orphan_count") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Stefan Bach <sfb@google.com> Cc: Neal Cardwell <ncardwell@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-10-14 21:41:26 +08:00
this_cpu_inc(tcp_orphan_count);
[TCP]: Fix sock_orphan dead lock Calling sock_orphan inside bh_lock_sock in tcp_close can lead to dead locks. For example, the inet_diag code holds sk_callback_lock without disabling BH. If an inbound packet arrives during that admittedly tiny window, it will cause a dead lock on bh_lock_sock. Another possible path would be through sock_wfree if the network device driver frees the tx skb in process context with BH enabled. We can fix this by moving sock_orphan out of bh_lock_sock. The tricky bit is to work out when we need to destroy the socket ourselves and when it has already been destroyed by someone else. By moving sock_orphan before the release_sock we can solve this problem. This is because as long as we own the socket lock its state cannot change. So we simply record the socket state before the release_sock and then check the state again after we regain the socket lock. If the socket state has transitioned to TCP_CLOSE in the time being, we know that the socket has been destroyed. Otherwise the socket is still ours to keep. Note that I've also moved the increment on the orphan count forward. This may look like a problem as we're increasing it even if the socket is just about to be destroyed where it'll be decreased again. However, this simply enlarges a window that already exists. This also changes the orphan count test by one. Considering what the orphan count is meant to do this is no big deal. This problem was discoverd by Ingo Molnar using his lock validator. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-05-04 14:31:35 +08:00
/* Have we already been destroyed by a softirq or backlog? */
if (state != TCP_CLOSE && sk->sk_state == TCP_CLOSE)
goto out;
/* This is a (useful) BSD violating of the RFC. There is a
* problem with TCP as specified in that the other end could
* keep a socket open forever with no application left this end.
* We use a 1 minute timeout (about the same as BSD) then kill
* our end. If they send after that then tough - BUT: long enough
* that we won't make the old 4*rto = almost no time - whoops
* reset mistake.
*
* Nope, it was not mistake. It is really desired behaviour
* f.e. on http servers, when such sockets are useless, but
* consume significant resources. Let's do it with special
* linger2 option. --ANK
*/
if (sk->sk_state == TCP_FIN_WAIT2) {
struct tcp_sock *tp = tcp_sk(sk);
if (READ_ONCE(tp->linger2) < 0) {
tcp_set_state(sk, TCP_CLOSE);
tcp_send_active_reset(sk, GFP_ATOMIC);
__NET_INC_STATS(sock_net(sk),
LINUX_MIB_TCPABORTONLINGER);
} else {
const int tmo = tcp_fin_time(sk);
if (tmo > TCP_TIMEWAIT_LEN) {
inet_csk_reset_keepalive_timer(sk,
tmo - TCP_TIMEWAIT_LEN);
} else {
tcp_time_wait(sk, TCP_FIN_WAIT2, tmo);
goto out;
}
}
}
if (sk->sk_state != TCP_CLOSE) {
if (tcp_check_oom(sk, 0)) {
tcp_set_state(sk, TCP_CLOSE);
tcp_send_active_reset(sk, GFP_ATOMIC);
__NET_INC_STATS(sock_net(sk),
LINUX_MIB_TCPABORTONMEMORY);
net: tcp: close sock if net namespace is exiting When a tcp socket is closed, if it detects that its net namespace is exiting, close immediately and do not wait for FIN sequence. For normal sockets, a reference is taken to their net namespace, so it will never exit while the socket is open. However, kernel sockets do not take a reference to their net namespace, so it may begin exiting while the kernel socket is still open. In this case if the kernel socket is a tcp socket, it will stay open trying to complete its close sequence. The sock's dst(s) hold a reference to their interface, which are all transferred to the namespace's loopback interface when the real interfaces are taken down. When the namespace tries to take down its loopback interface, it hangs waiting for all references to the loopback interface to release, which results in messages like: unregister_netdevice: waiting for lo to become free. Usage count = 1 These messages continue until the socket finally times out and closes. Since the net namespace cleanup holds the net_mutex while calling its registered pernet callbacks, any new net namespace initialization is blocked until the current net namespace finishes exiting. After this change, the tcp socket notices the exiting net namespace, and closes immediately, releasing its dst(s) and their reference to the loopback interface, which lets the net namespace continue exiting. Link: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1711407 Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=97811 Signed-off-by: Dan Streetman <ddstreet@canonical.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-01-19 05:14:26 +08:00
} else if (!check_net(sock_net(sk))) {
/* Not possible to send reset; just close */
tcp_set_state(sk, TCP_CLOSE);
}
}
if (sk->sk_state == TCP_CLOSE) {
struct request_sock *req;
req = rcu_dereference_protected(tcp_sk(sk)->fastopen_rsk,
lockdep_sock_is_held(sk));
/* We could get here with a non-NULL req if the socket is
* aborted (e.g., closed with unread data) before 3WHS
* finishes.
*/
if (req)
reqsk_fastopen_remove(sk, req, false);
inet_csk_destroy_sock(sk);
}
/* Otherwise, socket is reprieved until protocol close. */
out:
bh_unlock_sock(sk);
local_bh_enable();
}
void tcp_close(struct sock *sk, long timeout)
{
lock_sock(sk);
__tcp_close(sk, timeout);
release_sock(sk);
sock_put(sk);
}
EXPORT_SYMBOL(tcp_close);
/* These states need RST on ABORT according to RFC793 */
static inline bool tcp_need_reset(int state)
{
return (1 << state) &
(TCPF_ESTABLISHED | TCPF_CLOSE_WAIT | TCPF_FIN_WAIT1 |
Revert "tcp: Reset tcp connections in SYN-SENT state" This reverts commit e880f8b3a24a73704731a7227ed5fee14bd90192. 1) Patch has not been properly tested, and is wrong [1] 2) Patch submission did not include TCP maintainer (this is me) [1] divide error: 0000 [#1] PREEMPT SMP KASAN CPU: 0 PID: 8426 Comm: syz-executor478 Not tainted 5.12.0-rc4-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 RIP: 0010:__tcp_select_window+0x56d/0xad0 net/ipv4/tcp_output.c:3015 Code: 44 89 ff e8 d5 cd f0 f9 45 39 e7 0f 8d 20 ff ff ff e8 f7 c7 f0 f9 44 89 e3 e9 13 ff ff ff e8 ea c7 f0 f9 44 89 e0 44 89 e3 99 <f7> 7c 24 04 29 d3 e9 fc fe ff ff e8 d3 c7 f0 f9 41 f7 dc bf 1f 00 RSP: 0018:ffffc9000184fac0 EFLAGS: 00010293 RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffffffff87832e76 RDI: 0000000000000003 RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 R10: ffffffff87832e14 R11: 0000000000000000 R12: 0000000000000000 R13: 1ffff92000309f5c R14: 0000000000000000 R15: 0000000000000000 FS: 00000000023eb300(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fc2b5f426c0 CR3: 000000001c5cf000 CR4: 00000000001506f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: tcp_select_window net/ipv4/tcp_output.c:264 [inline] __tcp_transmit_skb+0xa82/0x38f0 net/ipv4/tcp_output.c:1351 tcp_transmit_skb net/ipv4/tcp_output.c:1423 [inline] tcp_send_active_reset+0x475/0x8e0 net/ipv4/tcp_output.c:3449 tcp_disconnect+0x15a9/0x1e60 net/ipv4/tcp.c:2955 inet_shutdown+0x260/0x430 net/ipv4/af_inet.c:905 __sys_shutdown_sock net/socket.c:2189 [inline] __sys_shutdown_sock net/socket.c:2183 [inline] __sys_shutdown+0xf1/0x1b0 net/socket.c:2201 __do_sys_shutdown net/socket.c:2209 [inline] __se_sys_shutdown net/socket.c:2207 [inline] __x64_sys_shutdown+0x50/0x70 net/socket.c:2207 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46 entry_SYSCALL_64_after_hwframe+0x44/0xae Fixes: e880f8b3a24a ("tcp: Reset tcp connections in SYN-SENT state") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Cc: Manoj Basapathi <manojbm@codeaurora.org> Cc: Sauvik Saha <ssaha@codeaurora.org> Link: https://lore.kernel.org/r/20210409170237.274904-1-eric.dumazet@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-04-10 01:02:37 +08:00
TCPF_FIN_WAIT2 | TCPF_SYN_RECV);
}
tcp: implement rb-tree based retransmit queue Using a linear list to store all skbs in write queue has been okay for quite a while : O(N) is not too bad when N < 500. Things get messy when N is the order of 100,000 : Modern TCP stacks want 10Gbit+ of throughput even with 200 ms RTT flows. 40 ns per cache line miss means a full scan can use 4 ms, blowing away CPU caches. SACK processing often can use various hints to avoid parsing whole retransmit queue. But with high packet losses and/or high reordering, hints no longer work. Sender has to process thousands of unfriendly SACK, accumulating a huge socket backlog, burning a cpu and massively dropping packets. Using an rb-tree for retransmit queue has been avoided for years because it added complexity and overhead, but now is the time to be more resistant and say no to quadratic behavior. 1) RTX queue is no longer part of the write queue : already sent skbs are stored in one rb-tree. 2) Since reaching the head of write queue no longer needs sk->sk_send_head, we added an union of sk_send_head and tcp_rtx_queue Tested: On receiver : netem on ingress : delay 150ms 200us loss 1 GRO disabled to force stress and SACK storms. for f in `seq 1 10` do ./netperf -H lpaa6 -l30 -- -K bbr -o THROUGHPUT|tail -1 done | awk '{print $0} {sum += $0} END {printf "%7u\n",sum}' Before patch : 323.87 351.48 339.59 338.62 306.72 204.07 304.93 291.88 202.47 176.88 2840 After patch: 1700.83 2207.98 2070.17 1544.26 2114.76 2124.89 1693.14 1080.91 2216.82 1299.94 18053 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-06 13:21:27 +08:00
static void tcp_rtx_queue_purge(struct sock *sk)
{
struct rb_node *p = rb_first(&sk->tcp_rtx_queue);
tcp: do not leave dangling pointers in tp->highest_sack Latest commit 853697504de0 ("tcp: Fix highest_sack and highest_sack_seq") apparently allowed syzbot to trigger various crashes in TCP stack [1] I believe this commit only made things easier for syzbot to find its way into triggering use-after-frees. But really the bugs could lead to bad TCP behavior or even plain crashes even for non malicious peers. I have audited all calls to tcp_rtx_queue_unlink() and tcp_rtx_queue_unlink_and_free() and made sure tp->highest_sack would be updated if we are removing from rtx queue the skb that tp->highest_sack points to. These updates were missing in three locations : 1) tcp_clean_rtx_queue() [This one seems quite serious, I have no idea why this was not caught earlier] 2) tcp_rtx_queue_purge() [Probably not a big deal for normal operations] 3) tcp_send_synack() [Probably not a big deal for normal operations] [1] BUG: KASAN: use-after-free in tcp_highest_sack_seq include/net/tcp.h:1864 [inline] BUG: KASAN: use-after-free in tcp_highest_sack_seq include/net/tcp.h:1856 [inline] BUG: KASAN: use-after-free in tcp_check_sack_reordering+0x33c/0x3a0 net/ipv4/tcp_input.c:891 Read of size 4 at addr ffff8880a488d068 by task ksoftirqd/1/16 CPU: 1 PID: 16 Comm: ksoftirqd/1 Not tainted 5.5.0-rc5-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x197/0x210 lib/dump_stack.c:118 print_address_description.constprop.0.cold+0xd4/0x30b mm/kasan/report.c:374 __kasan_report.cold+0x1b/0x41 mm/kasan/report.c:506 kasan_report+0x12/0x20 mm/kasan/common.c:639 __asan_report_load4_noabort+0x14/0x20 mm/kasan/generic_report.c:134 tcp_highest_sack_seq include/net/tcp.h:1864 [inline] tcp_highest_sack_seq include/net/tcp.h:1856 [inline] tcp_check_sack_reordering+0x33c/0x3a0 net/ipv4/tcp_input.c:891 tcp_try_undo_partial net/ipv4/tcp_input.c:2730 [inline] tcp_fastretrans_alert+0xf74/0x23f0 net/ipv4/tcp_input.c:2847 tcp_ack+0x2577/0x5bf0 net/ipv4/tcp_input.c:3710 tcp_rcv_established+0x6dd/0x1e90 net/ipv4/tcp_input.c:5706 tcp_v4_do_rcv+0x619/0x8d0 net/ipv4/tcp_ipv4.c:1619 tcp_v4_rcv+0x307f/0x3b40 net/ipv4/tcp_ipv4.c:2001 ip_protocol_deliver_rcu+0x5a/0x880 net/ipv4/ip_input.c:204 ip_local_deliver_finish+0x23b/0x380 net/ipv4/ip_input.c:231 NF_HOOK include/linux/netfilter.h:307 [inline] NF_HOOK include/linux/netfilter.h:301 [inline] ip_local_deliver+0x1e9/0x520 net/ipv4/ip_input.c:252 dst_input include/net/dst.h:442 [inline] ip_rcv_finish+0x1db/0x2f0 net/ipv4/ip_input.c:428 NF_HOOK include/linux/netfilter.h:307 [inline] NF_HOOK include/linux/netfilter.h:301 [inline] ip_rcv+0xe8/0x3f0 net/ipv4/ip_input.c:538 __netif_receive_skb_one_core+0x113/0x1a0 net/core/dev.c:5148 __netif_receive_skb+0x2c/0x1d0 net/core/dev.c:5262 process_backlog+0x206/0x750 net/core/dev.c:6093 napi_poll net/core/dev.c:6530 [inline] net_rx_action+0x508/0x1120 net/core/dev.c:6598 __do_softirq+0x262/0x98c kernel/softirq.c:292 run_ksoftirqd kernel/softirq.c:603 [inline] run_ksoftirqd+0x8e/0x110 kernel/softirq.c:595 smpboot_thread_fn+0x6a3/0xa40 kernel/smpboot.c:165 kthread+0x361/0x430 kernel/kthread.c:255 ret_from_fork+0x24/0x30 arch/x86/entry/entry_64.S:352 Allocated by task 10091: save_stack+0x23/0x90 mm/kasan/common.c:72 set_track mm/kasan/common.c:80 [inline] __kasan_kmalloc mm/kasan/common.c:513 [inline] __kasan_kmalloc.constprop.0+0xcf/0xe0 mm/kasan/common.c:486 kasan_slab_alloc+0xf/0x20 mm/kasan/common.c:521 slab_post_alloc_hook mm/slab.h:584 [inline] slab_alloc_node mm/slab.c:3263 [inline] kmem_cache_alloc_node+0x138/0x740 mm/slab.c:3575 __alloc_skb+0xd5/0x5e0 net/core/skbuff.c:198 alloc_skb_fclone include/linux/skbuff.h:1099 [inline] sk_stream_alloc_skb net/ipv4/tcp.c:875 [inline] sk_stream_alloc_skb+0x113/0xc90 net/ipv4/tcp.c:852 tcp_sendmsg_locked+0xcf9/0x3470 net/ipv4/tcp.c:1282 tcp_sendmsg+0x30/0x50 net/ipv4/tcp.c:1432 inet_sendmsg+0x9e/0xe0 net/ipv4/af_inet.c:807 sock_sendmsg_nosec net/socket.c:652 [inline] sock_sendmsg+0xd7/0x130 net/socket.c:672 __sys_sendto+0x262/0x380 net/socket.c:1998 __do_sys_sendto net/socket.c:2010 [inline] __se_sys_sendto net/socket.c:2006 [inline] __x64_sys_sendto+0xe1/0x1a0 net/socket.c:2006 do_syscall_64+0xfa/0x790 arch/x86/entry/common.c:294 entry_SYSCALL_64_after_hwframe+0x49/0xbe Freed by task 10095: save_stack+0x23/0x90 mm/kasan/common.c:72 set_track mm/kasan/common.c:80 [inline] kasan_set_free_info mm/kasan/common.c:335 [inline] __kasan_slab_free+0x102/0x150 mm/kasan/common.c:474 kasan_slab_free+0xe/0x10 mm/kasan/common.c:483 __cache_free mm/slab.c:3426 [inline] kmem_cache_free+0x86/0x320 mm/slab.c:3694 kfree_skbmem+0x178/0x1c0 net/core/skbuff.c:645 __kfree_skb+0x1e/0x30 net/core/skbuff.c:681 sk_eat_skb include/net/sock.h:2453 [inline] tcp_recvmsg+0x1252/0x2930 net/ipv4/tcp.c:2166 inet_recvmsg+0x136/0x610 net/ipv4/af_inet.c:838 sock_recvmsg_nosec net/socket.c:886 [inline] sock_recvmsg net/socket.c:904 [inline] sock_recvmsg+0xce/0x110 net/socket.c:900 __sys_recvfrom+0x1ff/0x350 net/socket.c:2055 __do_sys_recvfrom net/socket.c:2073 [inline] __se_sys_recvfrom net/socket.c:2069 [inline] __x64_sys_recvfrom+0xe1/0x1a0 net/socket.c:2069 do_syscall_64+0xfa/0x790 arch/x86/entry/common.c:294 entry_SYSCALL_64_after_hwframe+0x49/0xbe The buggy address belongs to the object at ffff8880a488d040 which belongs to the cache skbuff_fclone_cache of size 456 The buggy address is located 40 bytes inside of 456-byte region [ffff8880a488d040, ffff8880a488d208) The buggy address belongs to the page: page:ffffea0002922340 refcount:1 mapcount:0 mapping:ffff88821b057000 index:0x0 raw: 00fffe0000000200 ffffea00022a5788 ffffea0002624a48 ffff88821b057000 raw: 0000000000000000 ffff8880a488d040 0000000100000006 0000000000000000 page dumped because: kasan: bad access detected Memory state around the buggy address: ffff8880a488cf00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc ffff8880a488cf80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc >ffff8880a488d000: fc fc fc fc fc fc fc fc fb fb fb fb fb fb fb fb ^ ffff8880a488d080: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff8880a488d100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb Fixes: 853697504de0 ("tcp: Fix highest_sack and highest_sack_seq") Fixes: 50895b9de1d3 ("tcp: highest_sack fix") Fixes: 737ff314563c ("tcp: use sequence distance to detect reordering") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Cambda Zhu <cambda@linux.alibaba.com> Cc: Yuchung Cheng <ycheng@google.com> Cc: Neal Cardwell <ncardwell@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-01-23 13:03:00 +08:00
tcp_sk(sk)->highest_sack = NULL;
tcp: implement rb-tree based retransmit queue Using a linear list to store all skbs in write queue has been okay for quite a while : O(N) is not too bad when N < 500. Things get messy when N is the order of 100,000 : Modern TCP stacks want 10Gbit+ of throughput even with 200 ms RTT flows. 40 ns per cache line miss means a full scan can use 4 ms, blowing away CPU caches. SACK processing often can use various hints to avoid parsing whole retransmit queue. But with high packet losses and/or high reordering, hints no longer work. Sender has to process thousands of unfriendly SACK, accumulating a huge socket backlog, burning a cpu and massively dropping packets. Using an rb-tree for retransmit queue has been avoided for years because it added complexity and overhead, but now is the time to be more resistant and say no to quadratic behavior. 1) RTX queue is no longer part of the write queue : already sent skbs are stored in one rb-tree. 2) Since reaching the head of write queue no longer needs sk->sk_send_head, we added an union of sk_send_head and tcp_rtx_queue Tested: On receiver : netem on ingress : delay 150ms 200us loss 1 GRO disabled to force stress and SACK storms. for f in `seq 1 10` do ./netperf -H lpaa6 -l30 -- -K bbr -o THROUGHPUT|tail -1 done | awk '{print $0} {sum += $0} END {printf "%7u\n",sum}' Before patch : 323.87 351.48 339.59 338.62 306.72 204.07 304.93 291.88 202.47 176.88 2840 After patch: 1700.83 2207.98 2070.17 1544.26 2114.76 2124.89 1693.14 1080.91 2216.82 1299.94 18053 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-06 13:21:27 +08:00
while (p) {
struct sk_buff *skb = rb_to_skb(p);
p = rb_next(p);
/* Since we are deleting whole queue, no need to
* list_del(&skb->tcp_tsorted_anchor)
*/
tcp_rtx_queue_unlink(skb, sk);
tcp_wmem_free_skb(sk, skb);
tcp: implement rb-tree based retransmit queue Using a linear list to store all skbs in write queue has been okay for quite a while : O(N) is not too bad when N < 500. Things get messy when N is the order of 100,000 : Modern TCP stacks want 10Gbit+ of throughput even with 200 ms RTT flows. 40 ns per cache line miss means a full scan can use 4 ms, blowing away CPU caches. SACK processing often can use various hints to avoid parsing whole retransmit queue. But with high packet losses and/or high reordering, hints no longer work. Sender has to process thousands of unfriendly SACK, accumulating a huge socket backlog, burning a cpu and massively dropping packets. Using an rb-tree for retransmit queue has been avoided for years because it added complexity and overhead, but now is the time to be more resistant and say no to quadratic behavior. 1) RTX queue is no longer part of the write queue : already sent skbs are stored in one rb-tree. 2) Since reaching the head of write queue no longer needs sk->sk_send_head, we added an union of sk_send_head and tcp_rtx_queue Tested: On receiver : netem on ingress : delay 150ms 200us loss 1 GRO disabled to force stress and SACK storms. for f in `seq 1 10` do ./netperf -H lpaa6 -l30 -- -K bbr -o THROUGHPUT|tail -1 done | awk '{print $0} {sum += $0} END {printf "%7u\n",sum}' Before patch : 323.87 351.48 339.59 338.62 306.72 204.07 304.93 291.88 202.47 176.88 2840 After patch: 1700.83 2207.98 2070.17 1544.26 2114.76 2124.89 1693.14 1080.91 2216.82 1299.94 18053 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-06 13:21:27 +08:00
}
}
void tcp_write_queue_purge(struct sock *sk)
{
struct sk_buff *skb;
tcp_chrono_stop(sk, TCP_CHRONO_BUSY);
while ((skb = __skb_dequeue(&sk->sk_write_queue)) != NULL) {
tcp_skb_tsorted_anchor_cleanup(skb);
tcp_wmem_free_skb(sk, skb);
}
tcp: implement rb-tree based retransmit queue Using a linear list to store all skbs in write queue has been okay for quite a while : O(N) is not too bad when N < 500. Things get messy when N is the order of 100,000 : Modern TCP stacks want 10Gbit+ of throughput even with 200 ms RTT flows. 40 ns per cache line miss means a full scan can use 4 ms, blowing away CPU caches. SACK processing often can use various hints to avoid parsing whole retransmit queue. But with high packet losses and/or high reordering, hints no longer work. Sender has to process thousands of unfriendly SACK, accumulating a huge socket backlog, burning a cpu and massively dropping packets. Using an rb-tree for retransmit queue has been avoided for years because it added complexity and overhead, but now is the time to be more resistant and say no to quadratic behavior. 1) RTX queue is no longer part of the write queue : already sent skbs are stored in one rb-tree. 2) Since reaching the head of write queue no longer needs sk->sk_send_head, we added an union of sk_send_head and tcp_rtx_queue Tested: On receiver : netem on ingress : delay 150ms 200us loss 1 GRO disabled to force stress and SACK storms. for f in `seq 1 10` do ./netperf -H lpaa6 -l30 -- -K bbr -o THROUGHPUT|tail -1 done | awk '{print $0} {sum += $0} END {printf "%7u\n",sum}' Before patch : 323.87 351.48 339.59 338.62 306.72 204.07 304.93 291.88 202.47 176.88 2840 After patch: 1700.83 2207.98 2070.17 1544.26 2114.76 2124.89 1693.14 1080.91 2216.82 1299.94 18053 Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-06 13:21:27 +08:00
tcp_rtx_queue_purge(sk);
INIT_LIST_HEAD(&tcp_sk(sk)->tsorted_sent_queue);
tcp_clear_all_retrans_hints(tcp_sk(sk));
tcp_sk(sk)->packets_out = 0;
inet_csk(sk)->icsk_backoff = 0;
}
int tcp_disconnect(struct sock *sk, int flags)
{
struct inet_sock *inet = inet_sk(sk);
struct inet_connection_sock *icsk = inet_csk(sk);
struct tcp_sock *tp = tcp_sk(sk);
int old_state = sk->sk_state;
u32 seq;
if (old_state != TCP_CLOSE)
tcp_set_state(sk, TCP_CLOSE);
/* ABORT function of RFC793 */
if (old_state == TCP_LISTEN) {
inet_csk_listen_stop(sk);
} else if (unlikely(tp->repair)) {
WRITE_ONCE(sk->sk_err, ECONNABORTED);
} else if (tcp_need_reset(old_state) ||
(tp->snd_nxt != tp->write_seq &&
(1 << old_state) & (TCPF_CLOSING | TCPF_LAST_ACK))) {
/* The last check adjusts for discrepancy of Linux wrt. RFC
* states
*/
tcp_send_active_reset(sk, gfp_any());
WRITE_ONCE(sk->sk_err, ECONNRESET);
Revert "tcp: Reset tcp connections in SYN-SENT state" This reverts commit e880f8b3a24a73704731a7227ed5fee14bd90192. 1) Patch has not been properly tested, and is wrong [1] 2) Patch submission did not include TCP maintainer (this is me) [1] divide error: 0000 [#1] PREEMPT SMP KASAN CPU: 0 PID: 8426 Comm: syz-executor478 Not tainted 5.12.0-rc4-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 RIP: 0010:__tcp_select_window+0x56d/0xad0 net/ipv4/tcp_output.c:3015 Code: 44 89 ff e8 d5 cd f0 f9 45 39 e7 0f 8d 20 ff ff ff e8 f7 c7 f0 f9 44 89 e3 e9 13 ff ff ff e8 ea c7 f0 f9 44 89 e0 44 89 e3 99 <f7> 7c 24 04 29 d3 e9 fc fe ff ff e8 d3 c7 f0 f9 41 f7 dc bf 1f 00 RSP: 0018:ffffc9000184fac0 EFLAGS: 00010293 RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffffffff87832e76 RDI: 0000000000000003 RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 R10: ffffffff87832e14 R11: 0000000000000000 R12: 0000000000000000 R13: 1ffff92000309f5c R14: 0000000000000000 R15: 0000000000000000 FS: 00000000023eb300(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fc2b5f426c0 CR3: 000000001c5cf000 CR4: 00000000001506f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: tcp_select_window net/ipv4/tcp_output.c:264 [inline] __tcp_transmit_skb+0xa82/0x38f0 net/ipv4/tcp_output.c:1351 tcp_transmit_skb net/ipv4/tcp_output.c:1423 [inline] tcp_send_active_reset+0x475/0x8e0 net/ipv4/tcp_output.c:3449 tcp_disconnect+0x15a9/0x1e60 net/ipv4/tcp.c:2955 inet_shutdown+0x260/0x430 net/ipv4/af_inet.c:905 __sys_shutdown_sock net/socket.c:2189 [inline] __sys_shutdown_sock net/socket.c:2183 [inline] __sys_shutdown+0xf1/0x1b0 net/socket.c:2201 __do_sys_shutdown net/socket.c:2209 [inline] __se_sys_shutdown net/socket.c:2207 [inline] __x64_sys_shutdown+0x50/0x70 net/socket.c:2207 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46 entry_SYSCALL_64_after_hwframe+0x44/0xae Fixes: e880f8b3a24a ("tcp: Reset tcp connections in SYN-SENT state") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Cc: Manoj Basapathi <manojbm@codeaurora.org> Cc: Sauvik Saha <ssaha@codeaurora.org> Link: https://lore.kernel.org/r/20210409170237.274904-1-eric.dumazet@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-04-10 01:02:37 +08:00
} else if (old_state == TCP_SYN_SENT)
WRITE_ONCE(sk->sk_err, ECONNRESET);
tcp_clear_xmit_timers(sk);
__skb_queue_purge(&sk->sk_receive_queue);
WRITE_ONCE(tp->copied_seq, tp->rcv_nxt);
WRITE_ONCE(tp->urg_data, 0);
tcp_write_queue_purge(sk);
net/tcp_fastopen: Disable active side TFO in certain scenarios Middlebox firewall issues can potentially cause server's data being blackholed after a successful 3WHS using TFO. Following are the related reports from Apple: https://www.nanog.org/sites/default/files/Paasch_Network_Support.pdf Slide 31 identifies an issue where the client ACK to the server's data sent during a TFO'd handshake is dropped. C ---> syn-data ---> S C <--- syn/ack ----- S C (accept & write) C <---- data ------- S C ----- ACK -> X S [retry and timeout] https://www.ietf.org/proceedings/94/slides/slides-94-tcpm-13.pdf Slide 5 shows a similar situation that the server's data gets dropped after 3WHS. C ---- syn-data ---> S C <--- syn/ack ----- S C ---- ack --------> S S (accept & write) C? X <- data ------ S [retry and timeout] This is the worst failure b/c the client can not detect such behavior to mitigate the situation (such as disabling TFO). Failing to proceed, the application (e.g., SSL library) may simply timeout and retry with TFO again, and the process repeats indefinitely. The proposed solution is to disable active TFO globally under the following circumstances: 1. client side TFO socket detects out of order FIN 2. client side TFO socket receives out of order RST We disable active side TFO globally for 1hr at first. Then if it happens again, we disable it for 2h, then 4h, 8h, ... And we reset the timeout to 1hr if a client side TFO sockets not opened on loopback has successfully received data segs from server. And we examine this condition during close(). The rational behind it is that when such firewall issue happens, application running on the client should eventually close the socket as it is not able to get the data it is expecting. Or application running on the server should close the socket as it is not able to receive any response from client. In both cases, out of order FIN or RST will get received on the client given that the firewall will not block them as no data are in those frames. And we want to disable active TFO globally as it helps if the middle box is very close to the client and most of the connections are likely to fail. Also, add a debug sysctl: tcp_fastopen_blackhole_detect_timeout_sec: the initial timeout to use when firewall blackhole issue happens. This can be set and read. When setting it to 0, it means to disable the active disable logic. Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-04-21 05:45:46 +08:00
tcp_fastopen_active_disable_ofo_check(sk);
2016-09-08 05:49:28 +08:00
skb_rbtree_purge(&tp->out_of_order_queue);
inet->inet_dport = 0;
dccp/tcp: Fixup bhash2 bucket when connect() fails. If a socket bound to a wildcard address fails to connect(), we only reset saddr and keep the port. Then, we have to fix up the bhash2 bucket; otherwise, the bucket has an inconsistent address in the list. Also, listen() for such a socket will fire the WARN_ON() in inet_csk_get_port(). [0] Note that when a system runs out of memory, we give up fixing the bucket and unlink sk from bhash and bhash2 by inet_put_port(). [0]: WARNING: CPU: 0 PID: 207 at net/ipv4/inet_connection_sock.c:548 inet_csk_get_port (net/ipv4/inet_connection_sock.c:548 (discriminator 1)) Modules linked in: CPU: 0 PID: 207 Comm: bhash2_prev_rep Not tainted 6.1.0-rc3-00799-gc8421681c845 #63 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.0-1.amzn2022.0.1 04/01/2014 RIP: 0010:inet_csk_get_port (net/ipv4/inet_connection_sock.c:548 (discriminator 1)) Code: 74 a7 eb 93 48 8b 54 24 18 0f b7 cb 4c 89 e6 4c 89 ff e8 48 b2 ff ff 49 8b 87 18 04 00 00 e9 32 ff ff ff 0f 0b e9 34 ff ff ff <0f> 0b e9 42 ff ff ff 41 8b 7f 50 41 8b 4f 54 89 fe 81 f6 00 00 ff RSP: 0018:ffffc900003d7e50 EFLAGS: 00010202 RAX: ffff8881047fb500 RBX: 0000000000004e20 RCX: 0000000000000000 RDX: 000000000000000a RSI: 00000000fffffe00 RDI: 00000000ffffffff RBP: ffffffff8324dc00 R08: 0000000000000001 R09: 0000000000000001 R10: 0000000000000001 R11: 0000000000000001 R12: 0000000000000000 R13: 0000000000000001 R14: 0000000000004e20 R15: ffff8881054e1280 FS: 00007f8ac04dc740(0000) GS:ffff88842fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000020001540 CR3: 00000001055fa003 CR4: 0000000000770ef0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 PKRU: 55555554 Call Trace: <TASK> inet_csk_listen_start (net/ipv4/inet_connection_sock.c:1205) inet_listen (net/ipv4/af_inet.c:228) __sys_listen (net/socket.c:1810) __x64_sys_listen (net/socket.c:1819 net/socket.c:1817 net/socket.c:1817) do_syscall_64 (arch/x86/entry/common.c:50 arch/x86/entry/common.c:80) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:120) RIP: 0033:0x7f8ac051de5d Code: ff c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 93 af 1b 00 f7 d8 64 89 01 48 RSP: 002b:00007ffc1c177248 EFLAGS: 00000206 ORIG_RAX: 0000000000000032 RAX: ffffffffffffffda RBX: 0000000020001550 RCX: 00007f8ac051de5d RDX: ffffffffffffff80 RSI: 0000000000000000 RDI: 0000000000000004 RBP: 00007ffc1c177270 R08: 0000000000000018 R09: 0000000000000007 R10: 0000000020001540 R11: 0000000000000206 R12: 00007ffc1c177388 R13: 0000000000401169 R14: 0000000000403e18 R15: 00007f8ac0723000 </TASK> Fixes: 28044fc1d495 ("net: Add a bhash2 table hashed by port and address") Reported-by: syzbot <syzkaller@googlegroups.com> Reported-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Acked-by: Joanne Koong <joannelkoong@gmail.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2022-11-19 09:49:14 +08:00
inet_bhash2_reset_saddr(sk);
tcp: add annotations around sk->sk_shutdown accesses Now sk->sk_shutdown is no longer a bitfield, we can add standard READ_ONCE()/WRITE_ONCE() annotations to silence KCSAN reports like the following: BUG: KCSAN: data-race in tcp_disconnect / tcp_poll write to 0xffff88814588582c of 1 bytes by task 3404 on cpu 1: tcp_disconnect+0x4d6/0xdb0 net/ipv4/tcp.c:3121 __inet_stream_connect+0x5dd/0x6e0 net/ipv4/af_inet.c:715 inet_stream_connect+0x48/0x70 net/ipv4/af_inet.c:727 __sys_connect_file net/socket.c:2001 [inline] __sys_connect+0x19b/0x1b0 net/socket.c:2018 __do_sys_connect net/socket.c:2028 [inline] __se_sys_connect net/socket.c:2025 [inline] __x64_sys_connect+0x41/0x50 net/socket.c:2025 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd read to 0xffff88814588582c of 1 bytes by task 3374 on cpu 0: tcp_poll+0x2e6/0x7d0 net/ipv4/tcp.c:562 sock_poll+0x253/0x270 net/socket.c:1383 vfs_poll include/linux/poll.h:88 [inline] io_poll_check_events io_uring/poll.c:281 [inline] io_poll_task_func+0x15a/0x820 io_uring/poll.c:333 handle_tw_list io_uring/io_uring.c:1184 [inline] tctx_task_work+0x1fe/0x4d0 io_uring/io_uring.c:1246 task_work_run+0x123/0x160 kernel/task_work.c:179 get_signal+0xe64/0xff0 kernel/signal.c:2635 arch_do_signal_or_restart+0x89/0x2a0 arch/x86/kernel/signal.c:306 exit_to_user_mode_loop+0x6f/0xe0 kernel/entry/common.c:168 exit_to_user_mode_prepare+0x6c/0xb0 kernel/entry/common.c:204 __syscall_exit_to_user_mode_work kernel/entry/common.c:286 [inline] syscall_exit_to_user_mode+0x26/0x140 kernel/entry/common.c:297 do_syscall_64+0x4d/0xc0 arch/x86/entry/common.c:86 entry_SYSCALL_64_after_hwframe+0x63/0xcd value changed: 0x03 -> 0x00 Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-05-10 04:36:56 +08:00
WRITE_ONCE(sk->sk_shutdown, 0);
sock_reset_flag(sk, SOCK_DONE);
tcp: switch rtt estimations to usec resolution Upcoming congestion controls for TCP require usec resolution for RTT estimations. Millisecond resolution is simply not enough these days. FQ/pacing in DC environments also require this change for finer control and removal of bimodal behavior due to the current hack in tcp_update_pacing_rate() for 'small rtt' TCP_CONG_RTT_STAMP is no longer needed. As Julian Anastasov pointed out, we need to keep user compatibility : tcp_metrics used to export RTT and RTTVAR in msec resolution, so we added RTT_US and RTTVAR_US. An iproute2 patch is needed to use the new attributes if provided by the kernel. In this example ss command displays a srtt of 32 usecs (10Gbit link) lpk51:~# ./ss -i dst lpk52 Netid State Recv-Q Send-Q Local Address:Port Peer Address:Port tcp ESTAB 0 1 10.246.11.51:42959 10.246.11.52:64614 cubic wscale:6,6 rto:201 rtt:0.032/0.001 ato:40 mss:1448 cwnd:10 send 3620.0Mbps pacing_rate 7240.0Mbps unacked:1 rcv_rtt:993 rcv_space:29559 Updated iproute2 ip command displays : lpk51:~# ./ip tcp_metrics | grep 10.246.11.52 10.246.11.52 age 561.914sec cwnd 10 rtt 274us rttvar 213us source 10.246.11.51 Old binary displays : lpk51:~# ip tcp_metrics | grep 10.246.11.52 10.246.11.52 age 561.914sec cwnd 10 rtt 250us rttvar 125us source 10.246.11.51 With help from Julian Anastasov, Stephen Hemminger and Yuchung Cheng Signed-off-by: Eric Dumazet <edumazet@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Cc: Stephen Hemminger <stephen@networkplumber.org> Cc: Yuchung Cheng <ycheng@google.com> Cc: Larry Brakmo <brakmo@google.com> Cc: Julian Anastasov <ja@ssi.bg> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-27 06:02:48 +08:00
tp->srtt_us = 0;
tp->mdev_us = jiffies_to_usecs(TCP_TIMEOUT_INIT);
tp->rcv_rtt_last_tsecr = 0;
seq = tp->write_seq + tp->max_window + 2;
if (!seq)
seq = 1;
WRITE_ONCE(tp->write_seq, seq);
icsk->icsk_backoff = 0;
icsk->icsk_probes_out = 0;
icsk->icsk_probes_tstamp = 0;
icsk->icsk_rto = TCP_TIMEOUT_INIT;
icsk->icsk_rto_min = TCP_RTO_MIN;
icsk->icsk_delack_max = TCP_DELACK_MAX;
tp->snd_ssthresh = TCP_INFINITE_SSTHRESH;
tcp_snd_cwnd_set(tp, TCP_INIT_CWND);
tp->snd_cwnd_cnt = 0;
tcp: fix tcp_cwnd_validate() to not forget is_cwnd_limited This commit fixes a bug in the tracking of max_packets_out and is_cwnd_limited. This bug can cause the connection to fail to remember that is_cwnd_limited is true, causing the connection to fail to grow cwnd when it should, causing throughput to be lower than it should be. The following event sequence is an example that triggers the bug: (a) The connection is cwnd_limited, but packets_out is not at its peak due to TSO deferral deciding not to send another skb yet. In such cases the connection can advance max_packets_seq and set tp->is_cwnd_limited to true and max_packets_out to a small number. (b) Then later in the round trip the connection is pacing-limited (not cwnd-limited), and packets_out is larger. In such cases the connection would raise max_packets_out to a bigger number but (unexpectedly) flip tp->is_cwnd_limited from true to false. This commit fixes that bug. One straightforward fix would be to separately track (a) the next window after max_packets_out reaches a maximum, and (b) the next window after tp->is_cwnd_limited is set to true. But this would require consuming an extra u32 sequence number. Instead, to save space we track only the most important information. Specifically, we track the strongest available signal of the degree to which the cwnd is fully utilized: (1) If the connection is cwnd-limited then we remember that fact for the current window. (2) If the connection not cwnd-limited then we track the maximum number of outstanding packets in the current window. In particular, note that the new logic cannot trigger the buggy (a)/(b) sequence above because with the new logic a condition where tp->packets_out > tp->max_packets_out can only trigger an update of tp->is_cwnd_limited if tp->is_cwnd_limited is false. This first showed up in a testing of a BBRv2 dev branch, but this buggy behavior highlighted a general issue with the tcp_cwnd_validate() logic that can cause cwnd to fail to increase at the proper rate for any TCP congestion control, including Reno or CUBIC. Fixes: ca8a22634381 ("tcp: make cwnd-limited checks measurement-based, and gentler") Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: Kevin(Yudong) Yang <yyd@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2022-09-29 04:03:31 +08:00
tp->is_cwnd_limited = 0;
tp->max_packets_out = 0;
tp->window_clamp = 0;
tp->delivered = 0;
tp->delivered_ce = 0;
tcp: make sure listeners don't initialize congestion-control state syzkaller found its way into setsockopt with TCP_CONGESTION "cdg". tcp_cdg_init() does a kcalloc to store the gradients. As sk_clone_lock just copies all the memory, the allocated pointer will be copied as well, if the app called setsockopt(..., TCP_CONGESTION) on the listener. If now the socket will be destroyed before the congestion-control has properly been initialized (through a call to tcp_init_transfer), we will end up freeing memory that does not belong to that particular socket, opening the door to a double-free: [ 11.413102] ================================================================== [ 11.414181] BUG: KASAN: double-free or invalid-free in tcp_cleanup_congestion_control+0x58/0xd0 [ 11.415329] [ 11.415560] CPU: 3 PID: 4884 Comm: syz-executor.5 Not tainted 5.8.0-rc2 #80 [ 11.416544] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.1-0-ga5cab58e9a3f-prebuilt.qemu.org 04/01/2014 [ 11.418148] Call Trace: [ 11.418534] <IRQ> [ 11.418834] dump_stack+0x7d/0xb0 [ 11.419297] print_address_description.constprop.0+0x1a/0x210 [ 11.422079] kasan_report_invalid_free+0x51/0x80 [ 11.423433] __kasan_slab_free+0x15e/0x170 [ 11.424761] kfree+0x8c/0x230 [ 11.425157] tcp_cleanup_congestion_control+0x58/0xd0 [ 11.425872] tcp_v4_destroy_sock+0x57/0x5a0 [ 11.426493] inet_csk_destroy_sock+0x153/0x2c0 [ 11.427093] tcp_v4_syn_recv_sock+0xb29/0x1100 [ 11.427731] tcp_get_cookie_sock+0xc3/0x4a0 [ 11.429457] cookie_v4_check+0x13d0/0x2500 [ 11.433189] tcp_v4_do_rcv+0x60e/0x780 [ 11.433727] tcp_v4_rcv+0x2869/0x2e10 [ 11.437143] ip_protocol_deliver_rcu+0x23/0x190 [ 11.437810] ip_local_deliver+0x294/0x350 [ 11.439566] __netif_receive_skb_one_core+0x15d/0x1a0 [ 11.441995] process_backlog+0x1b1/0x6b0 [ 11.443148] net_rx_action+0x37e/0xc40 [ 11.445361] __do_softirq+0x18c/0x61a [ 11.445881] asm_call_on_stack+0x12/0x20 [ 11.446409] </IRQ> [ 11.446716] do_softirq_own_stack+0x34/0x40 [ 11.447259] do_softirq.part.0+0x26/0x30 [ 11.447827] __local_bh_enable_ip+0x46/0x50 [ 11.448406] ip_finish_output2+0x60f/0x1bc0 [ 11.450109] __ip_queue_xmit+0x71c/0x1b60 [ 11.451861] __tcp_transmit_skb+0x1727/0x3bb0 [ 11.453789] tcp_rcv_state_process+0x3070/0x4d3a [ 11.456810] tcp_v4_do_rcv+0x2ad/0x780 [ 11.457995] __release_sock+0x14b/0x2c0 [ 11.458529] release_sock+0x4a/0x170 [ 11.459005] __inet_stream_connect+0x467/0xc80 [ 11.461435] inet_stream_connect+0x4e/0xa0 [ 11.462043] __sys_connect+0x204/0x270 [ 11.465515] __x64_sys_connect+0x6a/0xb0 [ 11.466088] do_syscall_64+0x3e/0x70 [ 11.466617] entry_SYSCALL_64_after_hwframe+0x44/0xa9 [ 11.467341] RIP: 0033:0x7f56046dc469 [ 11.467844] Code: Bad RIP value. [ 11.468282] RSP: 002b:00007f5604dccdd8 EFLAGS: 00000246 ORIG_RAX: 000000000000002a [ 11.469326] RAX: ffffffffffffffda RBX: 000000000068bf00 RCX: 00007f56046dc469 [ 11.470379] RDX: 0000000000000010 RSI: 0000000020000000 RDI: 0000000000000004 [ 11.471311] RBP: 00000000ffffffff R08: 0000000000000000 R09: 0000000000000000 [ 11.472286] R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000 [ 11.473341] R13: 000000000041427c R14: 00007f5604dcd5c0 R15: 0000000000000003 [ 11.474321] [ 11.474527] Allocated by task 4884: [ 11.475031] save_stack+0x1b/0x40 [ 11.475548] __kasan_kmalloc.constprop.0+0xc2/0xd0 [ 11.476182] tcp_cdg_init+0xf0/0x150 [ 11.476744] tcp_init_congestion_control+0x9b/0x3a0 [ 11.477435] tcp_set_congestion_control+0x270/0x32f [ 11.478088] do_tcp_setsockopt.isra.0+0x521/0x1a00 [ 11.478744] __sys_setsockopt+0xff/0x1e0 [ 11.479259] __x64_sys_setsockopt+0xb5/0x150 [ 11.479895] do_syscall_64+0x3e/0x70 [ 11.480395] entry_SYSCALL_64_after_hwframe+0x44/0xa9 [ 11.481097] [ 11.481321] Freed by task 4872: [ 11.481783] save_stack+0x1b/0x40 [ 11.482230] __kasan_slab_free+0x12c/0x170 [ 11.482839] kfree+0x8c/0x230 [ 11.483240] tcp_cleanup_congestion_control+0x58/0xd0 [ 11.483948] tcp_v4_destroy_sock+0x57/0x5a0 [ 11.484502] inet_csk_destroy_sock+0x153/0x2c0 [ 11.485144] tcp_close+0x932/0xfe0 [ 11.485642] inet_release+0xc1/0x1c0 [ 11.486131] __sock_release+0xc0/0x270 [ 11.486697] sock_close+0xc/0x10 [ 11.487145] __fput+0x277/0x780 [ 11.487632] task_work_run+0xeb/0x180 [ 11.488118] __prepare_exit_to_usermode+0x15a/0x160 [ 11.488834] do_syscall_64+0x4a/0x70 [ 11.489326] entry_SYSCALL_64_after_hwframe+0x44/0xa9 Wei Wang fixed a part of these CDG-malloc issues with commit c12014440750 ("tcp: memset ca_priv data to 0 properly"). This patch here fixes the listener-scenario: We make sure that listeners setting the congestion-control through setsockopt won't initialize it (thus CDG never allocates on listeners). For those who use AF_UNSPEC to reuse a socket, tcp_disconnect() is changed to cleanup afterwards. (The issue can be reproduced at least down to v4.4.x.) Cc: Wei Wang <weiwan@google.com> Cc: Eric Dumazet <edumazet@google.com> Fixes: 2b0a8c9eee81 ("tcp: add CDG congestion control") Signed-off-by: Christoph Paasch <cpaasch@apple.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2020-07-09 07:18:34 +08:00
if (icsk->icsk_ca_ops->release)
icsk->icsk_ca_ops->release(sk);
memset(icsk->icsk_ca_priv, 0, sizeof(icsk->icsk_ca_priv));
icsk->icsk_ca_initialized = 0;
tcp_set_ca_state(sk, TCP_CA_Open);
tp->is_sack_reneg = 0;
tcp_clear_retrans(tp);
tp->total_retrans = 0;
inet_csk_delack_init(sk);
/* Initialize rcv_mss to TCP_MIN_MSS to avoid division by 0
* issue in __tcp_select_window()
*/
icsk->icsk_ack.rcv_mss = TCP_MIN_MSS;
memset(&tp->rx_opt, 0, sizeof(tp->rx_opt));
__sk_dst_reset(sk);
inet: fully convert sk->sk_rx_dst to RCU rules syzbot reported various issues around early demux, one being included in this changelog [1] sk->sk_rx_dst is using RCU protection without clearly documenting it. And following sequences in tcp_v4_do_rcv()/tcp_v6_do_rcv() are not following standard RCU rules. [a] dst_release(dst); [b] sk->sk_rx_dst = NULL; They look wrong because a delete operation of RCU protected pointer is supposed to clear the pointer before the call_rcu()/synchronize_rcu() guarding actual memory freeing. In some cases indeed, dst could be freed before [b] is done. We could cheat by clearing sk_rx_dst before calling dst_release(), but this seems the right time to stick to standard RCU annotations and debugging facilities. [1] BUG: KASAN: use-after-free in dst_check include/net/dst.h:470 [inline] BUG: KASAN: use-after-free in tcp_v4_early_demux+0x95b/0x960 net/ipv4/tcp_ipv4.c:1792 Read of size 2 at addr ffff88807f1cb73a by task syz-executor.5/9204 CPU: 0 PID: 9204 Comm: syz-executor.5 Not tainted 5.16.0-rc5-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: <TASK> __dump_stack lib/dump_stack.c:88 [inline] dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106 print_address_description.constprop.0.cold+0x8d/0x320 mm/kasan/report.c:247 __kasan_report mm/kasan/report.c:433 [inline] kasan_report.cold+0x83/0xdf mm/kasan/report.c:450 dst_check include/net/dst.h:470 [inline] tcp_v4_early_demux+0x95b/0x960 net/ipv4/tcp_ipv4.c:1792 ip_rcv_finish_core.constprop.0+0x15de/0x1e80 net/ipv4/ip_input.c:340 ip_list_rcv_finish.constprop.0+0x1b2/0x6e0 net/ipv4/ip_input.c:583 ip_sublist_rcv net/ipv4/ip_input.c:609 [inline] ip_list_rcv+0x34e/0x490 net/ipv4/ip_input.c:644 __netif_receive_skb_list_ptype net/core/dev.c:5508 [inline] __netif_receive_skb_list_core+0x549/0x8e0 net/core/dev.c:5556 __netif_receive_skb_list net/core/dev.c:5608 [inline] netif_receive_skb_list_internal+0x75e/0xd80 net/core/dev.c:5699 gro_normal_list net/core/dev.c:5853 [inline] gro_normal_list net/core/dev.c:5849 [inline] napi_complete_done+0x1f1/0x880 net/core/dev.c:6590 virtqueue_napi_complete drivers/net/virtio_net.c:339 [inline] virtnet_poll+0xca2/0x11b0 drivers/net/virtio_net.c:1557 __napi_poll+0xaf/0x440 net/core/dev.c:7023 napi_poll net/core/dev.c:7090 [inline] net_rx_action+0x801/0xb40 net/core/dev.c:7177 __do_softirq+0x29b/0x9c2 kernel/softirq.c:558 invoke_softirq kernel/softirq.c:432 [inline] __irq_exit_rcu+0x123/0x180 kernel/softirq.c:637 irq_exit_rcu+0x5/0x20 kernel/softirq.c:649 common_interrupt+0x52/0xc0 arch/x86/kernel/irq.c:240 asm_common_interrupt+0x1e/0x40 arch/x86/include/asm/idtentry.h:629 RIP: 0033:0x7f5e972bfd57 Code: 39 d1 73 14 0f 1f 80 00 00 00 00 48 8b 50 f8 48 83 e8 08 48 39 ca 77 f3 48 39 c3 73 3e 48 89 13 48 8b 50 f8 48 89 38 49 8b 0e <48> 8b 3e 48 83 c3 08 48 83 c6 08 eb bc 48 39 d1 72 9e 48 39 d0 73 RSP: 002b:00007fff8a413210 EFLAGS: 00000283 RAX: 00007f5e97108990 RBX: 00007f5e97108338 RCX: ffffffff81d3aa45 RDX: ffffffff81d3aa45 RSI: 00007f5e97108340 RDI: ffffffff81d3aa45 RBP: 00007f5e97107eb8 R08: 00007f5e97108d88 R09: 0000000093c2e8d9 R10: 0000000000000000 R11: 0000000000000000 R12: 00007f5e97107eb0 R13: 00007f5e97108338 R14: 00007f5e97107ea8 R15: 0000000000000019 </TASK> Allocated by task 13: kasan_save_stack+0x1e/0x50 mm/kasan/common.c:38 kasan_set_track mm/kasan/common.c:46 [inline] set_alloc_info mm/kasan/common.c:434 [inline] __kasan_slab_alloc+0x90/0xc0 mm/kasan/common.c:467 kasan_slab_alloc include/linux/kasan.h:259 [inline] slab_post_alloc_hook mm/slab.h:519 [inline] slab_alloc_node mm/slub.c:3234 [inline] slab_alloc mm/slub.c:3242 [inline] kmem_cache_alloc+0x202/0x3a0 mm/slub.c:3247 dst_alloc+0x146/0x1f0 net/core/dst.c:92 rt_dst_alloc+0x73/0x430 net/ipv4/route.c:1613 ip_route_input_slow+0x1817/0x3a20 net/ipv4/route.c:2340 ip_route_input_rcu net/ipv4/route.c:2470 [inline] ip_route_input_noref+0x116/0x2a0 net/ipv4/route.c:2415 ip_rcv_finish_core.constprop.0+0x288/0x1e80 net/ipv4/ip_input.c:354 ip_list_rcv_finish.constprop.0+0x1b2/0x6e0 net/ipv4/ip_input.c:583 ip_sublist_rcv net/ipv4/ip_input.c:609 [inline] ip_list_rcv+0x34e/0x490 net/ipv4/ip_input.c:644 __netif_receive_skb_list_ptype net/core/dev.c:5508 [inline] __netif_receive_skb_list_core+0x549/0x8e0 net/core/dev.c:5556 __netif_receive_skb_list net/core/dev.c:5608 [inline] netif_receive_skb_list_internal+0x75e/0xd80 net/core/dev.c:5699 gro_normal_list net/core/dev.c:5853 [inline] gro_normal_list net/core/dev.c:5849 [inline] napi_complete_done+0x1f1/0x880 net/core/dev.c:6590 virtqueue_napi_complete drivers/net/virtio_net.c:339 [inline] virtnet_poll+0xca2/0x11b0 drivers/net/virtio_net.c:1557 __napi_poll+0xaf/0x440 net/core/dev.c:7023 napi_poll net/core/dev.c:7090 [inline] net_rx_action+0x801/0xb40 net/core/dev.c:7177 __do_softirq+0x29b/0x9c2 kernel/softirq.c:558 Freed by task 13: kasan_save_stack+0x1e/0x50 mm/kasan/common.c:38 kasan_set_track+0x21/0x30 mm/kasan/common.c:46 kasan_set_free_info+0x20/0x30 mm/kasan/generic.c:370 ____kasan_slab_free mm/kasan/common.c:366 [inline] ____kasan_slab_free mm/kasan/common.c:328 [inline] __kasan_slab_free+0xff/0x130 mm/kasan/common.c:374 kasan_slab_free include/linux/kasan.h:235 [inline] slab_free_hook mm/slub.c:1723 [inline] slab_free_freelist_hook+0x8b/0x1c0 mm/slub.c:1749 slab_free mm/slub.c:3513 [inline] kmem_cache_free+0xbd/0x5d0 mm/slub.c:3530 dst_destroy+0x2d6/0x3f0 net/core/dst.c:127 rcu_do_batch kernel/rcu/tree.c:2506 [inline] rcu_core+0x7ab/0x1470 kernel/rcu/tree.c:2741 __do_softirq+0x29b/0x9c2 kernel/softirq.c:558 Last potentially related work creation: kasan_save_stack+0x1e/0x50 mm/kasan/common.c:38 __kasan_record_aux_stack+0xf5/0x120 mm/kasan/generic.c:348 __call_rcu kernel/rcu/tree.c:2985 [inline] call_rcu+0xb1/0x740 kernel/rcu/tree.c:3065 dst_release net/core/dst.c:177 [inline] dst_release+0x79/0xe0 net/core/dst.c:167 tcp_v4_do_rcv+0x612/0x8d0 net/ipv4/tcp_ipv4.c:1712 sk_backlog_rcv include/net/sock.h:1030 [inline] __release_sock+0x134/0x3b0 net/core/sock.c:2768 release_sock+0x54/0x1b0 net/core/sock.c:3300 tcp_sendmsg+0x36/0x40 net/ipv4/tcp.c:1441 inet_sendmsg+0x99/0xe0 net/ipv4/af_inet.c:819 sock_sendmsg_nosec net/socket.c:704 [inline] sock_sendmsg+0xcf/0x120 net/socket.c:724 sock_write_iter+0x289/0x3c0 net/socket.c:1057 call_write_iter include/linux/fs.h:2162 [inline] new_sync_write+0x429/0x660 fs/read_write.c:503 vfs_write+0x7cd/0xae0 fs/read_write.c:590 ksys_write+0x1ee/0x250 fs/read_write.c:643 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x44/0xae The buggy address belongs to the object at ffff88807f1cb700 which belongs to the cache ip_dst_cache of size 176 The buggy address is located 58 bytes inside of 176-byte region [ffff88807f1cb700, ffff88807f1cb7b0) The buggy address belongs to the page: page:ffffea0001fc72c0 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x7f1cb flags: 0xfff00000000200(slab|node=0|zone=1|lastcpupid=0x7ff) raw: 00fff00000000200 dead000000000100 dead000000000122 ffff8881413bb780 raw: 0000000000000000 0000000000100010 00000001ffffffff 0000000000000000 page dumped because: kasan: bad access detected page_owner tracks the page as allocated page last allocated via order 0, migratetype Unmovable, gfp_mask 0x112a20(GFP_ATOMIC|__GFP_NOWARN|__GFP_NORETRY|__GFP_HARDWALL), pid 5, ts 108466983062, free_ts 108048976062 prep_new_page mm/page_alloc.c:2418 [inline] get_page_from_freelist+0xa72/0x2f50 mm/page_alloc.c:4149 __alloc_pages+0x1b2/0x500 mm/page_alloc.c:5369 alloc_pages+0x1a7/0x300 mm/mempolicy.c:2191 alloc_slab_page mm/slub.c:1793 [inline] allocate_slab mm/slub.c:1930 [inline] new_slab+0x32d/0x4a0 mm/slub.c:1993 ___slab_alloc+0x918/0xfe0 mm/slub.c:3022 __slab_alloc.constprop.0+0x4d/0xa0 mm/slub.c:3109 slab_alloc_node mm/slub.c:3200 [inline] slab_alloc mm/slub.c:3242 [inline] kmem_cache_alloc+0x35c/0x3a0 mm/slub.c:3247 dst_alloc+0x146/0x1f0 net/core/dst.c:92 rt_dst_alloc+0x73/0x430 net/ipv4/route.c:1613 __mkroute_output net/ipv4/route.c:2564 [inline] ip_route_output_key_hash_rcu+0x921/0x2d00 net/ipv4/route.c:2791 ip_route_output_key_hash+0x18b/0x300 net/ipv4/route.c:2619 __ip_route_output_key include/net/route.h:126 [inline] ip_route_output_flow+0x23/0x150 net/ipv4/route.c:2850 ip_route_output_key include/net/route.h:142 [inline] geneve_get_v4_rt+0x3a6/0x830 drivers/net/geneve.c:809 geneve_xmit_skb drivers/net/geneve.c:899 [inline] geneve_xmit+0xc4a/0x3540 drivers/net/geneve.c:1082 __netdev_start_xmit include/linux/netdevice.h:4994 [inline] netdev_start_xmit include/linux/netdevice.h:5008 [inline] xmit_one net/core/dev.c:3590 [inline] dev_hard_start_xmit+0x1eb/0x920 net/core/dev.c:3606 __dev_queue_xmit+0x299a/0x3650 net/core/dev.c:4229 page last free stack trace: reset_page_owner include/linux/page_owner.h:24 [inline] free_pages_prepare mm/page_alloc.c:1338 [inline] free_pcp_prepare+0x374/0x870 mm/page_alloc.c:1389 free_unref_page_prepare mm/page_alloc.c:3309 [inline] free_unref_page+0x19/0x690 mm/page_alloc.c:3388 qlink_free mm/kasan/quarantine.c:146 [inline] qlist_free_all+0x5a/0xc0 mm/kasan/quarantine.c:165 kasan_quarantine_reduce+0x180/0x200 mm/kasan/quarantine.c:272 __kasan_slab_alloc+0xa2/0xc0 mm/kasan/common.c:444 kasan_slab_alloc include/linux/kasan.h:259 [inline] slab_post_alloc_hook mm/slab.h:519 [inline] slab_alloc_node mm/slub.c:3234 [inline] kmem_cache_alloc_node+0x255/0x3f0 mm/slub.c:3270 __alloc_skb+0x215/0x340 net/core/skbuff.c:414 alloc_skb include/linux/skbuff.h:1126 [inline] alloc_skb_with_frags+0x93/0x620 net/core/skbuff.c:6078 sock_alloc_send_pskb+0x783/0x910 net/core/sock.c:2575 mld_newpack+0x1df/0x770 net/ipv6/mcast.c:1754 add_grhead+0x265/0x330 net/ipv6/mcast.c:1857 add_grec+0x1053/0x14e0 net/ipv6/mcast.c:1995 mld_send_initial_cr.part.0+0xf6/0x230 net/ipv6/mcast.c:2242 mld_send_initial_cr net/ipv6/mcast.c:1232 [inline] mld_dad_work+0x1d3/0x690 net/ipv6/mcast.c:2268 process_one_work+0x9b2/0x1690 kernel/workqueue.c:2298 worker_thread+0x658/0x11f0 kernel/workqueue.c:2445 Memory state around the buggy address: ffff88807f1cb600: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff88807f1cb680: fb fb fb fb fb fb fc fc fc fc fc fc fc fc fc fc >ffff88807f1cb700: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ^ ffff88807f1cb780: fb fb fb fb fb fb fc fc fc fc fc fc fc fc fc fc ffff88807f1cb800: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb Fixes: 41063e9dd119 ("ipv4: Early TCP socket demux.") Signed-off-by: Eric Dumazet <edumazet@google.com> Link: https://lore.kernel.org/r/20211220143330.680945-1-eric.dumazet@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2021-12-20 22:33:30 +08:00
dst_release(xchg((__force struct dst_entry **)&sk->sk_rx_dst, NULL));
tcp_saved_syn_free(tp);
tp->compressed_ack = 0;
tp->segs_in = 0;
tp->segs_out = 0;
tp->bytes_sent = 0;
tp->bytes_acked = 0;
tp->bytes_received = 0;
tp->bytes_retrans = 0;
tp->data_segs_in = 0;
tp->data_segs_out = 0;
tp->duplicate_sack[0].start_seq = 0;
tp->duplicate_sack[0].end_seq = 0;
tp->dsack_dups = 0;
tp->reord_seen = 0;
tp->retrans_out = 0;
tp->sacked_out = 0;
tp->tlp_high_seq = 0;
tp->last_oow_ack_time = 0;
tp->plb_rehash = 0;
/* There's a bubble in the pipe until at least the first ACK. */
tp->app_limited = ~0U;
tp->rate_app_limited = 1;
tp->rack.mstamp = 0;
tp->rack.advanced = 0;
tp->rack.reo_wnd_steps = 1;
tp->rack.last_delivered = 0;
tp->rack.reo_wnd_persist = 0;
tp->rack.dsack_seen = 0;
tp->syn_data_acked = 0;
tp->rx_opt.saw_tstamp = 0;
tp->rx_opt.dsack = 0;
tp->rx_opt.num_sacks = 0;
tp->rcv_ooopack = 0;
/* Clean up fastopen related fields */
tcp_free_fastopen_req(tp);
inet_clear_bit(DEFER_CONNECT, sk);
tcp: add TCP_INFO status for failed client TFO The TCPI_OPT_SYN_DATA bit as part of tcpi_options currently reports whether or not data-in-SYN was ack'd on both the client and server side. We'd like to gather more information on the client-side in the failure case in order to indicate the reason for the failure. This can be useful for not only debugging TFO, but also for creating TFO socket policies. For example, if a middle box removes the TFO option or drops a data-in-SYN, we can can detect this case, and turn off TFO for these connections saving the extra retransmits. The newly added tcpi_fastopen_client_fail status is 2 bits and has the following 4 states: 1) TFO_STATUS_UNSPEC Catch-all state which includes when TFO is disabled via black hole detection, which is indicated via LINUX_MIB_TCPFASTOPENBLACKHOLE. 2) TFO_COOKIE_UNAVAILABLE If TFO_CLIENT_NO_COOKIE mode is off, this state indicates that no cookie is available in the cache. 3) TFO_DATA_NOT_ACKED Data was sent with SYN, we received a SYN/ACK but it did not cover the data portion. Cookie is not accepted by server because the cookie may be invalid or the server may be overloaded. 4) TFO_SYN_RETRANSMITTED Data was sent with SYN, we received a SYN/ACK which did not cover the data after at least 1 additional SYN was sent (without data). It may be the case that a middle-box is dropping data-in-SYN packets. Thus, it would be more efficient to not use TFO on this connection to avoid extra retransmits during connection establishment. These new fields do not cover all the cases where TFO may fail, but other failures, such as SYN/ACK + data being dropped, will result in the connection not becoming established. And a connection blackhole after session establishment shows up as a stalled connection. Signed-off-by: Jason Baron <jbaron@akamai.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Christoph Paasch <cpaasch@apple.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2019-10-23 23:09:26 +08:00
tp->fastopen_client_fail = 0;
WARN_ON(inet->inet_num && !icsk->icsk_bind_hash);
if (sk->sk_frag.page) {
put_page(sk->sk_frag.page);
sk->sk_frag.page = NULL;
sk->sk_frag.offset = 0;
}
sk_error_report(sk);
return 0;
}
EXPORT_SYMBOL(tcp_disconnect);
static inline bool tcp_can_repair_sock(const struct sock *sk)
{
return sockopt_ns_capable(sock_net(sk)->user_ns, CAP_NET_ADMIN) &&
(sk->sk_state != TCP_LISTEN);
}
static int tcp_repair_set_window(struct tcp_sock *tp, sockptr_t optbuf, int len)
{
struct tcp_repair_window opt;
if (!tp->repair)
return -EPERM;
if (len != sizeof(opt))
return -EINVAL;
if (copy_from_sockptr(&opt, optbuf, sizeof(opt)))
return -EFAULT;
if (opt.max_window < opt.snd_wnd)
return -EINVAL;
if (after(opt.snd_wl1, tp->rcv_nxt + opt.rcv_wnd))
return -EINVAL;
if (after(opt.rcv_wup, tp->rcv_nxt))
return -EINVAL;
tp->snd_wl1 = opt.snd_wl1;
tp->snd_wnd = opt.snd_wnd;
tp->max_window = opt.max_window;
tp->rcv_wnd = opt.rcv_wnd;
tp->rcv_wup = opt.rcv_wup;
return 0;
}
static int tcp_repair_options_est(struct sock *sk, sockptr_t optbuf,
unsigned int len)
{
struct tcp_sock *tp = tcp_sk(sk);
struct tcp_repair_opt opt;
size_t offset = 0;
while (len >= sizeof(opt)) {
if (copy_from_sockptr_offset(&opt, optbuf, offset, sizeof(opt)))
return -EFAULT;
offset += sizeof(opt);
len -= sizeof(opt);
switch (opt.opt_code) {
case TCPOPT_MSS:
tp->rx_opt.mss_clamp = opt.opt_val;
tcp_mtup_init(sk);
break;
case TCPOPT_WINDOW:
{
u16 snd_wscale = opt.opt_val & 0xFFFF;
u16 rcv_wscale = opt.opt_val >> 16;
if (snd_wscale > TCP_MAX_WSCALE || rcv_wscale > TCP_MAX_WSCALE)
return -EFBIG;
tp->rx_opt.snd_wscale = snd_wscale;
tp->rx_opt.rcv_wscale = rcv_wscale;
tp->rx_opt.wscale_ok = 1;
}
break;
case TCPOPT_SACK_PERM:
if (opt.opt_val != 0)
return -EINVAL;
tp->rx_opt.sack_ok |= TCP_SACK_SEEN;
break;
case TCPOPT_TIMESTAMP:
if (opt.opt_val != 0)
return -EINVAL;
tp->rx_opt.tstamp_ok = 1;
break;
}
}
return 0;
}
tcp: add optional per socket transmit delay Adding delays to TCP flows is crucial for studying behavior of TCP stacks, including congestion control modules. Linux offers netem module, but it has unpractical constraints : - Need root access to change qdisc - Hard to setup on egress if combined with non trivial qdisc like FQ - Single delay for all flows. EDT (Earliest Departure Time) adoption in TCP stack allows us to enable a per socket delay at a very small cost. Networking tools can now establish thousands of flows, each of them with a different delay, simulating real world conditions. This requires FQ packet scheduler or a EDT-enabled NIC. This patchs adds TCP_TX_DELAY socket option, to set a delay in usec units. unsigned int tx_delay = 10000; /* 10 msec */ setsockopt(fd, SOL_TCP, TCP_TX_DELAY, &tx_delay, sizeof(tx_delay)); Note that FQ packet scheduler limits might need some tweaking : man tc-fq PARAMETERS limit Hard limit on the real queue size. When this limit is reached, new packets are dropped. If the value is lowered, packets are dropped so that the new limit is met. Default is 10000 packets. flow_limit Hard limit on the maximum number of packets queued per flow. Default value is 100. Use of TCP_TX_DELAY option will increase number of skbs in FQ qdisc, so packets would be dropped if any of the previous limit is hit. Use of a jump label makes this support runtime-free, for hosts never using the option. Also note that TSQ (TCP Small Queues) limits are slightly changed with this patch : we need to account that skbs artificially delayed wont stop us providind more skbs to feed the pipe (netem uses skb_orphan_partial() for this purpose, but FQ can not use this trick) Because of that, using big delays might very well trigger old bugs in TSO auto defer logic and/or sndbuf limited detection. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2019-06-13 02:57:25 +08:00
DEFINE_STATIC_KEY_FALSE(tcp_tx_delay_enabled);
EXPORT_SYMBOL(tcp_tx_delay_enabled);
static void tcp_enable_tx_delay(void)
{
if (!static_branch_unlikely(&tcp_tx_delay_enabled)) {
static int __tcp_tx_delay_enabled = 0;
if (cmpxchg(&__tcp_tx_delay_enabled, 0, 1) == 0) {
static_branch_enable(&tcp_tx_delay_enabled);
pr_info("TCP_TX_DELAY enabled\n");
}
}
}
/* When set indicates to always queue non-full frames. Later the user clears
* this option and we transmit any pending partial frames in the queue. This is
* meant to be used alongside sendfile() to get properly filled frames when the
* user (for example) must write out headers with a write() call first and then
* use sendfile to send out the data parts.
*
* TCP_CORK can be set together with TCP_NODELAY and it is stronger than
* TCP_NODELAY.
*/
void __tcp_sock_set_cork(struct sock *sk, bool on)
{
struct tcp_sock *tp = tcp_sk(sk);
if (on) {
tp->nonagle |= TCP_NAGLE_CORK;
} else {
tp->nonagle &= ~TCP_NAGLE_CORK;
if (tp->nonagle & TCP_NAGLE_OFF)
tp->nonagle |= TCP_NAGLE_PUSH;
tcp_push_pending_frames(sk);
}
}
void tcp_sock_set_cork(struct sock *sk, bool on)
{
lock_sock(sk);
__tcp_sock_set_cork(sk, on);
release_sock(sk);
}
EXPORT_SYMBOL(tcp_sock_set_cork);
/* TCP_NODELAY is weaker than TCP_CORK, so that this option on corked socket is
* remembered, but it is not activated until cork is cleared.
*
* However, when TCP_NODELAY is set we make an explicit push, which overrides
* even TCP_CORK for currently queued segments.
*/
void __tcp_sock_set_nodelay(struct sock *sk, bool on)
{
if (on) {
tcp_sk(sk)->nonagle |= TCP_NAGLE_OFF|TCP_NAGLE_PUSH;
tcp_push_pending_frames(sk);
} else {
tcp_sk(sk)->nonagle &= ~TCP_NAGLE_OFF;
}
}
void tcp_sock_set_nodelay(struct sock *sk)
{
lock_sock(sk);
__tcp_sock_set_nodelay(sk, true);
release_sock(sk);
}
EXPORT_SYMBOL(tcp_sock_set_nodelay);
static void __tcp_sock_set_quickack(struct sock *sk, int val)
{
if (!val) {
inet_csk_enter_pingpong_mode(sk);
return;
}
inet_csk_exit_pingpong_mode(sk);
if ((1 << sk->sk_state) & (TCPF_ESTABLISHED | TCPF_CLOSE_WAIT) &&
inet_csk_ack_scheduled(sk)) {
inet_csk(sk)->icsk_ack.pending |= ICSK_ACK_PUSHED;
tcp_cleanup_rbuf(sk, 1);
if (!(val & 1))
inet_csk_enter_pingpong_mode(sk);
}
}
void tcp_sock_set_quickack(struct sock *sk, int val)
{
lock_sock(sk);
__tcp_sock_set_quickack(sk, val);
release_sock(sk);
}
EXPORT_SYMBOL(tcp_sock_set_quickack);
int tcp_sock_set_syncnt(struct sock *sk, int val)
{
if (val < 1 || val > MAX_TCP_SYNCNT)
return -EINVAL;
WRITE_ONCE(inet_csk(sk)->icsk_syn_retries, val);
return 0;
}
EXPORT_SYMBOL(tcp_sock_set_syncnt);
int tcp_sock_set_user_timeout(struct sock *sk, int val)
{
/* Cap the max time in ms TCP will retry or probe the window
* before giving up and aborting (ETIMEDOUT) a connection.
*/
if (val < 0)
return -EINVAL;
WRITE_ONCE(inet_csk(sk)->icsk_user_timeout, val);
return 0;
}
EXPORT_SYMBOL(tcp_sock_set_user_timeout);
int tcp_sock_set_keepidle_locked(struct sock *sk, int val)
{
struct tcp_sock *tp = tcp_sk(sk);
if (val < 1 || val > MAX_TCP_KEEPIDLE)
return -EINVAL;
/* Paired with WRITE_ONCE() in keepalive_time_when() */
WRITE_ONCE(tp->keepalive_time, val * HZ);
if (sock_flag(sk, SOCK_KEEPOPEN) &&
!((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN))) {
u32 elapsed = keepalive_time_elapsed(tp);
if (tp->keepalive_time > elapsed)
elapsed = tp->keepalive_time - elapsed;
else
elapsed = 0;
inet_csk_reset_keepalive_timer(sk, elapsed);
}
return 0;
}
int tcp_sock_set_keepidle(struct sock *sk, int val)
{
int err;
lock_sock(sk);
err = tcp_sock_set_keepidle_locked(sk, val);
release_sock(sk);
return err;
}
EXPORT_SYMBOL(tcp_sock_set_keepidle);
int tcp_sock_set_keepintvl(struct sock *sk, int val)
{
if (val < 1 || val > MAX_TCP_KEEPINTVL)
return -EINVAL;
WRITE_ONCE(tcp_sk(sk)->keepalive_intvl, val * HZ);
return 0;
}
EXPORT_SYMBOL(tcp_sock_set_keepintvl);
int tcp_sock_set_keepcnt(struct sock *sk, int val)
{
if (val < 1 || val > MAX_TCP_KEEPCNT)
return -EINVAL;
/* Paired with READ_ONCE() in keepalive_probes() */
WRITE_ONCE(tcp_sk(sk)->keepalive_probes, val);
return 0;
}
EXPORT_SYMBOL(tcp_sock_set_keepcnt);
int tcp_set_window_clamp(struct sock *sk, int val)
{
struct tcp_sock *tp = tcp_sk(sk);
if (!val) {
if (sk->sk_state != TCP_CLOSE)
return -EINVAL;
tp->window_clamp = 0;
} else {
u32 new_rcv_ssthresh, old_window_clamp = tp->window_clamp;
u32 new_window_clamp = val < SOCK_MIN_RCVBUF / 2 ?
SOCK_MIN_RCVBUF / 2 : val;
if (new_window_clamp == old_window_clamp)
return 0;
tp->window_clamp = new_window_clamp;
if (new_window_clamp < old_window_clamp) {
/* need to apply the reserved mem provisioning only
* when shrinking the window clamp
*/
__tcp_adjust_rcv_ssthresh(sk, tp->window_clamp);
} else {
new_rcv_ssthresh = min(tp->rcv_wnd, tp->window_clamp);
tp->rcv_ssthresh = max(new_rcv_ssthresh,
tp->rcv_ssthresh);
}
}
return 0;
}
/*
* Socket option code for TCP.
*/
int do_tcp_setsockopt(struct sock *sk, int level, int optname,
sockptr_t optval, unsigned int optlen)
{
struct tcp_sock *tp = tcp_sk(sk);
struct inet_connection_sock *icsk = inet_csk(sk);
struct net *net = sock_net(sk);
int val;
int err = 0;
/* These are data/string values, all the others are ints */
switch (optname) {
case TCP_CONGESTION: {
char name[TCP_CA_NAME_MAX];
if (optlen < 1)
return -EINVAL;
val = strncpy_from_sockptr(name, optval,
min_t(long, TCP_CA_NAME_MAX-1, optlen));
if (val < 0)
return -EFAULT;
name[val] = 0;
sockopt_lock_sock(sk);
err = tcp_set_congestion_control(sk, name, !has_current_bpf_ctx(),
sockopt_ns_capable(sock_net(sk)->user_ns,
CAP_NET_ADMIN));
sockopt_release_sock(sk);
return err;
}
case TCP_ULP: {
char name[TCP_ULP_NAME_MAX];
if (optlen < 1)
return -EINVAL;
val = strncpy_from_sockptr(name, optval,
min_t(long, TCP_ULP_NAME_MAX - 1,
optlen));
if (val < 0)
return -EFAULT;
name[val] = 0;
sockopt_lock_sock(sk);
err = tcp_set_ulp(sk, name);
sockopt_release_sock(sk);
return err;
}
case TCP_FASTOPEN_KEY: {
__u8 key[TCP_FASTOPEN_KEY_BUF_LENGTH];
__u8 *backup_key = NULL;
/* Allow a backup key as well to facilitate key rotation
* First key is the active one.
*/
if (optlen != TCP_FASTOPEN_KEY_LENGTH &&
optlen != TCP_FASTOPEN_KEY_BUF_LENGTH)
return -EINVAL;
if (copy_from_sockptr(key, optval, optlen))
return -EFAULT;
if (optlen == TCP_FASTOPEN_KEY_BUF_LENGTH)
backup_key = key + TCP_FASTOPEN_KEY_LENGTH;
return tcp_fastopen_reset_cipher(net, sk, key, backup_key);
}
default:
/* fallthru */
break;
}
if (optlen < sizeof(int))
return -EINVAL;
if (copy_from_sockptr(&val, optval, sizeof(val)))
return -EFAULT;
/* Handle options that can be set without locking the socket. */
switch (optname) {
case TCP_SYNCNT:
return tcp_sock_set_syncnt(sk, val);
case TCP_USER_TIMEOUT:
return tcp_sock_set_user_timeout(sk, val);
case TCP_KEEPINTVL:
return tcp_sock_set_keepintvl(sk, val);
case TCP_KEEPCNT:
return tcp_sock_set_keepcnt(sk, val);
case TCP_LINGER2:
if (val < 0)
WRITE_ONCE(tp->linger2, -1);
else if (val > TCP_FIN_TIMEOUT_MAX / HZ)
WRITE_ONCE(tp->linger2, TCP_FIN_TIMEOUT_MAX);
else
WRITE_ONCE(tp->linger2, val * HZ);
return 0;
case TCP_DEFER_ACCEPT:
/* Translate value in seconds to number of retransmits */
WRITE_ONCE(icsk->icsk_accept_queue.rskq_defer_accept,
secs_to_retrans(val, TCP_TIMEOUT_INIT / HZ,
TCP_RTO_MAX / HZ));
return 0;
}
sockopt_lock_sock(sk);
switch (optname) {
case TCP_MAXSEG:
/* Values greater than interface MTU won't take effect. However
* at the point when this call is done we typically don't yet
* know which interface is going to be used
*/
if (val && (val < TCP_MIN_MSS || val > MAX_TCP_WINDOW)) {
err = -EINVAL;
break;
}
tp->rx_opt.user_mss = val;
break;
case TCP_NODELAY:
__tcp_sock_set_nodelay(sk, val);
break;
case TCP_THIN_LINEAR_TIMEOUTS:
if (val < 0 || val > 1)
err = -EINVAL;
else
tp->thin_lto = val;
break;
case TCP_THIN_DUPACK:
if (val < 0 || val > 1)
err = -EINVAL;
break;
case TCP_REPAIR:
if (!tcp_can_repair_sock(sk))
err = -EPERM;
else if (val == TCP_REPAIR_ON) {
tp->repair = 1;
sk->sk_reuse = SK_FORCE_REUSE;
tp->repair_queue = TCP_NO_QUEUE;
} else if (val == TCP_REPAIR_OFF) {
tp->repair = 0;
sk->sk_reuse = SK_NO_REUSE;
tcp_send_window_probe(sk);
} else if (val == TCP_REPAIR_OFF_NO_WP) {
tp->repair = 0;
sk->sk_reuse = SK_NO_REUSE;
} else
err = -EINVAL;
break;
case TCP_REPAIR_QUEUE:
if (!tp->repair)
err = -EPERM;
else if ((unsigned int)val < TCP_QUEUES_NR)
tp->repair_queue = val;
else
err = -EINVAL;
break;
case TCP_QUEUE_SEQ:
tcp: add sanity tests to TCP_QUEUE_SEQ Qingyu Li reported a syzkaller bug where the repro changes RCV SEQ _after_ restoring data in the receive queue. mprotect(0x4aa000, 12288, PROT_READ) = 0 mmap(0x1ffff000, 4096, PROT_NONE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x1ffff000 mmap(0x20000000, 16777216, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x20000000 mmap(0x21000000, 4096, PROT_NONE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x21000000 socket(AF_INET6, SOCK_STREAM, IPPROTO_IP) = 3 setsockopt(3, SOL_TCP, TCP_REPAIR, [1], 4) = 0 connect(3, {sa_family=AF_INET6, sin6_port=htons(0), sin6_flowinfo=htonl(0), inet_pton(AF_INET6, "::1", &sin6_addr), sin6_scope_id=0}, 28) = 0 setsockopt(3, SOL_TCP, TCP_REPAIR_QUEUE, [1], 4) = 0 sendmsg(3, {msg_name=NULL, msg_namelen=0, msg_iov=[{iov_base="0x0000000000000003\0\0", iov_len=20}], msg_iovlen=1, msg_controllen=0, msg_flags=0}, 0) = 20 setsockopt(3, SOL_TCP, TCP_REPAIR, [0], 4) = 0 setsockopt(3, SOL_TCP, TCP_QUEUE_SEQ, [128], 4) = 0 recvfrom(3, NULL, 20, 0, NULL, NULL) = -1 ECONNRESET (Connection reset by peer) syslog shows: [ 111.205099] TCP recvmsg seq # bug 2: copied 80, seq 0, rcvnxt 80, fl 0 [ 111.207894] WARNING: CPU: 1 PID: 356 at net/ipv4/tcp.c:2343 tcp_recvmsg_locked+0x90e/0x29a0 This should not be allowed. TCP_QUEUE_SEQ should only be used when queues are empty. This patch fixes this case, and the tx path as well. Fixes: ee9952831cfd ("tcp: Initial repair mode") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Pavel Emelyanov <xemul@parallels.com> Link: https://bugzilla.kernel.org/show_bug.cgi?id=212005 Reported-by: Qingyu Li <ieatmuttonchuan@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-03-02 02:29:17 +08:00
if (sk->sk_state != TCP_CLOSE) {
err = -EPERM;
tcp: add sanity tests to TCP_QUEUE_SEQ Qingyu Li reported a syzkaller bug where the repro changes RCV SEQ _after_ restoring data in the receive queue. mprotect(0x4aa000, 12288, PROT_READ) = 0 mmap(0x1ffff000, 4096, PROT_NONE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x1ffff000 mmap(0x20000000, 16777216, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x20000000 mmap(0x21000000, 4096, PROT_NONE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x21000000 socket(AF_INET6, SOCK_STREAM, IPPROTO_IP) = 3 setsockopt(3, SOL_TCP, TCP_REPAIR, [1], 4) = 0 connect(3, {sa_family=AF_INET6, sin6_port=htons(0), sin6_flowinfo=htonl(0), inet_pton(AF_INET6, "::1", &sin6_addr), sin6_scope_id=0}, 28) = 0 setsockopt(3, SOL_TCP, TCP_REPAIR_QUEUE, [1], 4) = 0 sendmsg(3, {msg_name=NULL, msg_namelen=0, msg_iov=[{iov_base="0x0000000000000003\0\0", iov_len=20}], msg_iovlen=1, msg_controllen=0, msg_flags=0}, 0) = 20 setsockopt(3, SOL_TCP, TCP_REPAIR, [0], 4) = 0 setsockopt(3, SOL_TCP, TCP_QUEUE_SEQ, [128], 4) = 0 recvfrom(3, NULL, 20, 0, NULL, NULL) = -1 ECONNRESET (Connection reset by peer) syslog shows: [ 111.205099] TCP recvmsg seq # bug 2: copied 80, seq 0, rcvnxt 80, fl 0 [ 111.207894] WARNING: CPU: 1 PID: 356 at net/ipv4/tcp.c:2343 tcp_recvmsg_locked+0x90e/0x29a0 This should not be allowed. TCP_QUEUE_SEQ should only be used when queues are empty. This patch fixes this case, and the tx path as well. Fixes: ee9952831cfd ("tcp: Initial repair mode") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Pavel Emelyanov <xemul@parallels.com> Link: https://bugzilla.kernel.org/show_bug.cgi?id=212005 Reported-by: Qingyu Li <ieatmuttonchuan@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-03-02 02:29:17 +08:00
} else if (tp->repair_queue == TCP_SEND_QUEUE) {
if (!tcp_rtx_queue_empty(sk))
err = -EPERM;
else
WRITE_ONCE(tp->write_seq, val);
} else if (tp->repair_queue == TCP_RECV_QUEUE) {
if (tp->rcv_nxt != tp->copied_seq) {
err = -EPERM;
} else {
WRITE_ONCE(tp->rcv_nxt, val);
WRITE_ONCE(tp->copied_seq, val);
}
} else {
err = -EINVAL;
tcp: add sanity tests to TCP_QUEUE_SEQ Qingyu Li reported a syzkaller bug where the repro changes RCV SEQ _after_ restoring data in the receive queue. mprotect(0x4aa000, 12288, PROT_READ) = 0 mmap(0x1ffff000, 4096, PROT_NONE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x1ffff000 mmap(0x20000000, 16777216, PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x20000000 mmap(0x21000000, 4096, PROT_NONE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x21000000 socket(AF_INET6, SOCK_STREAM, IPPROTO_IP) = 3 setsockopt(3, SOL_TCP, TCP_REPAIR, [1], 4) = 0 connect(3, {sa_family=AF_INET6, sin6_port=htons(0), sin6_flowinfo=htonl(0), inet_pton(AF_INET6, "::1", &sin6_addr), sin6_scope_id=0}, 28) = 0 setsockopt(3, SOL_TCP, TCP_REPAIR_QUEUE, [1], 4) = 0 sendmsg(3, {msg_name=NULL, msg_namelen=0, msg_iov=[{iov_base="0x0000000000000003\0\0", iov_len=20}], msg_iovlen=1, msg_controllen=0, msg_flags=0}, 0) = 20 setsockopt(3, SOL_TCP, TCP_REPAIR, [0], 4) = 0 setsockopt(3, SOL_TCP, TCP_QUEUE_SEQ, [128], 4) = 0 recvfrom(3, NULL, 20, 0, NULL, NULL) = -1 ECONNRESET (Connection reset by peer) syslog shows: [ 111.205099] TCP recvmsg seq # bug 2: copied 80, seq 0, rcvnxt 80, fl 0 [ 111.207894] WARNING: CPU: 1 PID: 356 at net/ipv4/tcp.c:2343 tcp_recvmsg_locked+0x90e/0x29a0 This should not be allowed. TCP_QUEUE_SEQ should only be used when queues are empty. This patch fixes this case, and the tx path as well. Fixes: ee9952831cfd ("tcp: Initial repair mode") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Pavel Emelyanov <xemul@parallels.com> Link: https://bugzilla.kernel.org/show_bug.cgi?id=212005 Reported-by: Qingyu Li <ieatmuttonchuan@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-03-02 02:29:17 +08:00
}
break;
case TCP_REPAIR_OPTIONS:
if (!tp->repair)
err = -EINVAL;
tcp: prohibit TCP_REPAIR_OPTIONS if data was already sent If setsockopt with option name of TCP_REPAIR_OPTIONS and opt_code of TCPOPT_SACK_PERM is called to enable sack after data is sent and dupacks are received , it will trigger a warning in function tcp_verify_left_out() as follows: ============================================ WARNING: CPU: 8 PID: 0 at net/ipv4/tcp_input.c:2132 tcp_timeout_mark_lost+0x154/0x160 tcp_enter_loss+0x2b/0x290 tcp_retransmit_timer+0x50b/0x640 tcp_write_timer_handler+0x1c8/0x340 tcp_write_timer+0xe5/0x140 call_timer_fn+0x3a/0x1b0 __run_timers.part.0+0x1bf/0x2d0 run_timer_softirq+0x43/0xb0 __do_softirq+0xfd/0x373 __irq_exit_rcu+0xf6/0x140 The warning is caused in the following steps: 1. a socket named socketA is created 2. socketA enters repair mode without build a connection 3. socketA calls connect() and its state is changed to TCP_ESTABLISHED directly 4. socketA leaves repair mode 5. socketA calls sendmsg() to send data, packets_out and sack_outs(dup ack receives) increase 6. socketA enters repair mode again 7. socketA calls setsockopt with TCPOPT_SACK_PERM to enable sack 8. retransmit timer expires, it calls tcp_timeout_mark_lost(), lost_out increases 9. sack_outs + lost_out > packets_out triggers since lost_out and sack_outs increase repeatly In function tcp_timeout_mark_lost(), tp->sacked_out will be cleared if Step7 not happen and the warning will not be triggered. As suggested by Denis and Eric, TCP_REPAIR_OPTIONS should be prohibited if data was already sent. socket-tcp tests in CRIU has been tested as follows: $ sudo ./test/zdtm.py run -t zdtm/static/socket-tcp* --keep-going \ --ignore-taint socket-tcp* represent all socket-tcp tests in test/zdtm/static/. Fixes: b139ba4e90dc ("tcp: Repair connection-time negotiated parameters") Signed-off-by: Lu Wei <luwei32@huawei.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2022-11-04 10:27:23 +08:00
else if (sk->sk_state == TCP_ESTABLISHED && !tp->bytes_sent)
err = tcp_repair_options_est(sk, optval, optlen);
else
err = -EPERM;
break;
case TCP_CORK:
__tcp_sock_set_cork(sk, val);
break;
case TCP_KEEPIDLE:
err = tcp_sock_set_keepidle_locked(sk, val);
break;
case TCP_SAVE_SYN:
/* 0: disable, 1: enable, 2: start from ether_header */
if (val < 0 || val > 2)
err = -EINVAL;
else
tp->save_syn = val;
break;
case TCP_WINDOW_CLAMP:
err = tcp_set_window_clamp(sk, val);
break;
case TCP_QUICKACK:
__tcp_sock_set_quickack(sk, val);
break;
case TCP_AO_REPAIR:
if (!tcp_can_repair_sock(sk)) {
err = -EPERM;
break;
}
err = tcp_ao_set_repair(sk, optval, optlen);
break;
#ifdef CONFIG_TCP_AO
case TCP_AO_ADD_KEY:
case TCP_AO_DEL_KEY:
case TCP_AO_INFO: {
/* If this is the first TCP-AO setsockopt() on the socket,
* sk_state has to be LISTEN or CLOSE. Allow TCP_REPAIR
* in any state.
*/
if ((1 << sk->sk_state) & (TCPF_LISTEN | TCPF_CLOSE))
goto ao_parse;
if (rcu_dereference_protected(tcp_sk(sk)->ao_info,
lockdep_sock_is_held(sk)))
goto ao_parse;
if (tp->repair)
goto ao_parse;
err = -EISCONN;
break;
ao_parse:
err = tp->af_specific->ao_parse(sk, optname, optval, optlen);
break;
}
#endif
#ifdef CONFIG_TCP_MD5SIG
case TCP_MD5SIG:
case TCP_MD5SIG_EXT:
err = tp->af_specific->md5_parse(sk, optname, optval, optlen);
break;
#endif
case TCP_FASTOPEN:
if (val >= 0 && ((1 << sk->sk_state) & (TCPF_CLOSE |
tcp: Do not call tcp_fastopen_reset_cipher from interrupt context tcp_fastopen_reset_cipher really cannot be called from interrupt context. It allocates the tcp_fastopen_context with GFP_KERNEL and calls crypto_alloc_cipher, which allocates all kind of stuff with GFP_KERNEL. Thus, we might sleep when the key-generation is triggered by an incoming TFO cookie-request which would then happen in interrupt- context, as shown by enabling CONFIG_DEBUG_ATOMIC_SLEEP: [ 36.001813] BUG: sleeping function called from invalid context at mm/slub.c:1266 [ 36.003624] in_atomic(): 1, irqs_disabled(): 0, pid: 1016, name: packetdrill [ 36.004859] CPU: 1 PID: 1016 Comm: packetdrill Not tainted 4.1.0-rc7 #14 [ 36.006085] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.7.5-0-ge51488c-20140602_164612-nilsson.home.kraxel.org 04/01/2014 [ 36.008250] 00000000000004f2 ffff88007f8838a8 ffffffff8171d53a ffff880075a084a8 [ 36.009630] ffff880075a08000 ffff88007f8838c8 ffffffff810967d3 ffff88007f883928 [ 36.011076] 0000000000000000 ffff88007f8838f8 ffffffff81096892 ffff88007f89be00 [ 36.012494] Call Trace: [ 36.012953] <IRQ> [<ffffffff8171d53a>] dump_stack+0x4f/0x6d [ 36.014085] [<ffffffff810967d3>] ___might_sleep+0x103/0x170 [ 36.015117] [<ffffffff81096892>] __might_sleep+0x52/0x90 [ 36.016117] [<ffffffff8118e887>] kmem_cache_alloc_trace+0x47/0x190 [ 36.017266] [<ffffffff81680d82>] ? tcp_fastopen_reset_cipher+0x42/0x130 [ 36.018485] [<ffffffff81680d82>] tcp_fastopen_reset_cipher+0x42/0x130 [ 36.019679] [<ffffffff81680f01>] tcp_fastopen_init_key_once+0x61/0x70 [ 36.020884] [<ffffffff81680f2c>] __tcp_fastopen_cookie_gen+0x1c/0x60 [ 36.022058] [<ffffffff816814ff>] tcp_try_fastopen+0x58f/0x730 [ 36.023118] [<ffffffff81671788>] tcp_conn_request+0x3e8/0x7b0 [ 36.024185] [<ffffffff810e3872>] ? __module_text_address+0x12/0x60 [ 36.025327] [<ffffffff8167b2e1>] tcp_v4_conn_request+0x51/0x60 [ 36.026410] [<ffffffff816727e0>] tcp_rcv_state_process+0x190/0xda0 [ 36.027556] [<ffffffff81661f97>] ? __inet_lookup_established+0x47/0x170 [ 36.028784] [<ffffffff8167c2ad>] tcp_v4_do_rcv+0x16d/0x3d0 [ 36.029832] [<ffffffff812e6806>] ? security_sock_rcv_skb+0x16/0x20 [ 36.030936] [<ffffffff8167cc8a>] tcp_v4_rcv+0x77a/0x7b0 [ 36.031875] [<ffffffff816af8c3>] ? iptable_filter_hook+0x33/0x70 [ 36.032953] [<ffffffff81657d22>] ip_local_deliver_finish+0x92/0x1f0 [ 36.034065] [<ffffffff81657f1a>] ip_local_deliver+0x9a/0xb0 [ 36.035069] [<ffffffff81657c90>] ? ip_rcv+0x3d0/0x3d0 [ 36.035963] [<ffffffff81657569>] ip_rcv_finish+0x119/0x330 [ 36.036950] [<ffffffff81657ba7>] ip_rcv+0x2e7/0x3d0 [ 36.037847] [<ffffffff81610652>] __netif_receive_skb_core+0x552/0x930 [ 36.038994] [<ffffffff81610a57>] __netif_receive_skb+0x27/0x70 [ 36.040033] [<ffffffff81610b72>] process_backlog+0xd2/0x1f0 [ 36.041025] [<ffffffff81611482>] net_rx_action+0x122/0x310 [ 36.042007] [<ffffffff81076743>] __do_softirq+0x103/0x2f0 [ 36.042978] [<ffffffff81723e3c>] do_softirq_own_stack+0x1c/0x30 This patch moves the call to tcp_fastopen_init_key_once to the places where a listener socket creates its TFO-state, which always happens in user-context (either from the setsockopt, or implicitly during the listen()-call) Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Hannes Frederic Sowa <hannes@stressinduktion.org> Fixes: 222e83d2e0ae ("tcp: switch tcp_fastopen key generation to net_get_random_once") Signed-off-by: Christoph Paasch <cpaasch@apple.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-06-19 00:15:34 +08:00
TCPF_LISTEN))) {
tcp_fastopen_init_key_once(net);
tcp: Do not call tcp_fastopen_reset_cipher from interrupt context tcp_fastopen_reset_cipher really cannot be called from interrupt context. It allocates the tcp_fastopen_context with GFP_KERNEL and calls crypto_alloc_cipher, which allocates all kind of stuff with GFP_KERNEL. Thus, we might sleep when the key-generation is triggered by an incoming TFO cookie-request which would then happen in interrupt- context, as shown by enabling CONFIG_DEBUG_ATOMIC_SLEEP: [ 36.001813] BUG: sleeping function called from invalid context at mm/slub.c:1266 [ 36.003624] in_atomic(): 1, irqs_disabled(): 0, pid: 1016, name: packetdrill [ 36.004859] CPU: 1 PID: 1016 Comm: packetdrill Not tainted 4.1.0-rc7 #14 [ 36.006085] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.7.5-0-ge51488c-20140602_164612-nilsson.home.kraxel.org 04/01/2014 [ 36.008250] 00000000000004f2 ffff88007f8838a8 ffffffff8171d53a ffff880075a084a8 [ 36.009630] ffff880075a08000 ffff88007f8838c8 ffffffff810967d3 ffff88007f883928 [ 36.011076] 0000000000000000 ffff88007f8838f8 ffffffff81096892 ffff88007f89be00 [ 36.012494] Call Trace: [ 36.012953] <IRQ> [<ffffffff8171d53a>] dump_stack+0x4f/0x6d [ 36.014085] [<ffffffff810967d3>] ___might_sleep+0x103/0x170 [ 36.015117] [<ffffffff81096892>] __might_sleep+0x52/0x90 [ 36.016117] [<ffffffff8118e887>] kmem_cache_alloc_trace+0x47/0x190 [ 36.017266] [<ffffffff81680d82>] ? tcp_fastopen_reset_cipher+0x42/0x130 [ 36.018485] [<ffffffff81680d82>] tcp_fastopen_reset_cipher+0x42/0x130 [ 36.019679] [<ffffffff81680f01>] tcp_fastopen_init_key_once+0x61/0x70 [ 36.020884] [<ffffffff81680f2c>] __tcp_fastopen_cookie_gen+0x1c/0x60 [ 36.022058] [<ffffffff816814ff>] tcp_try_fastopen+0x58f/0x730 [ 36.023118] [<ffffffff81671788>] tcp_conn_request+0x3e8/0x7b0 [ 36.024185] [<ffffffff810e3872>] ? __module_text_address+0x12/0x60 [ 36.025327] [<ffffffff8167b2e1>] tcp_v4_conn_request+0x51/0x60 [ 36.026410] [<ffffffff816727e0>] tcp_rcv_state_process+0x190/0xda0 [ 36.027556] [<ffffffff81661f97>] ? __inet_lookup_established+0x47/0x170 [ 36.028784] [<ffffffff8167c2ad>] tcp_v4_do_rcv+0x16d/0x3d0 [ 36.029832] [<ffffffff812e6806>] ? security_sock_rcv_skb+0x16/0x20 [ 36.030936] [<ffffffff8167cc8a>] tcp_v4_rcv+0x77a/0x7b0 [ 36.031875] [<ffffffff816af8c3>] ? iptable_filter_hook+0x33/0x70 [ 36.032953] [<ffffffff81657d22>] ip_local_deliver_finish+0x92/0x1f0 [ 36.034065] [<ffffffff81657f1a>] ip_local_deliver+0x9a/0xb0 [ 36.035069] [<ffffffff81657c90>] ? ip_rcv+0x3d0/0x3d0 [ 36.035963] [<ffffffff81657569>] ip_rcv_finish+0x119/0x330 [ 36.036950] [<ffffffff81657ba7>] ip_rcv+0x2e7/0x3d0 [ 36.037847] [<ffffffff81610652>] __netif_receive_skb_core+0x552/0x930 [ 36.038994] [<ffffffff81610a57>] __netif_receive_skb+0x27/0x70 [ 36.040033] [<ffffffff81610b72>] process_backlog+0xd2/0x1f0 [ 36.041025] [<ffffffff81611482>] net_rx_action+0x122/0x310 [ 36.042007] [<ffffffff81076743>] __do_softirq+0x103/0x2f0 [ 36.042978] [<ffffffff81723e3c>] do_softirq_own_stack+0x1c/0x30 This patch moves the call to tcp_fastopen_init_key_once to the places where a listener socket creates its TFO-state, which always happens in user-context (either from the setsockopt, or implicitly during the listen()-call) Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Hannes Frederic Sowa <hannes@stressinduktion.org> Fixes: 222e83d2e0ae ("tcp: switch tcp_fastopen key generation to net_get_random_once") Signed-off-by: Christoph Paasch <cpaasch@apple.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-06-19 00:15:34 +08:00
fastopen_queue_tune(sk, val);
tcp: Do not call tcp_fastopen_reset_cipher from interrupt context tcp_fastopen_reset_cipher really cannot be called from interrupt context. It allocates the tcp_fastopen_context with GFP_KERNEL and calls crypto_alloc_cipher, which allocates all kind of stuff with GFP_KERNEL. Thus, we might sleep when the key-generation is triggered by an incoming TFO cookie-request which would then happen in interrupt- context, as shown by enabling CONFIG_DEBUG_ATOMIC_SLEEP: [ 36.001813] BUG: sleeping function called from invalid context at mm/slub.c:1266 [ 36.003624] in_atomic(): 1, irqs_disabled(): 0, pid: 1016, name: packetdrill [ 36.004859] CPU: 1 PID: 1016 Comm: packetdrill Not tainted 4.1.0-rc7 #14 [ 36.006085] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.7.5-0-ge51488c-20140602_164612-nilsson.home.kraxel.org 04/01/2014 [ 36.008250] 00000000000004f2 ffff88007f8838a8 ffffffff8171d53a ffff880075a084a8 [ 36.009630] ffff880075a08000 ffff88007f8838c8 ffffffff810967d3 ffff88007f883928 [ 36.011076] 0000000000000000 ffff88007f8838f8 ffffffff81096892 ffff88007f89be00 [ 36.012494] Call Trace: [ 36.012953] <IRQ> [<ffffffff8171d53a>] dump_stack+0x4f/0x6d [ 36.014085] [<ffffffff810967d3>] ___might_sleep+0x103/0x170 [ 36.015117] [<ffffffff81096892>] __might_sleep+0x52/0x90 [ 36.016117] [<ffffffff8118e887>] kmem_cache_alloc_trace+0x47/0x190 [ 36.017266] [<ffffffff81680d82>] ? tcp_fastopen_reset_cipher+0x42/0x130 [ 36.018485] [<ffffffff81680d82>] tcp_fastopen_reset_cipher+0x42/0x130 [ 36.019679] [<ffffffff81680f01>] tcp_fastopen_init_key_once+0x61/0x70 [ 36.020884] [<ffffffff81680f2c>] __tcp_fastopen_cookie_gen+0x1c/0x60 [ 36.022058] [<ffffffff816814ff>] tcp_try_fastopen+0x58f/0x730 [ 36.023118] [<ffffffff81671788>] tcp_conn_request+0x3e8/0x7b0 [ 36.024185] [<ffffffff810e3872>] ? __module_text_address+0x12/0x60 [ 36.025327] [<ffffffff8167b2e1>] tcp_v4_conn_request+0x51/0x60 [ 36.026410] [<ffffffff816727e0>] tcp_rcv_state_process+0x190/0xda0 [ 36.027556] [<ffffffff81661f97>] ? __inet_lookup_established+0x47/0x170 [ 36.028784] [<ffffffff8167c2ad>] tcp_v4_do_rcv+0x16d/0x3d0 [ 36.029832] [<ffffffff812e6806>] ? security_sock_rcv_skb+0x16/0x20 [ 36.030936] [<ffffffff8167cc8a>] tcp_v4_rcv+0x77a/0x7b0 [ 36.031875] [<ffffffff816af8c3>] ? iptable_filter_hook+0x33/0x70 [ 36.032953] [<ffffffff81657d22>] ip_local_deliver_finish+0x92/0x1f0 [ 36.034065] [<ffffffff81657f1a>] ip_local_deliver+0x9a/0xb0 [ 36.035069] [<ffffffff81657c90>] ? ip_rcv+0x3d0/0x3d0 [ 36.035963] [<ffffffff81657569>] ip_rcv_finish+0x119/0x330 [ 36.036950] [<ffffffff81657ba7>] ip_rcv+0x2e7/0x3d0 [ 36.037847] [<ffffffff81610652>] __netif_receive_skb_core+0x552/0x930 [ 36.038994] [<ffffffff81610a57>] __netif_receive_skb+0x27/0x70 [ 36.040033] [<ffffffff81610b72>] process_backlog+0xd2/0x1f0 [ 36.041025] [<ffffffff81611482>] net_rx_action+0x122/0x310 [ 36.042007] [<ffffffff81076743>] __do_softirq+0x103/0x2f0 [ 36.042978] [<ffffffff81723e3c>] do_softirq_own_stack+0x1c/0x30 This patch moves the call to tcp_fastopen_init_key_once to the places where a listener socket creates its TFO-state, which always happens in user-context (either from the setsockopt, or implicitly during the listen()-call) Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Hannes Frederic Sowa <hannes@stressinduktion.org> Fixes: 222e83d2e0ae ("tcp: switch tcp_fastopen key generation to net_get_random_once") Signed-off-by: Christoph Paasch <cpaasch@apple.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-06-19 00:15:34 +08:00
} else {
err = -EINVAL;
tcp: Do not call tcp_fastopen_reset_cipher from interrupt context tcp_fastopen_reset_cipher really cannot be called from interrupt context. It allocates the tcp_fastopen_context with GFP_KERNEL and calls crypto_alloc_cipher, which allocates all kind of stuff with GFP_KERNEL. Thus, we might sleep when the key-generation is triggered by an incoming TFO cookie-request which would then happen in interrupt- context, as shown by enabling CONFIG_DEBUG_ATOMIC_SLEEP: [ 36.001813] BUG: sleeping function called from invalid context at mm/slub.c:1266 [ 36.003624] in_atomic(): 1, irqs_disabled(): 0, pid: 1016, name: packetdrill [ 36.004859] CPU: 1 PID: 1016 Comm: packetdrill Not tainted 4.1.0-rc7 #14 [ 36.006085] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.7.5-0-ge51488c-20140602_164612-nilsson.home.kraxel.org 04/01/2014 [ 36.008250] 00000000000004f2 ffff88007f8838a8 ffffffff8171d53a ffff880075a084a8 [ 36.009630] ffff880075a08000 ffff88007f8838c8 ffffffff810967d3 ffff88007f883928 [ 36.011076] 0000000000000000 ffff88007f8838f8 ffffffff81096892 ffff88007f89be00 [ 36.012494] Call Trace: [ 36.012953] <IRQ> [<ffffffff8171d53a>] dump_stack+0x4f/0x6d [ 36.014085] [<ffffffff810967d3>] ___might_sleep+0x103/0x170 [ 36.015117] [<ffffffff81096892>] __might_sleep+0x52/0x90 [ 36.016117] [<ffffffff8118e887>] kmem_cache_alloc_trace+0x47/0x190 [ 36.017266] [<ffffffff81680d82>] ? tcp_fastopen_reset_cipher+0x42/0x130 [ 36.018485] [<ffffffff81680d82>] tcp_fastopen_reset_cipher+0x42/0x130 [ 36.019679] [<ffffffff81680f01>] tcp_fastopen_init_key_once+0x61/0x70 [ 36.020884] [<ffffffff81680f2c>] __tcp_fastopen_cookie_gen+0x1c/0x60 [ 36.022058] [<ffffffff816814ff>] tcp_try_fastopen+0x58f/0x730 [ 36.023118] [<ffffffff81671788>] tcp_conn_request+0x3e8/0x7b0 [ 36.024185] [<ffffffff810e3872>] ? __module_text_address+0x12/0x60 [ 36.025327] [<ffffffff8167b2e1>] tcp_v4_conn_request+0x51/0x60 [ 36.026410] [<ffffffff816727e0>] tcp_rcv_state_process+0x190/0xda0 [ 36.027556] [<ffffffff81661f97>] ? __inet_lookup_established+0x47/0x170 [ 36.028784] [<ffffffff8167c2ad>] tcp_v4_do_rcv+0x16d/0x3d0 [ 36.029832] [<ffffffff812e6806>] ? security_sock_rcv_skb+0x16/0x20 [ 36.030936] [<ffffffff8167cc8a>] tcp_v4_rcv+0x77a/0x7b0 [ 36.031875] [<ffffffff816af8c3>] ? iptable_filter_hook+0x33/0x70 [ 36.032953] [<ffffffff81657d22>] ip_local_deliver_finish+0x92/0x1f0 [ 36.034065] [<ffffffff81657f1a>] ip_local_deliver+0x9a/0xb0 [ 36.035069] [<ffffffff81657c90>] ? ip_rcv+0x3d0/0x3d0 [ 36.035963] [<ffffffff81657569>] ip_rcv_finish+0x119/0x330 [ 36.036950] [<ffffffff81657ba7>] ip_rcv+0x2e7/0x3d0 [ 36.037847] [<ffffffff81610652>] __netif_receive_skb_core+0x552/0x930 [ 36.038994] [<ffffffff81610a57>] __netif_receive_skb+0x27/0x70 [ 36.040033] [<ffffffff81610b72>] process_backlog+0xd2/0x1f0 [ 36.041025] [<ffffffff81611482>] net_rx_action+0x122/0x310 [ 36.042007] [<ffffffff81076743>] __do_softirq+0x103/0x2f0 [ 36.042978] [<ffffffff81723e3c>] do_softirq_own_stack+0x1c/0x30 This patch moves the call to tcp_fastopen_init_key_once to the places where a listener socket creates its TFO-state, which always happens in user-context (either from the setsockopt, or implicitly during the listen()-call) Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Hannes Frederic Sowa <hannes@stressinduktion.org> Fixes: 222e83d2e0ae ("tcp: switch tcp_fastopen key generation to net_get_random_once") Signed-off-by: Christoph Paasch <cpaasch@apple.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-06-19 00:15:34 +08:00
}
break;
net/tcp-fastopen: Add new API support This patch adds a new socket option, TCP_FASTOPEN_CONNECT, as an alternative way to perform Fast Open on the active side (client). Prior to this patch, a client needs to replace the connect() call with sendto(MSG_FASTOPEN). This can be cumbersome for applications who want to use Fast Open: these socket operations are often done in lower layer libraries used by many other applications. Changing these libraries and/or the socket call sequences are not trivial. A more convenient approach is to perform Fast Open by simply enabling a socket option when the socket is created w/o changing other socket calls sequence: s = socket() create a new socket setsockopt(s, IPPROTO_TCP, TCP_FASTOPEN_CONNECT …); newly introduced sockopt If set, new functionality described below will be used. Return ENOTSUPP if TFO is not supported or not enabled in the kernel. connect() With cookie present, return 0 immediately. With no cookie, initiate 3WHS with TFO cookie-request option and return -1 with errno = EINPROGRESS. write()/sendmsg() With cookie present, send out SYN with data and return the number of bytes buffered. With no cookie, and 3WHS not yet completed, return -1 with errno = EINPROGRESS. No MSG_FASTOPEN flag is needed. read() Return -1 with errno = EWOULDBLOCK/EAGAIN if connect() is called but write() is not called yet. Return -1 with errno = EWOULDBLOCK/EAGAIN if connection is established but no msg is received yet. Return number of bytes read if socket is established and there is msg received. The new API simplifies life for applications that always perform a write() immediately after a successful connect(). Such applications can now take advantage of Fast Open by merely making one new setsockopt() call at the time of creating the socket. Nothing else about the application's socket call sequence needs to change. Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-24 02:59:22 +08:00
case TCP_FASTOPEN_CONNECT:
if (val > 1 || val < 0) {
err = -EINVAL;
} else if (READ_ONCE(net->ipv4.sysctl_tcp_fastopen) &
TFO_CLIENT_ENABLE) {
net/tcp-fastopen: Add new API support This patch adds a new socket option, TCP_FASTOPEN_CONNECT, as an alternative way to perform Fast Open on the active side (client). Prior to this patch, a client needs to replace the connect() call with sendto(MSG_FASTOPEN). This can be cumbersome for applications who want to use Fast Open: these socket operations are often done in lower layer libraries used by many other applications. Changing these libraries and/or the socket call sequences are not trivial. A more convenient approach is to perform Fast Open by simply enabling a socket option when the socket is created w/o changing other socket calls sequence: s = socket() create a new socket setsockopt(s, IPPROTO_TCP, TCP_FASTOPEN_CONNECT …); newly introduced sockopt If set, new functionality described below will be used. Return ENOTSUPP if TFO is not supported or not enabled in the kernel. connect() With cookie present, return 0 immediately. With no cookie, initiate 3WHS with TFO cookie-request option and return -1 with errno = EINPROGRESS. write()/sendmsg() With cookie present, send out SYN with data and return the number of bytes buffered. With no cookie, and 3WHS not yet completed, return -1 with errno = EINPROGRESS. No MSG_FASTOPEN flag is needed. read() Return -1 with errno = EWOULDBLOCK/EAGAIN if connect() is called but write() is not called yet. Return -1 with errno = EWOULDBLOCK/EAGAIN if connection is established but no msg is received yet. Return number of bytes read if socket is established and there is msg received. The new API simplifies life for applications that always perform a write() immediately after a successful connect(). Such applications can now take advantage of Fast Open by merely making one new setsockopt() call at the time of creating the socket. Nothing else about the application's socket call sequence needs to change. Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-24 02:59:22 +08:00
if (sk->sk_state == TCP_CLOSE)
tp->fastopen_connect = val;
else
err = -EINVAL;
} else {
err = -EOPNOTSUPP;
}
break;
case TCP_FASTOPEN_NO_COOKIE:
if (val > 1 || val < 0)
err = -EINVAL;
else if (!((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN)))
err = -EINVAL;
else
tp->fastopen_no_cookie = val;
break;
case TCP_TIMESTAMP:
if (!tp->repair) {
err = -EPERM;
break;
}
/* val is an opaque field,
* and low order bit contains usec_ts enable bit.
* Its a best effort, and we do not care if user makes an error.
*/
tp->tcp_usec_ts = val & 1;
WRITE_ONCE(tp->tsoffset, val - tcp_clock_ts(tp->tcp_usec_ts));
break;
case TCP_REPAIR_WINDOW:
err = tcp_repair_set_window(tp, optval, optlen);
break;
tcp: TCP_NOTSENT_LOWAT socket option Idea of this patch is to add optional limitation of number of unsent bytes in TCP sockets, to reduce usage of kernel memory. TCP receiver might announce a big window, and TCP sender autotuning might allow a large amount of bytes in write queue, but this has little performance impact if a large part of this buffering is wasted : Write queue needs to be large only to deal with large BDP, not necessarily to cope with scheduling delays (incoming ACKS make room for the application to queue more bytes) For most workloads, using a value of 128 KB or less is OK to give applications enough time to react to POLLOUT events in time (or being awaken in a blocking sendmsg()) This patch adds two ways to set the limit : 1) Per socket option TCP_NOTSENT_LOWAT 2) A sysctl (/proc/sys/net/ipv4/tcp_notsent_lowat) for sockets not using TCP_NOTSENT_LOWAT socket option (or setting a zero value) Default value being UINT_MAX (0xFFFFFFFF), meaning this has no effect. This changes poll()/select()/epoll() to report POLLOUT only if number of unsent bytes is below tp->nosent_lowat Note this might increase number of sendmsg()/sendfile() calls when using non blocking sockets, and increase number of context switches for blocking sockets. Note this is not related to SO_SNDLOWAT (as SO_SNDLOWAT is defined as : Specify the minimum number of bytes in the buffer until the socket layer will pass the data to the protocol) Tested: netperf sessions, and watching /proc/net/protocols "memory" column for TCP With 200 concurrent netperf -t TCP_STREAM sessions, amount of kernel memory used by TCP buffers shrinks by ~55 % (20567 pages instead of 45458) lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols TCPv6 1880 2 45458 no 208 yes ipv6 y y y y y y y y y y y y y n y y y y y TCP 1696 508 45458 no 208 yes kernel y y y y y y y y y y y y y n y y y y y lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols TCPv6 1880 2 20567 no 208 yes ipv6 y y y y y y y y y y y y y n y y y y y TCP 1696 508 20567 no 208 yes kernel y y y y y y y y y y y y y n y y y y y Using 128KB has no bad effect on the throughput or cpu usage of a single flow, although there is an increase of context switches. A bonus is that we hold socket lock for a shorter amount of time and should improve latencies of ACK processing. lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3 OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf. Local Remote Local Elapsed Throughput Throughput Local Local Remote Remote Local Remote Service Send Socket Recv Socket Send Time Units CPU CPU CPU CPU Service Service Demand Size Size Size (sec) Util Util Util Util Demand Demand Units Final Final % Method % Method 1651584 6291456 16384 20.00 17447.90 10^6bits/s 3.13 S -1.00 U 0.353 -1.000 usec/KB Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3': 412,514 context-switches 200.034645535 seconds time elapsed lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3 OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf. Local Remote Local Elapsed Throughput Throughput Local Local Remote Remote Local Remote Service Send Socket Recv Socket Send Time Units CPU CPU CPU CPU Service Service Demand Size Size Size (sec) Util Util Util Util Demand Demand Units Final Final % Method % Method 1593240 6291456 16384 20.00 17321.16 10^6bits/s 3.35 S -1.00 U 0.381 -1.000 usec/KB Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3': 2,675,818 context-switches 200.029651391 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-By: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-07-23 11:27:07 +08:00
case TCP_NOTSENT_LOWAT:
WRITE_ONCE(tp->notsent_lowat, val);
tcp: TCP_NOTSENT_LOWAT socket option Idea of this patch is to add optional limitation of number of unsent bytes in TCP sockets, to reduce usage of kernel memory. TCP receiver might announce a big window, and TCP sender autotuning might allow a large amount of bytes in write queue, but this has little performance impact if a large part of this buffering is wasted : Write queue needs to be large only to deal with large BDP, not necessarily to cope with scheduling delays (incoming ACKS make room for the application to queue more bytes) For most workloads, using a value of 128 KB or less is OK to give applications enough time to react to POLLOUT events in time (or being awaken in a blocking sendmsg()) This patch adds two ways to set the limit : 1) Per socket option TCP_NOTSENT_LOWAT 2) A sysctl (/proc/sys/net/ipv4/tcp_notsent_lowat) for sockets not using TCP_NOTSENT_LOWAT socket option (or setting a zero value) Default value being UINT_MAX (0xFFFFFFFF), meaning this has no effect. This changes poll()/select()/epoll() to report POLLOUT only if number of unsent bytes is below tp->nosent_lowat Note this might increase number of sendmsg()/sendfile() calls when using non blocking sockets, and increase number of context switches for blocking sockets. Note this is not related to SO_SNDLOWAT (as SO_SNDLOWAT is defined as : Specify the minimum number of bytes in the buffer until the socket layer will pass the data to the protocol) Tested: netperf sessions, and watching /proc/net/protocols "memory" column for TCP With 200 concurrent netperf -t TCP_STREAM sessions, amount of kernel memory used by TCP buffers shrinks by ~55 % (20567 pages instead of 45458) lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols TCPv6 1880 2 45458 no 208 yes ipv6 y y y y y y y y y y y y y n y y y y y TCP 1696 508 45458 no 208 yes kernel y y y y y y y y y y y y y n y y y y y lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols TCPv6 1880 2 20567 no 208 yes ipv6 y y y y y y y y y y y y y n y y y y y TCP 1696 508 20567 no 208 yes kernel y y y y y y y y y y y y y n y y y y y Using 128KB has no bad effect on the throughput or cpu usage of a single flow, although there is an increase of context switches. A bonus is that we hold socket lock for a shorter amount of time and should improve latencies of ACK processing. lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3 OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf. Local Remote Local Elapsed Throughput Throughput Local Local Remote Remote Local Remote Service Send Socket Recv Socket Send Time Units CPU CPU CPU CPU Service Service Demand Size Size Size (sec) Util Util Util Util Demand Demand Units Final Final % Method % Method 1651584 6291456 16384 20.00 17447.90 10^6bits/s 3.13 S -1.00 U 0.353 -1.000 usec/KB Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3': 412,514 context-switches 200.034645535 seconds time elapsed lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3 OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf. Local Remote Local Elapsed Throughput Throughput Local Local Remote Remote Local Remote Service Send Socket Recv Socket Send Time Units CPU CPU CPU CPU Service Service Demand Size Size Size (sec) Util Util Util Util Demand Demand Units Final Final % Method % Method 1593240 6291456 16384 20.00 17321.16 10^6bits/s 3.35 S -1.00 U 0.381 -1.000 usec/KB Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3': 2,675,818 context-switches 200.029651391 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-By: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-07-23 11:27:07 +08:00
sk->sk_write_space(sk);
break;
tcp: send in-queue bytes in cmsg upon read Applications with many concurrent connections, high variance in receive queue length and tight memory bounds cannot allocate worst-case buffer size to drain sockets. Knowing the size of receive queue length, applications can optimize how they allocate buffers to read from the socket. The number of bytes pending on the socket is directly available through ioctl(FIONREAD/SIOCINQ) and can be approximated using getsockopt(MEMINFO) (rmem_alloc includes skb overheads in addition to application data). But, both of these options add an extra syscall per recvmsg. Moreover, ioctl(FIONREAD/SIOCINQ) takes the socket lock. Add the TCP_INQ socket option to TCP. When this socket option is set, recvmsg() relays the number of bytes available on the socket for reading to the application via the TCP_CM_INQ control message. Calculate the number of bytes after releasing the socket lock to include the processed backlog, if any. To avoid an extra branch in the hot path of recvmsg() for this new control message, move all cmsg processing inside an existing branch for processing receive timestamps. Since the socket lock is not held when calculating the size of receive queue, TCP_INQ is a hint. For example, it can overestimate the queue size by one byte, if FIN is received. With this method, applications can start reading from the socket using a small buffer, and then use larger buffers based on the remaining data when needed. V3 change-log: As suggested by David Miller, added loads with barrier to check whether we have multiple threads calling recvmsg in parallel. When that happens we lock the socket to calculate inq. V4 change-log: Removed inline from a static function. Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Willem de Bruijn <willemb@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Neal Cardwell <ncardwell@google.com> Suggested-by: David Miller <davem@davemloft.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-05-02 03:39:15 +08:00
case TCP_INQ:
if (val > 1 || val < 0)
err = -EINVAL;
else
tp->recvmsg_inq = val;
break;
tcp: add optional per socket transmit delay Adding delays to TCP flows is crucial for studying behavior of TCP stacks, including congestion control modules. Linux offers netem module, but it has unpractical constraints : - Need root access to change qdisc - Hard to setup on egress if combined with non trivial qdisc like FQ - Single delay for all flows. EDT (Earliest Departure Time) adoption in TCP stack allows us to enable a per socket delay at a very small cost. Networking tools can now establish thousands of flows, each of them with a different delay, simulating real world conditions. This requires FQ packet scheduler or a EDT-enabled NIC. This patchs adds TCP_TX_DELAY socket option, to set a delay in usec units. unsigned int tx_delay = 10000; /* 10 msec */ setsockopt(fd, SOL_TCP, TCP_TX_DELAY, &tx_delay, sizeof(tx_delay)); Note that FQ packet scheduler limits might need some tweaking : man tc-fq PARAMETERS limit Hard limit on the real queue size. When this limit is reached, new packets are dropped. If the value is lowered, packets are dropped so that the new limit is met. Default is 10000 packets. flow_limit Hard limit on the maximum number of packets queued per flow. Default value is 100. Use of TCP_TX_DELAY option will increase number of skbs in FQ qdisc, so packets would be dropped if any of the previous limit is hit. Use of a jump label makes this support runtime-free, for hosts never using the option. Also note that TSQ (TCP Small Queues) limits are slightly changed with this patch : we need to account that skbs artificially delayed wont stop us providind more skbs to feed the pipe (netem uses skb_orphan_partial() for this purpose, but FQ can not use this trick) Because of that, using big delays might very well trigger old bugs in TSO auto defer logic and/or sndbuf limited detection. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2019-06-13 02:57:25 +08:00
case TCP_TX_DELAY:
if (val)
tcp_enable_tx_delay();
WRITE_ONCE(tp->tcp_tx_delay, val);
tcp: add optional per socket transmit delay Adding delays to TCP flows is crucial for studying behavior of TCP stacks, including congestion control modules. Linux offers netem module, but it has unpractical constraints : - Need root access to change qdisc - Hard to setup on egress if combined with non trivial qdisc like FQ - Single delay for all flows. EDT (Earliest Departure Time) adoption in TCP stack allows us to enable a per socket delay at a very small cost. Networking tools can now establish thousands of flows, each of them with a different delay, simulating real world conditions. This requires FQ packet scheduler or a EDT-enabled NIC. This patchs adds TCP_TX_DELAY socket option, to set a delay in usec units. unsigned int tx_delay = 10000; /* 10 msec */ setsockopt(fd, SOL_TCP, TCP_TX_DELAY, &tx_delay, sizeof(tx_delay)); Note that FQ packet scheduler limits might need some tweaking : man tc-fq PARAMETERS limit Hard limit on the real queue size. When this limit is reached, new packets are dropped. If the value is lowered, packets are dropped so that the new limit is met. Default is 10000 packets. flow_limit Hard limit on the maximum number of packets queued per flow. Default value is 100. Use of TCP_TX_DELAY option will increase number of skbs in FQ qdisc, so packets would be dropped if any of the previous limit is hit. Use of a jump label makes this support runtime-free, for hosts never using the option. Also note that TSQ (TCP Small Queues) limits are slightly changed with this patch : we need to account that skbs artificially delayed wont stop us providind more skbs to feed the pipe (netem uses skb_orphan_partial() for this purpose, but FQ can not use this trick) Because of that, using big delays might very well trigger old bugs in TSO auto defer logic and/or sndbuf limited detection. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2019-06-13 02:57:25 +08:00
break;
default:
err = -ENOPROTOOPT;
break;
}
sockopt_release_sock(sk);
return err;
}
int tcp_setsockopt(struct sock *sk, int level, int optname, sockptr_t optval,
unsigned int optlen)
{
const struct inet_connection_sock *icsk = inet_csk(sk);
if (level != SOL_TCP)
tcp: Fix data races around icsk->icsk_af_ops. setsockopt(IPV6_ADDRFORM) and tcp_v6_connect() change icsk->icsk_af_ops under lock_sock(), but tcp_(get|set)sockopt() read it locklessly. To avoid load/store tearing, we need to add READ_ONCE() and WRITE_ONCE() for the reads and writes. Thanks to Eric Dumazet for providing the syzbot report: BUG: KCSAN: data-race in tcp_setsockopt / tcp_v6_connect write to 0xffff88813c624518 of 8 bytes by task 23936 on cpu 0: tcp_v6_connect+0x5b3/0xce0 net/ipv6/tcp_ipv6.c:240 __inet_stream_connect+0x159/0x6d0 net/ipv4/af_inet.c:660 inet_stream_connect+0x44/0x70 net/ipv4/af_inet.c:724 __sys_connect_file net/socket.c:1976 [inline] __sys_connect+0x197/0x1b0 net/socket.c:1993 __do_sys_connect net/socket.c:2003 [inline] __se_sys_connect net/socket.c:2000 [inline] __x64_sys_connect+0x3d/0x50 net/socket.c:2000 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x2b/0x70 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd read to 0xffff88813c624518 of 8 bytes by task 23937 on cpu 1: tcp_setsockopt+0x147/0x1c80 net/ipv4/tcp.c:3789 sock_common_setsockopt+0x5d/0x70 net/core/sock.c:3585 __sys_setsockopt+0x212/0x2b0 net/socket.c:2252 __do_sys_setsockopt net/socket.c:2263 [inline] __se_sys_setsockopt net/socket.c:2260 [inline] __x64_sys_setsockopt+0x62/0x70 net/socket.c:2260 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x2b/0x70 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd value changed: 0xffffffff8539af68 -> 0xffffffff8539aff8 Reported by Kernel Concurrency Sanitizer on: CPU: 1 PID: 23937 Comm: syz-executor.5 Not tainted 6.0.0-rc4-syzkaller-00331-g4ed9c1e971b1-dirty #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 08/26/2022 Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot <syzkaller@googlegroups.com> Reported-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2022-10-07 02:53:49 +08:00
/* Paired with WRITE_ONCE() in do_ipv6_setsockopt() and tcp_v6_connect() */
return READ_ONCE(icsk->icsk_af_ops)->setsockopt(sk, level, optname,
optval, optlen);
return do_tcp_setsockopt(sk, level, optname, optval, optlen);
}
EXPORT_SYMBOL(tcp_setsockopt);
static void tcp_get_info_chrono_stats(const struct tcp_sock *tp,
struct tcp_info *info)
{
u64 stats[__TCP_CHRONO_MAX], total = 0;
enum tcp_chrono i;
for (i = TCP_CHRONO_BUSY; i < __TCP_CHRONO_MAX; ++i) {
stats[i] = tp->chrono_stat[i - 1];
if (i == tp->chrono_type)
stats[i] += tcp_jiffies32 - tp->chrono_start;
stats[i] *= USEC_PER_SEC / HZ;
total += stats[i];
}
info->tcpi_busy_time = total;
info->tcpi_rwnd_limited = stats[TCP_CHRONO_RWND_LIMITED];
info->tcpi_sndbuf_limited = stats[TCP_CHRONO_SNDBUF_LIMITED];
}
/* Return information about state of tcp endpoint in API format. */
void tcp_get_info(struct sock *sk, struct tcp_info *info)
{
const struct tcp_sock *tp = tcp_sk(sk); /* iff sk_type == SOCK_STREAM */
const struct inet_connection_sock *icsk = inet_csk(sk);
unsigned long rate;
u32 now;
u64 rate64;
bool slow;
memset(info, 0, sizeof(*info));
if (sk->sk_type != SOCK_STREAM)
return;
info->tcpi_state = inet_sk_state_load(sk);
/* Report meaningful fields for all TCP states, including listeners */
rate = READ_ONCE(sk->sk_pacing_rate);
rate64 = (rate != ~0UL) ? rate : ~0ULL;
info->tcpi_pacing_rate = rate64;
rate = READ_ONCE(sk->sk_max_pacing_rate);
rate64 = (rate != ~0UL) ? rate : ~0ULL;
info->tcpi_max_pacing_rate = rate64;
info->tcpi_reordering = tp->reordering;
info->tcpi_snd_cwnd = tcp_snd_cwnd(tp);
if (info->tcpi_state == TCP_LISTEN) {
/* listeners aliased fields :
* tcpi_unacked -> Number of children ready for accept()
* tcpi_sacked -> max backlog
*/
info->tcpi_unacked = READ_ONCE(sk->sk_ack_backlog);
info->tcpi_sacked = READ_ONCE(sk->sk_max_ack_backlog);
return;
}
slow = lock_sock_fast(sk);
info->tcpi_ca_state = icsk->icsk_ca_state;
info->tcpi_retransmits = icsk->icsk_retransmits;
info->tcpi_probes = icsk->icsk_probes_out;
info->tcpi_backoff = icsk->icsk_backoff;
if (tp->rx_opt.tstamp_ok)
info->tcpi_options |= TCPI_OPT_TIMESTAMPS;
if (tcp_is_sack(tp))
info->tcpi_options |= TCPI_OPT_SACK;
if (tp->rx_opt.wscale_ok) {
info->tcpi_options |= TCPI_OPT_WSCALE;
info->tcpi_snd_wscale = tp->rx_opt.snd_wscale;
info->tcpi_rcv_wscale = tp->rx_opt.rcv_wscale;
}
if (tp->ecn_flags & TCP_ECN_OK)
info->tcpi_options |= TCPI_OPT_ECN;
if (tp->ecn_flags & TCP_ECN_SEEN)
info->tcpi_options |= TCPI_OPT_ECN_SEEN;
if (tp->syn_data_acked)
info->tcpi_options |= TCPI_OPT_SYN_DATA;
if (tp->tcp_usec_ts)
info->tcpi_options |= TCPI_OPT_USEC_TS;
info->tcpi_rto = jiffies_to_usecs(icsk->icsk_rto);
info->tcpi_ato = jiffies_to_usecs(min_t(u32, icsk->icsk_ack.ato,
tcp_delack_max(sk)));
info->tcpi_snd_mss = tp->mss_cache;
info->tcpi_rcv_mss = icsk->icsk_ack.rcv_mss;
info->tcpi_unacked = tp->packets_out;
info->tcpi_sacked = tp->sacked_out;
info->tcpi_lost = tp->lost_out;
info->tcpi_retrans = tp->retrans_out;
now = tcp_jiffies32;
info->tcpi_last_data_sent = jiffies_to_msecs(now - tp->lsndtime);
info->tcpi_last_data_recv = jiffies_to_msecs(now - icsk->icsk_ack.lrcvtime);
info->tcpi_last_ack_recv = jiffies_to_msecs(now - tp->rcv_tstamp);
info->tcpi_pmtu = icsk->icsk_pmtu_cookie;
info->tcpi_rcv_ssthresh = tp->rcv_ssthresh;
tcp: switch rtt estimations to usec resolution Upcoming congestion controls for TCP require usec resolution for RTT estimations. Millisecond resolution is simply not enough these days. FQ/pacing in DC environments also require this change for finer control and removal of bimodal behavior due to the current hack in tcp_update_pacing_rate() for 'small rtt' TCP_CONG_RTT_STAMP is no longer needed. As Julian Anastasov pointed out, we need to keep user compatibility : tcp_metrics used to export RTT and RTTVAR in msec resolution, so we added RTT_US and RTTVAR_US. An iproute2 patch is needed to use the new attributes if provided by the kernel. In this example ss command displays a srtt of 32 usecs (10Gbit link) lpk51:~# ./ss -i dst lpk52 Netid State Recv-Q Send-Q Local Address:Port Peer Address:Port tcp ESTAB 0 1 10.246.11.51:42959 10.246.11.52:64614 cubic wscale:6,6 rto:201 rtt:0.032/0.001 ato:40 mss:1448 cwnd:10 send 3620.0Mbps pacing_rate 7240.0Mbps unacked:1 rcv_rtt:993 rcv_space:29559 Updated iproute2 ip command displays : lpk51:~# ./ip tcp_metrics | grep 10.246.11.52 10.246.11.52 age 561.914sec cwnd 10 rtt 274us rttvar 213us source 10.246.11.51 Old binary displays : lpk51:~# ip tcp_metrics | grep 10.246.11.52 10.246.11.52 age 561.914sec cwnd 10 rtt 250us rttvar 125us source 10.246.11.51 With help from Julian Anastasov, Stephen Hemminger and Yuchung Cheng Signed-off-by: Eric Dumazet <edumazet@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Cc: Stephen Hemminger <stephen@networkplumber.org> Cc: Yuchung Cheng <ycheng@google.com> Cc: Larry Brakmo <brakmo@google.com> Cc: Julian Anastasov <ja@ssi.bg> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-27 06:02:48 +08:00
info->tcpi_rtt = tp->srtt_us >> 3;
info->tcpi_rttvar = tp->mdev_us >> 2;
info->tcpi_snd_ssthresh = tp->snd_ssthresh;
info->tcpi_advmss = tp->advmss;
info->tcpi_rcv_rtt = tp->rcv_rtt_est.rtt_us >> 3;
info->tcpi_rcv_space = tp->rcvq_space.space;
info->tcpi_total_retrans = tp->total_retrans;
info->tcpi_bytes_acked = tp->bytes_acked;
info->tcpi_bytes_received = tp->bytes_received;
info->tcpi_notsent_bytes = max_t(int, 0, tp->write_seq - tp->snd_nxt);
tcp_get_info_chrono_stats(tp, info);
info->tcpi_segs_out = tp->segs_out;
/* segs_in and data_segs_in can be updated from tcp_segs_in() from BH */
info->tcpi_segs_in = READ_ONCE(tp->segs_in);
info->tcpi_data_segs_in = READ_ONCE(tp->data_segs_in);
info->tcpi_min_rtt = tcp_min_rtt(tp);
tcp: Add RFC4898 tcpEStatsPerfDataSegsOut/In Per RFC4898, they count segments sent/received containing a positive length data segment (that includes retransmission segments carrying data). Unlike tcpi_segs_out/in, tcpi_data_segs_out/in excludes segments carrying no data (e.g. pure ack). The patch also updates the segs_in in tcp_fastopen_add_skb() so that segs_in >= data_segs_in property is kept. Together with retransmission data, tcpi_data_segs_out gives a better signal on the rxmit rate. v6: Rebase on the latest net-next v5: Eric pointed out that checking skb->len is still needed in tcp_fastopen_add_skb() because skb can carry a FIN without data. Hence, instead of open coding segs_in and data_segs_in, tcp_segs_in() helper is used. Comment is added to the fastopen case to explain why segs_in has to be reset and tcp_segs_in() has to be called before __skb_pull(). v4: Add comment to the changes in tcp_fastopen_add_skb() and also add remark on this case in the commit message. v3: Add const modifier to the skb parameter in tcp_segs_in() v2: Rework based on recent fix by Eric: commit a9d99ce28ed3 ("tcp: fix tcpi_segs_in after connection establishment") Signed-off-by: Martin KaFai Lau <kafai@fb.com> Cc: Chris Rapier <rapier@psc.edu> Cc: Eric Dumazet <edumazet@google.com> Cc: Marcelo Ricardo Leitner <mleitner@redhat.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2016-03-15 01:52:15 +08:00
info->tcpi_data_segs_out = tp->data_segs_out;
info->tcpi_delivery_rate_app_limited = tp->rate_app_limited ? 1 : 0;
rate64 = tcp_compute_delivery_rate(tp);
if (rate64)
info->tcpi_delivery_rate = rate64;
info->tcpi_delivered = tp->delivered;
info->tcpi_delivered_ce = tp->delivered_ce;
info->tcpi_bytes_sent = tp->bytes_sent;
info->tcpi_bytes_retrans = tp->bytes_retrans;
info->tcpi_dsack_dups = tp->dsack_dups;
info->tcpi_reord_seen = tp->reord_seen;
info->tcpi_rcv_ooopack = tp->rcv_ooopack;
info->tcpi_snd_wnd = tp->snd_wnd;
info->tcpi_rcv_wnd = tp->rcv_wnd;
info->tcpi_rehash = tp->plb_rehash + tp->timeout_rehash;
tcp: add TCP_INFO status for failed client TFO The TCPI_OPT_SYN_DATA bit as part of tcpi_options currently reports whether or not data-in-SYN was ack'd on both the client and server side. We'd like to gather more information on the client-side in the failure case in order to indicate the reason for the failure. This can be useful for not only debugging TFO, but also for creating TFO socket policies. For example, if a middle box removes the TFO option or drops a data-in-SYN, we can can detect this case, and turn off TFO for these connections saving the extra retransmits. The newly added tcpi_fastopen_client_fail status is 2 bits and has the following 4 states: 1) TFO_STATUS_UNSPEC Catch-all state which includes when TFO is disabled via black hole detection, which is indicated via LINUX_MIB_TCPFASTOPENBLACKHOLE. 2) TFO_COOKIE_UNAVAILABLE If TFO_CLIENT_NO_COOKIE mode is off, this state indicates that no cookie is available in the cache. 3) TFO_DATA_NOT_ACKED Data was sent with SYN, we received a SYN/ACK but it did not cover the data portion. Cookie is not accepted by server because the cookie may be invalid or the server may be overloaded. 4) TFO_SYN_RETRANSMITTED Data was sent with SYN, we received a SYN/ACK which did not cover the data after at least 1 additional SYN was sent (without data). It may be the case that a middle-box is dropping data-in-SYN packets. Thus, it would be more efficient to not use TFO on this connection to avoid extra retransmits during connection establishment. These new fields do not cover all the cases where TFO may fail, but other failures, such as SYN/ACK + data being dropped, will result in the connection not becoming established. And a connection blackhole after session establishment shows up as a stalled connection. Signed-off-by: Jason Baron <jbaron@akamai.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Christoph Paasch <cpaasch@apple.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2019-10-23 23:09:26 +08:00
info->tcpi_fastopen_client_fail = tp->fastopen_client_fail;
info->tcpi_total_rto = tp->total_rto;
info->tcpi_total_rto_recoveries = tp->total_rto_recoveries;
info->tcpi_total_rto_time = tp->total_rto_time;
if (tp->rto_stamp)
info->tcpi_total_rto_time += tcp_clock_ms() - tp->rto_stamp;
unlock_sock_fast(sk, slow);
}
EXPORT_SYMBOL_GPL(tcp_get_info);
static size_t tcp_opt_stats_get_size(void)
{
return
nla_total_size_64bit(sizeof(u64)) + /* TCP_NLA_BUSY */
nla_total_size_64bit(sizeof(u64)) + /* TCP_NLA_RWND_LIMITED */
nla_total_size_64bit(sizeof(u64)) + /* TCP_NLA_SNDBUF_LIMITED */
nla_total_size_64bit(sizeof(u64)) + /* TCP_NLA_DATA_SEGS_OUT */
nla_total_size_64bit(sizeof(u64)) + /* TCP_NLA_TOTAL_RETRANS */
nla_total_size_64bit(sizeof(u64)) + /* TCP_NLA_PACING_RATE */
nla_total_size_64bit(sizeof(u64)) + /* TCP_NLA_DELIVERY_RATE */
nla_total_size(sizeof(u32)) + /* TCP_NLA_SND_CWND */
nla_total_size(sizeof(u32)) + /* TCP_NLA_REORDERING */
nla_total_size(sizeof(u32)) + /* TCP_NLA_MIN_RTT */
nla_total_size(sizeof(u8)) + /* TCP_NLA_RECUR_RETRANS */
nla_total_size(sizeof(u8)) + /* TCP_NLA_DELIVERY_RATE_APP_LMT */
nla_total_size(sizeof(u32)) + /* TCP_NLA_SNDQ_SIZE */
nla_total_size(sizeof(u8)) + /* TCP_NLA_CA_STATE */
nla_total_size(sizeof(u32)) + /* TCP_NLA_SND_SSTHRESH */
nla_total_size(sizeof(u32)) + /* TCP_NLA_DELIVERED */
nla_total_size(sizeof(u32)) + /* TCP_NLA_DELIVERED_CE */
nla_total_size_64bit(sizeof(u64)) + /* TCP_NLA_BYTES_SENT */
nla_total_size_64bit(sizeof(u64)) + /* TCP_NLA_BYTES_RETRANS */
nla_total_size(sizeof(u32)) + /* TCP_NLA_DSACK_DUPS */
nla_total_size(sizeof(u32)) + /* TCP_NLA_REORD_SEEN */
nla_total_size(sizeof(u32)) + /* TCP_NLA_SRTT */
nla_total_size(sizeof(u16)) + /* TCP_NLA_TIMEOUT_REHASH */
nla_total_size(sizeof(u32)) + /* TCP_NLA_BYTES_NOTSENT */
nla_total_size_64bit(sizeof(u64)) + /* TCP_NLA_EDT */
nla_total_size(sizeof(u8)) + /* TCP_NLA_TTL */
nla_total_size(sizeof(u32)) + /* TCP_NLA_REHASH */
0;
}
/* Returns TTL or hop limit of an incoming packet from skb. */
static u8 tcp_skb_ttl_or_hop_limit(const struct sk_buff *skb)
{
if (skb->protocol == htons(ETH_P_IP))
return ip_hdr(skb)->ttl;
else if (skb->protocol == htons(ETH_P_IPV6))
return ipv6_hdr(skb)->hop_limit;
else
return 0;
}
struct sk_buff *tcp_get_timestamping_opt_stats(const struct sock *sk,
const struct sk_buff *orig_skb,
const struct sk_buff *ack_skb)
{
const struct tcp_sock *tp = tcp_sk(sk);
struct sk_buff *stats;
struct tcp_info info;
unsigned long rate;
u64 rate64;
stats = alloc_skb(tcp_opt_stats_get_size(), GFP_ATOMIC);
if (!stats)
return NULL;
tcp_get_info_chrono_stats(tp, &info);
nla_put_u64_64bit(stats, TCP_NLA_BUSY,
info.tcpi_busy_time, TCP_NLA_PAD);
nla_put_u64_64bit(stats, TCP_NLA_RWND_LIMITED,
info.tcpi_rwnd_limited, TCP_NLA_PAD);
nla_put_u64_64bit(stats, TCP_NLA_SNDBUF_LIMITED,
info.tcpi_sndbuf_limited, TCP_NLA_PAD);
nla_put_u64_64bit(stats, TCP_NLA_DATA_SEGS_OUT,
tp->data_segs_out, TCP_NLA_PAD);
nla_put_u64_64bit(stats, TCP_NLA_TOTAL_RETRANS,
tp->total_retrans, TCP_NLA_PAD);
rate = READ_ONCE(sk->sk_pacing_rate);
rate64 = (rate != ~0UL) ? rate : ~0ULL;
nla_put_u64_64bit(stats, TCP_NLA_PACING_RATE, rate64, TCP_NLA_PAD);
rate64 = tcp_compute_delivery_rate(tp);
nla_put_u64_64bit(stats, TCP_NLA_DELIVERY_RATE, rate64, TCP_NLA_PAD);
nla_put_u32(stats, TCP_NLA_SND_CWND, tcp_snd_cwnd(tp));
nla_put_u32(stats, TCP_NLA_REORDERING, tp->reordering);
nla_put_u32(stats, TCP_NLA_MIN_RTT, tcp_min_rtt(tp));
nla_put_u8(stats, TCP_NLA_RECUR_RETRANS, inet_csk(sk)->icsk_retransmits);
nla_put_u8(stats, TCP_NLA_DELIVERY_RATE_APP_LMT, !!tp->rate_app_limited);
nla_put_u32(stats, TCP_NLA_SND_SSTHRESH, tp->snd_ssthresh);
nla_put_u32(stats, TCP_NLA_DELIVERED, tp->delivered);
nla_put_u32(stats, TCP_NLA_DELIVERED_CE, tp->delivered_ce);
nla_put_u32(stats, TCP_NLA_SNDQ_SIZE, tp->write_seq - tp->snd_una);
nla_put_u8(stats, TCP_NLA_CA_STATE, inet_csk(sk)->icsk_ca_state);
nla_put_u64_64bit(stats, TCP_NLA_BYTES_SENT, tp->bytes_sent,
TCP_NLA_PAD);
nla_put_u64_64bit(stats, TCP_NLA_BYTES_RETRANS, tp->bytes_retrans,
TCP_NLA_PAD);
nla_put_u32(stats, TCP_NLA_DSACK_DUPS, tp->dsack_dups);
nla_put_u32(stats, TCP_NLA_REORD_SEEN, tp->reord_seen);
nla_put_u32(stats, TCP_NLA_SRTT, tp->srtt_us >> 3);
nla_put_u16(stats, TCP_NLA_TIMEOUT_REHASH, tp->timeout_rehash);
nla_put_u32(stats, TCP_NLA_BYTES_NOTSENT,
max_t(int, 0, tp->write_seq - tp->snd_nxt));
nla_put_u64_64bit(stats, TCP_NLA_EDT, orig_skb->skb_mstamp_ns,
TCP_NLA_PAD);
if (ack_skb)
nla_put_u8(stats, TCP_NLA_TTL,
tcp_skb_ttl_or_hop_limit(ack_skb));
nla_put_u32(stats, TCP_NLA_REHASH, tp->plb_rehash + tp->timeout_rehash);
return stats;
}
int do_tcp_getsockopt(struct sock *sk, int level,
int optname, sockptr_t optval, sockptr_t optlen)
{
struct inet_connection_sock *icsk = inet_csk(sk);
struct tcp_sock *tp = tcp_sk(sk);
struct net *net = sock_net(sk);
int val, len;
if (copy_from_sockptr(&len, optlen, sizeof(int)))
return -EFAULT;
len = min_t(unsigned int, len, sizeof(int));
if (len < 0)
return -EINVAL;
switch (optname) {
case TCP_MAXSEG:
val = tp->mss_cache;
if (tp->rx_opt.user_mss &&
((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN)))
val = tp->rx_opt.user_mss;
if (tp->repair)
val = tp->rx_opt.mss_clamp;
break;
case TCP_NODELAY:
val = !!(tp->nonagle&TCP_NAGLE_OFF);
break;
case TCP_CORK:
val = !!(tp->nonagle&TCP_NAGLE_CORK);
break;
case TCP_KEEPIDLE:
val = keepalive_time_when(tp) / HZ;
break;
case TCP_KEEPINTVL:
val = keepalive_intvl_when(tp) / HZ;
break;
case TCP_KEEPCNT:
val = keepalive_probes(tp);
break;
case TCP_SYNCNT:
val = READ_ONCE(icsk->icsk_syn_retries) ? :
READ_ONCE(net->ipv4.sysctl_tcp_syn_retries);
break;
case TCP_LINGER2:
val = READ_ONCE(tp->linger2);
if (val >= 0)
val = (val ? : READ_ONCE(net->ipv4.sysctl_tcp_fin_timeout)) / HZ;
break;
case TCP_DEFER_ACCEPT:
val = READ_ONCE(icsk->icsk_accept_queue.rskq_defer_accept);
val = retrans_to_secs(val, TCP_TIMEOUT_INIT / HZ,
TCP_RTO_MAX / HZ);
break;
case TCP_WINDOW_CLAMP:
val = tp->window_clamp;
break;
case TCP_INFO: {
struct tcp_info info;
if (copy_from_sockptr(&len, optlen, sizeof(int)))
return -EFAULT;
tcp_get_info(sk, &info);
len = min_t(unsigned int, len, sizeof(info));
if (copy_to_sockptr(optlen, &len, sizeof(int)))
return -EFAULT;
if (copy_to_sockptr(optval, &info, len))
return -EFAULT;
return 0;
}
case TCP_CC_INFO: {
const struct tcp_congestion_ops *ca_ops;
union tcp_cc_info info;
size_t sz = 0;
int attr;
if (copy_from_sockptr(&len, optlen, sizeof(int)))
return -EFAULT;
ca_ops = icsk->icsk_ca_ops;
if (ca_ops && ca_ops->get_info)
sz = ca_ops->get_info(sk, ~0U, &attr, &info);
len = min_t(unsigned int, len, sz);
if (copy_to_sockptr(optlen, &len, sizeof(int)))
return -EFAULT;
if (copy_to_sockptr(optval, &info, len))
return -EFAULT;
return 0;
}
case TCP_QUICKACK:
val = !inet_csk_in_pingpong_mode(sk);
break;
case TCP_CONGESTION:
if (copy_from_sockptr(&len, optlen, sizeof(int)))
return -EFAULT;
len = min_t(unsigned int, len, TCP_CA_NAME_MAX);
if (copy_to_sockptr(optlen, &len, sizeof(int)))
return -EFAULT;
if (copy_to_sockptr(optval, icsk->icsk_ca_ops->name, len))
return -EFAULT;
return 0;
case TCP_ULP:
if (copy_from_sockptr(&len, optlen, sizeof(int)))
return -EFAULT;
len = min_t(unsigned int, len, TCP_ULP_NAME_MAX);
if (!icsk->icsk_ulp_ops) {
len = 0;
if (copy_to_sockptr(optlen, &len, sizeof(int)))
return -EFAULT;
return 0;
}
if (copy_to_sockptr(optlen, &len, sizeof(int)))
return -EFAULT;
if (copy_to_sockptr(optval, icsk->icsk_ulp_ops->name, len))
return -EFAULT;
return 0;
case TCP_FASTOPEN_KEY: {
u64 key[TCP_FASTOPEN_KEY_BUF_LENGTH / sizeof(u64)];
unsigned int key_len;
if (copy_from_sockptr(&len, optlen, sizeof(int)))
return -EFAULT;
key_len = tcp_fastopen_get_cipher(net, icsk, key) *
TCP_FASTOPEN_KEY_LENGTH;
len = min_t(unsigned int, len, key_len);
if (copy_to_sockptr(optlen, &len, sizeof(int)))
return -EFAULT;
if (copy_to_sockptr(optval, key, len))
return -EFAULT;
return 0;
}
case TCP_THIN_LINEAR_TIMEOUTS:
val = tp->thin_lto;
break;
case TCP_THIN_DUPACK:
val = 0;
break;
tcp: Add TCP_USER_TIMEOUT socket option. This patch provides a "user timeout" support as described in RFC793. The socket option is also needed for the the local half of RFC5482 "TCP User Timeout Option". TCP_USER_TIMEOUT is a TCP level socket option that takes an unsigned int, when > 0, to specify the maximum amount of time in ms that transmitted data may remain unacknowledged before TCP will forcefully close the corresponding connection and return ETIMEDOUT to the application. If 0 is given, TCP will continue to use the system default. Increasing the user timeouts allows a TCP connection to survive extended periods without end-to-end connectivity. Decreasing the user timeouts allows applications to "fail fast" if so desired. Otherwise it may take upto 20 minutes with the current system defaults in a normal WAN environment. The socket option can be made during any state of a TCP connection, but is only effective during the synchronized states of a connection (ESTABLISHED, FIN-WAIT-1, FIN-WAIT-2, CLOSE-WAIT, CLOSING, or LAST-ACK). Moreover, when used with the TCP keepalive (SO_KEEPALIVE) option, TCP_USER_TIMEOUT will overtake keepalive to determine when to close a connection due to keepalive failure. The option does not change in anyway when TCP retransmits a packet, nor when a keepalive probe will be sent. This option, like many others, will be inherited by an acceptor from its listener. Signed-off-by: H.K. Jerry Chu <hkchu@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-28 03:13:28 +08:00
case TCP_REPAIR:
val = tp->repair;
break;
case TCP_REPAIR_QUEUE:
if (tp->repair)
val = tp->repair_queue;
else
return -EINVAL;
break;
case TCP_REPAIR_WINDOW: {
struct tcp_repair_window opt;
if (copy_from_sockptr(&len, optlen, sizeof(int)))
return -EFAULT;
if (len != sizeof(opt))
return -EINVAL;
if (!tp->repair)
return -EPERM;
opt.snd_wl1 = tp->snd_wl1;
opt.snd_wnd = tp->snd_wnd;
opt.max_window = tp->max_window;
opt.rcv_wnd = tp->rcv_wnd;
opt.rcv_wup = tp->rcv_wup;
if (copy_to_sockptr(optval, &opt, len))
return -EFAULT;
return 0;
}
case TCP_QUEUE_SEQ:
if (tp->repair_queue == TCP_SEND_QUEUE)
val = tp->write_seq;
else if (tp->repair_queue == TCP_RECV_QUEUE)
val = tp->rcv_nxt;
else
return -EINVAL;
break;
tcp: Add TCP_USER_TIMEOUT socket option. This patch provides a "user timeout" support as described in RFC793. The socket option is also needed for the the local half of RFC5482 "TCP User Timeout Option". TCP_USER_TIMEOUT is a TCP level socket option that takes an unsigned int, when > 0, to specify the maximum amount of time in ms that transmitted data may remain unacknowledged before TCP will forcefully close the corresponding connection and return ETIMEDOUT to the application. If 0 is given, TCP will continue to use the system default. Increasing the user timeouts allows a TCP connection to survive extended periods without end-to-end connectivity. Decreasing the user timeouts allows applications to "fail fast" if so desired. Otherwise it may take upto 20 minutes with the current system defaults in a normal WAN environment. The socket option can be made during any state of a TCP connection, but is only effective during the synchronized states of a connection (ESTABLISHED, FIN-WAIT-1, FIN-WAIT-2, CLOSE-WAIT, CLOSING, or LAST-ACK). Moreover, when used with the TCP keepalive (SO_KEEPALIVE) option, TCP_USER_TIMEOUT will overtake keepalive to determine when to close a connection due to keepalive failure. The option does not change in anyway when TCP retransmits a packet, nor when a keepalive probe will be sent. This option, like many others, will be inherited by an acceptor from its listener. Signed-off-by: H.K. Jerry Chu <hkchu@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-28 03:13:28 +08:00
case TCP_USER_TIMEOUT:
val = READ_ONCE(icsk->icsk_user_timeout);
tcp: Add TCP_USER_TIMEOUT socket option. This patch provides a "user timeout" support as described in RFC793. The socket option is also needed for the the local half of RFC5482 "TCP User Timeout Option". TCP_USER_TIMEOUT is a TCP level socket option that takes an unsigned int, when > 0, to specify the maximum amount of time in ms that transmitted data may remain unacknowledged before TCP will forcefully close the corresponding connection and return ETIMEDOUT to the application. If 0 is given, TCP will continue to use the system default. Increasing the user timeouts allows a TCP connection to survive extended periods without end-to-end connectivity. Decreasing the user timeouts allows applications to "fail fast" if so desired. Otherwise it may take upto 20 minutes with the current system defaults in a normal WAN environment. The socket option can be made during any state of a TCP connection, but is only effective during the synchronized states of a connection (ESTABLISHED, FIN-WAIT-1, FIN-WAIT-2, CLOSE-WAIT, CLOSING, or LAST-ACK). Moreover, when used with the TCP keepalive (SO_KEEPALIVE) option, TCP_USER_TIMEOUT will overtake keepalive to determine when to close a connection due to keepalive failure. The option does not change in anyway when TCP retransmits a packet, nor when a keepalive probe will be sent. This option, like many others, will be inherited by an acceptor from its listener. Signed-off-by: H.K. Jerry Chu <hkchu@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-28 03:13:28 +08:00
break;
case TCP_FASTOPEN:
val = READ_ONCE(icsk->icsk_accept_queue.fastopenq.max_qlen);
break;
net/tcp-fastopen: Add new API support This patch adds a new socket option, TCP_FASTOPEN_CONNECT, as an alternative way to perform Fast Open on the active side (client). Prior to this patch, a client needs to replace the connect() call with sendto(MSG_FASTOPEN). This can be cumbersome for applications who want to use Fast Open: these socket operations are often done in lower layer libraries used by many other applications. Changing these libraries and/or the socket call sequences are not trivial. A more convenient approach is to perform Fast Open by simply enabling a socket option when the socket is created w/o changing other socket calls sequence: s = socket() create a new socket setsockopt(s, IPPROTO_TCP, TCP_FASTOPEN_CONNECT …); newly introduced sockopt If set, new functionality described below will be used. Return ENOTSUPP if TFO is not supported or not enabled in the kernel. connect() With cookie present, return 0 immediately. With no cookie, initiate 3WHS with TFO cookie-request option and return -1 with errno = EINPROGRESS. write()/sendmsg() With cookie present, send out SYN with data and return the number of bytes buffered. With no cookie, and 3WHS not yet completed, return -1 with errno = EINPROGRESS. No MSG_FASTOPEN flag is needed. read() Return -1 with errno = EWOULDBLOCK/EAGAIN if connect() is called but write() is not called yet. Return -1 with errno = EWOULDBLOCK/EAGAIN if connection is established but no msg is received yet. Return number of bytes read if socket is established and there is msg received. The new API simplifies life for applications that always perform a write() immediately after a successful connect(). Such applications can now take advantage of Fast Open by merely making one new setsockopt() call at the time of creating the socket. Nothing else about the application's socket call sequence needs to change. Signed-off-by: Wei Wang <weiwan@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-01-24 02:59:22 +08:00
case TCP_FASTOPEN_CONNECT:
val = tp->fastopen_connect;
break;
case TCP_FASTOPEN_NO_COOKIE:
val = tp->fastopen_no_cookie;
break;
tcp: add optional per socket transmit delay Adding delays to TCP flows is crucial for studying behavior of TCP stacks, including congestion control modules. Linux offers netem module, but it has unpractical constraints : - Need root access to change qdisc - Hard to setup on egress if combined with non trivial qdisc like FQ - Single delay for all flows. EDT (Earliest Departure Time) adoption in TCP stack allows us to enable a per socket delay at a very small cost. Networking tools can now establish thousands of flows, each of them with a different delay, simulating real world conditions. This requires FQ packet scheduler or a EDT-enabled NIC. This patchs adds TCP_TX_DELAY socket option, to set a delay in usec units. unsigned int tx_delay = 10000; /* 10 msec */ setsockopt(fd, SOL_TCP, TCP_TX_DELAY, &tx_delay, sizeof(tx_delay)); Note that FQ packet scheduler limits might need some tweaking : man tc-fq PARAMETERS limit Hard limit on the real queue size. When this limit is reached, new packets are dropped. If the value is lowered, packets are dropped so that the new limit is met. Default is 10000 packets. flow_limit Hard limit on the maximum number of packets queued per flow. Default value is 100. Use of TCP_TX_DELAY option will increase number of skbs in FQ qdisc, so packets would be dropped if any of the previous limit is hit. Use of a jump label makes this support runtime-free, for hosts never using the option. Also note that TSQ (TCP Small Queues) limits are slightly changed with this patch : we need to account that skbs artificially delayed wont stop us providind more skbs to feed the pipe (netem uses skb_orphan_partial() for this purpose, but FQ can not use this trick) Because of that, using big delays might very well trigger old bugs in TSO auto defer logic and/or sndbuf limited detection. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2019-06-13 02:57:25 +08:00
case TCP_TX_DELAY:
val = READ_ONCE(tp->tcp_tx_delay);
tcp: add optional per socket transmit delay Adding delays to TCP flows is crucial for studying behavior of TCP stacks, including congestion control modules. Linux offers netem module, but it has unpractical constraints : - Need root access to change qdisc - Hard to setup on egress if combined with non trivial qdisc like FQ - Single delay for all flows. EDT (Earliest Departure Time) adoption in TCP stack allows us to enable a per socket delay at a very small cost. Networking tools can now establish thousands of flows, each of them with a different delay, simulating real world conditions. This requires FQ packet scheduler or a EDT-enabled NIC. This patchs adds TCP_TX_DELAY socket option, to set a delay in usec units. unsigned int tx_delay = 10000; /* 10 msec */ setsockopt(fd, SOL_TCP, TCP_TX_DELAY, &tx_delay, sizeof(tx_delay)); Note that FQ packet scheduler limits might need some tweaking : man tc-fq PARAMETERS limit Hard limit on the real queue size. When this limit is reached, new packets are dropped. If the value is lowered, packets are dropped so that the new limit is met. Default is 10000 packets. flow_limit Hard limit on the maximum number of packets queued per flow. Default value is 100. Use of TCP_TX_DELAY option will increase number of skbs in FQ qdisc, so packets would be dropped if any of the previous limit is hit. Use of a jump label makes this support runtime-free, for hosts never using the option. Also note that TSQ (TCP Small Queues) limits are slightly changed with this patch : we need to account that skbs artificially delayed wont stop us providind more skbs to feed the pipe (netem uses skb_orphan_partial() for this purpose, but FQ can not use this trick) Because of that, using big delays might very well trigger old bugs in TSO auto defer logic and/or sndbuf limited detection. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2019-06-13 02:57:25 +08:00
break;
case TCP_TIMESTAMP:
val = tcp_clock_ts(tp->tcp_usec_ts) + READ_ONCE(tp->tsoffset);
if (tp->tcp_usec_ts)
val |= 1;
else
val &= ~1;
break;
tcp: TCP_NOTSENT_LOWAT socket option Idea of this patch is to add optional limitation of number of unsent bytes in TCP sockets, to reduce usage of kernel memory. TCP receiver might announce a big window, and TCP sender autotuning might allow a large amount of bytes in write queue, but this has little performance impact if a large part of this buffering is wasted : Write queue needs to be large only to deal with large BDP, not necessarily to cope with scheduling delays (incoming ACKS make room for the application to queue more bytes) For most workloads, using a value of 128 KB or less is OK to give applications enough time to react to POLLOUT events in time (or being awaken in a blocking sendmsg()) This patch adds two ways to set the limit : 1) Per socket option TCP_NOTSENT_LOWAT 2) A sysctl (/proc/sys/net/ipv4/tcp_notsent_lowat) for sockets not using TCP_NOTSENT_LOWAT socket option (or setting a zero value) Default value being UINT_MAX (0xFFFFFFFF), meaning this has no effect. This changes poll()/select()/epoll() to report POLLOUT only if number of unsent bytes is below tp->nosent_lowat Note this might increase number of sendmsg()/sendfile() calls when using non blocking sockets, and increase number of context switches for blocking sockets. Note this is not related to SO_SNDLOWAT (as SO_SNDLOWAT is defined as : Specify the minimum number of bytes in the buffer until the socket layer will pass the data to the protocol) Tested: netperf sessions, and watching /proc/net/protocols "memory" column for TCP With 200 concurrent netperf -t TCP_STREAM sessions, amount of kernel memory used by TCP buffers shrinks by ~55 % (20567 pages instead of 45458) lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols TCPv6 1880 2 45458 no 208 yes ipv6 y y y y y y y y y y y y y n y y y y y TCP 1696 508 45458 no 208 yes kernel y y y y y y y y y y y y y n y y y y y lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols TCPv6 1880 2 20567 no 208 yes ipv6 y y y y y y y y y y y y y n y y y y y TCP 1696 508 20567 no 208 yes kernel y y y y y y y y y y y y y n y y y y y Using 128KB has no bad effect on the throughput or cpu usage of a single flow, although there is an increase of context switches. A bonus is that we hold socket lock for a shorter amount of time and should improve latencies of ACK processing. lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3 OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf. Local Remote Local Elapsed Throughput Throughput Local Local Remote Remote Local Remote Service Send Socket Recv Socket Send Time Units CPU CPU CPU CPU Service Service Demand Size Size Size (sec) Util Util Util Util Demand Demand Units Final Final % Method % Method 1651584 6291456 16384 20.00 17447.90 10^6bits/s 3.13 S -1.00 U 0.353 -1.000 usec/KB Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3': 412,514 context-switches 200.034645535 seconds time elapsed lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3 OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf. Local Remote Local Elapsed Throughput Throughput Local Local Remote Remote Local Remote Service Send Socket Recv Socket Send Time Units CPU CPU CPU CPU Service Service Demand Size Size Size (sec) Util Util Util Util Demand Demand Units Final Final % Method % Method 1593240 6291456 16384 20.00 17321.16 10^6bits/s 3.35 S -1.00 U 0.381 -1.000 usec/KB Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3': 2,675,818 context-switches 200.029651391 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-By: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-07-23 11:27:07 +08:00
case TCP_NOTSENT_LOWAT:
val = READ_ONCE(tp->notsent_lowat);
tcp: TCP_NOTSENT_LOWAT socket option Idea of this patch is to add optional limitation of number of unsent bytes in TCP sockets, to reduce usage of kernel memory. TCP receiver might announce a big window, and TCP sender autotuning might allow a large amount of bytes in write queue, but this has little performance impact if a large part of this buffering is wasted : Write queue needs to be large only to deal with large BDP, not necessarily to cope with scheduling delays (incoming ACKS make room for the application to queue more bytes) For most workloads, using a value of 128 KB or less is OK to give applications enough time to react to POLLOUT events in time (or being awaken in a blocking sendmsg()) This patch adds two ways to set the limit : 1) Per socket option TCP_NOTSENT_LOWAT 2) A sysctl (/proc/sys/net/ipv4/tcp_notsent_lowat) for sockets not using TCP_NOTSENT_LOWAT socket option (or setting a zero value) Default value being UINT_MAX (0xFFFFFFFF), meaning this has no effect. This changes poll()/select()/epoll() to report POLLOUT only if number of unsent bytes is below tp->nosent_lowat Note this might increase number of sendmsg()/sendfile() calls when using non blocking sockets, and increase number of context switches for blocking sockets. Note this is not related to SO_SNDLOWAT (as SO_SNDLOWAT is defined as : Specify the minimum number of bytes in the buffer until the socket layer will pass the data to the protocol) Tested: netperf sessions, and watching /proc/net/protocols "memory" column for TCP With 200 concurrent netperf -t TCP_STREAM sessions, amount of kernel memory used by TCP buffers shrinks by ~55 % (20567 pages instead of 45458) lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols TCPv6 1880 2 45458 no 208 yes ipv6 y y y y y y y y y y y y y n y y y y y TCP 1696 508 45458 no 208 yes kernel y y y y y y y y y y y y y n y y y y y lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols TCPv6 1880 2 20567 no 208 yes ipv6 y y y y y y y y y y y y y n y y y y y TCP 1696 508 20567 no 208 yes kernel y y y y y y y y y y y y y n y y y y y Using 128KB has no bad effect on the throughput or cpu usage of a single flow, although there is an increase of context switches. A bonus is that we hold socket lock for a shorter amount of time and should improve latencies of ACK processing. lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3 OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf. Local Remote Local Elapsed Throughput Throughput Local Local Remote Remote Local Remote Service Send Socket Recv Socket Send Time Units CPU CPU CPU CPU Service Service Demand Size Size Size (sec) Util Util Util Util Demand Demand Units Final Final % Method % Method 1651584 6291456 16384 20.00 17447.90 10^6bits/s 3.13 S -1.00 U 0.353 -1.000 usec/KB Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3': 412,514 context-switches 200.034645535 seconds time elapsed lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3 OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf. Local Remote Local Elapsed Throughput Throughput Local Local Remote Remote Local Remote Service Send Socket Recv Socket Send Time Units CPU CPU CPU CPU Service Service Demand Size Size Size (sec) Util Util Util Util Demand Demand Units Final Final % Method % Method 1593240 6291456 16384 20.00 17321.16 10^6bits/s 3.35 S -1.00 U 0.381 -1.000 usec/KB Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3': 2,675,818 context-switches 200.029651391 seconds time elapsed Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-By: Yuchung Cheng <ycheng@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-07-23 11:27:07 +08:00
break;
tcp: send in-queue bytes in cmsg upon read Applications with many concurrent connections, high variance in receive queue length and tight memory bounds cannot allocate worst-case buffer size to drain sockets. Knowing the size of receive queue length, applications can optimize how they allocate buffers to read from the socket. The number of bytes pending on the socket is directly available through ioctl(FIONREAD/SIOCINQ) and can be approximated using getsockopt(MEMINFO) (rmem_alloc includes skb overheads in addition to application data). But, both of these options add an extra syscall per recvmsg. Moreover, ioctl(FIONREAD/SIOCINQ) takes the socket lock. Add the TCP_INQ socket option to TCP. When this socket option is set, recvmsg() relays the number of bytes available on the socket for reading to the application via the TCP_CM_INQ control message. Calculate the number of bytes after releasing the socket lock to include the processed backlog, if any. To avoid an extra branch in the hot path of recvmsg() for this new control message, move all cmsg processing inside an existing branch for processing receive timestamps. Since the socket lock is not held when calculating the size of receive queue, TCP_INQ is a hint. For example, it can overestimate the queue size by one byte, if FIN is received. With this method, applications can start reading from the socket using a small buffer, and then use larger buffers based on the remaining data when needed. V3 change-log: As suggested by David Miller, added loads with barrier to check whether we have multiple threads calling recvmsg in parallel. When that happens we lock the socket to calculate inq. V4 change-log: Removed inline from a static function. Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Willem de Bruijn <willemb@google.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Neal Cardwell <ncardwell@google.com> Suggested-by: David Miller <davem@davemloft.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-05-02 03:39:15 +08:00
case TCP_INQ:
val = tp->recvmsg_inq;
break;
case TCP_SAVE_SYN:
val = tp->save_syn;
break;
case TCP_SAVED_SYN: {
if (copy_from_sockptr(&len, optlen, sizeof(int)))
return -EFAULT;
sockopt_lock_sock(sk);
if (tp->saved_syn) {
if (len < tcp_saved_syn_len(tp->saved_syn)) {
len = tcp_saved_syn_len(tp->saved_syn);
if (copy_to_sockptr(optlen, &len, sizeof(int))) {
sockopt_release_sock(sk);
return -EFAULT;
}
sockopt_release_sock(sk);
return -EINVAL;
}
len = tcp_saved_syn_len(tp->saved_syn);
if (copy_to_sockptr(optlen, &len, sizeof(int))) {
sockopt_release_sock(sk);
return -EFAULT;
}
if (copy_to_sockptr(optval, tp->saved_syn->data, len)) {
sockopt_release_sock(sk);
return -EFAULT;
}
tcp_saved_syn_free(tp);
sockopt_release_sock(sk);
} else {
sockopt_release_sock(sk);
len = 0;
if (copy_to_sockptr(optlen, &len, sizeof(int)))
return -EFAULT;
}
return 0;
}
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
#ifdef CONFIG_MMU
case TCP_ZEROCOPY_RECEIVE: {
struct scm_timestamping_internal tss;
struct tcp_zerocopy_receive zc = {};
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
int err;
if (copy_from_sockptr(&len, optlen, sizeof(int)))
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
return -EFAULT;
if (len < 0 ||
len < offsetofend(struct tcp_zerocopy_receive, length))
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
return -EINVAL;
if (unlikely(len > sizeof(zc))) {
err = check_zeroed_sockptr(optval, sizeof(zc),
len - sizeof(zc));
if (err < 1)
return err == 0 ? -EINVAL : err;
len = sizeof(zc);
if (copy_to_sockptr(optlen, &len, sizeof(int)))
return -EFAULT;
}
if (copy_from_sockptr(&zc, optval, len))
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
return -EFAULT;
if (zc.reserved)
return -EINVAL;
if (zc.msg_flags & ~(TCP_VALID_ZC_MSG_FLAGS))
return -EINVAL;
sockopt_lock_sock(sk);
err = tcp_zerocopy_receive(sk, &zc, &tss);
err = BPF_CGROUP_RUN_PROG_GETSOCKOPT_KERN(sk, level, optname,
&zc, &len, err);
sockopt_release_sock(sk);
if (len >= offsetofend(struct tcp_zerocopy_receive, msg_flags))
goto zerocopy_rcv_cmsg;
switch (len) {
case offsetofend(struct tcp_zerocopy_receive, msg_flags):
goto zerocopy_rcv_cmsg;
case offsetofend(struct tcp_zerocopy_receive, msg_controllen):
case offsetofend(struct tcp_zerocopy_receive, msg_control):
case offsetofend(struct tcp_zerocopy_receive, flags):
case offsetofend(struct tcp_zerocopy_receive, copybuf_len):
case offsetofend(struct tcp_zerocopy_receive, copybuf_address):
case offsetofend(struct tcp_zerocopy_receive, err):
goto zerocopy_rcv_sk_err;
case offsetofend(struct tcp_zerocopy_receive, inq):
goto zerocopy_rcv_inq;
case offsetofend(struct tcp_zerocopy_receive, length):
default:
goto zerocopy_rcv_out;
}
zerocopy_rcv_cmsg:
if (zc.msg_flags & TCP_CMSG_TS)
tcp_zc_finalize_rx_tstamp(sk, &zc, &tss);
else
zc.msg_flags = 0;
zerocopy_rcv_sk_err:
if (!err)
zc.err = sock_error(sk);
zerocopy_rcv_inq:
zc.inq = tcp_inq_hint(sk);
zerocopy_rcv_out:
if (!err && copy_to_sockptr(optval, &zc, len))
tcp: add TCP_ZEROCOPY_RECEIVE support for zerocopy receive When adding tcp mmap() implementation, I forgot that socket lock had to be taken before current->mm->mmap_sem. syzbot eventually caught the bug. Since we can not lock the socket in tcp mmap() handler we have to split the operation in two phases. 1) mmap() on a tcp socket simply reserves VMA space, and nothing else. This operation does not involve any TCP locking. 2) getsockopt(fd, IPPROTO_TCP, TCP_ZEROCOPY_RECEIVE, ...) implements the transfert of pages from skbs to one VMA. This operation only uses down_read(&current->mm->mmap_sem) after holding TCP lock, thus solving the lockdep issue. This new implementation was suggested by Andy Lutomirski with great details. Benefits are : - Better scalability, in case multiple threads reuse VMAS (without mmap()/munmap() calls) since mmap_sem wont be write locked. - Better error recovery. The previous mmap() model had to provide the expected size of the mapping. If for some reason one part could not be mapped (partial MSS), the whole operation had to be aborted. With the tcp_zerocopy_receive struct, kernel can report how many bytes were successfuly mapped, and how many bytes should be read to skip the problematic sequence. - No more memory allocation to hold an array of page pointers. 16 MB mappings needed 32 KB for this array, potentially using vmalloc() :/ - skbs are freed while mmap_sem has been released Following patch makes the change in tcp_mmap tool to demonstrate one possible use of mmap() and setsockopt(... TCP_ZEROCOPY_RECEIVE ...) Note that memcg might require additional changes. Fixes: 93ab6cc69162 ("tcp: implement mmap() for zero copy receive") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Suggested-by: Andy Lutomirski <luto@kernel.org> Cc: linux-mm@kvack.org Acked-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-27 23:58:08 +08:00
err = -EFAULT;
return err;
}
#endif
case TCP_AO_REPAIR:
if (!tcp_can_repair_sock(sk))
return -EPERM;
return tcp_ao_get_repair(sk, optval, optlen);
case TCP_AO_GET_KEYS:
case TCP_AO_INFO: {
int err;
sockopt_lock_sock(sk);
if (optname == TCP_AO_GET_KEYS)
err = tcp_ao_get_mkts(sk, optval, optlen);
else
err = tcp_ao_get_sock_info(sk, optval, optlen);
sockopt_release_sock(sk);
return err;
}
default:
return -ENOPROTOOPT;
}
if (copy_to_sockptr(optlen, &len, sizeof(int)))
return -EFAULT;
if (copy_to_sockptr(optval, &val, len))
return -EFAULT;
return 0;
}
bool tcp_bpf_bypass_getsockopt(int level, int optname)
{
/* TCP do_tcp_getsockopt has optimized getsockopt implementation
* to avoid extra socket lock for TCP_ZEROCOPY_RECEIVE.
*/
if (level == SOL_TCP && optname == TCP_ZEROCOPY_RECEIVE)
return true;
return false;
}
EXPORT_SYMBOL(tcp_bpf_bypass_getsockopt);
int tcp_getsockopt(struct sock *sk, int level, int optname, char __user *optval,
int __user *optlen)
{
struct inet_connection_sock *icsk = inet_csk(sk);
if (level != SOL_TCP)
tcp: Fix data races around icsk->icsk_af_ops. setsockopt(IPV6_ADDRFORM) and tcp_v6_connect() change icsk->icsk_af_ops under lock_sock(), but tcp_(get|set)sockopt() read it locklessly. To avoid load/store tearing, we need to add READ_ONCE() and WRITE_ONCE() for the reads and writes. Thanks to Eric Dumazet for providing the syzbot report: BUG: KCSAN: data-race in tcp_setsockopt / tcp_v6_connect write to 0xffff88813c624518 of 8 bytes by task 23936 on cpu 0: tcp_v6_connect+0x5b3/0xce0 net/ipv6/tcp_ipv6.c:240 __inet_stream_connect+0x159/0x6d0 net/ipv4/af_inet.c:660 inet_stream_connect+0x44/0x70 net/ipv4/af_inet.c:724 __sys_connect_file net/socket.c:1976 [inline] __sys_connect+0x197/0x1b0 net/socket.c:1993 __do_sys_connect net/socket.c:2003 [inline] __se_sys_connect net/socket.c:2000 [inline] __x64_sys_connect+0x3d/0x50 net/socket.c:2000 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x2b/0x70 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd read to 0xffff88813c624518 of 8 bytes by task 23937 on cpu 1: tcp_setsockopt+0x147/0x1c80 net/ipv4/tcp.c:3789 sock_common_setsockopt+0x5d/0x70 net/core/sock.c:3585 __sys_setsockopt+0x212/0x2b0 net/socket.c:2252 __do_sys_setsockopt net/socket.c:2263 [inline] __se_sys_setsockopt net/socket.c:2260 [inline] __x64_sys_setsockopt+0x62/0x70 net/socket.c:2260 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x2b/0x70 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd value changed: 0xffffffff8539af68 -> 0xffffffff8539aff8 Reported by Kernel Concurrency Sanitizer on: CPU: 1 PID: 23937 Comm: syz-executor.5 Not tainted 6.0.0-rc4-syzkaller-00331-g4ed9c1e971b1-dirty #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 08/26/2022 Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot <syzkaller@googlegroups.com> Reported-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2022-10-07 02:53:49 +08:00
/* Paired with WRITE_ONCE() in do_ipv6_setsockopt() and tcp_v6_connect() */
return READ_ONCE(icsk->icsk_af_ops)->getsockopt(sk, level, optname,
optval, optlen);
return do_tcp_getsockopt(sk, level, optname, USER_SOCKPTR(optval),
USER_SOCKPTR(optlen));
}
EXPORT_SYMBOL(tcp_getsockopt);
#ifdef CONFIG_TCP_MD5SIG
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
int tcp_md5_sigpool_id = -1;
EXPORT_SYMBOL_GPL(tcp_md5_sigpool_id);
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
int tcp_md5_alloc_sigpool(void)
{
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
size_t scratch_size;
int ret;
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
scratch_size = sizeof(union tcp_md5sum_block) + sizeof(struct tcphdr);
ret = tcp_sigpool_alloc_ahash("md5", scratch_size);
if (ret >= 0) {
/* As long as any md5 sigpool was allocated, the return
* id would stay the same. Re-write the id only for the case
* when previously all MD5 keys were deleted and this call
* allocates the first MD5 key, which may return a different
* sigpool id than was used previously.
*/
WRITE_ONCE(tcp_md5_sigpool_id, ret); /* Avoids the compiler potentially being smart here */
return 0;
}
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
return ret;
}
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
void tcp_md5_release_sigpool(void)
{
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
tcp_sigpool_release(READ_ONCE(tcp_md5_sigpool_id));
}
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
void tcp_md5_add_sigpool(void)
{
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
tcp_sigpool_get(READ_ONCE(tcp_md5_sigpool_id));
}
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
int tcp_md5_hash_key(struct tcp_sigpool *hp,
const struct tcp_md5sig_key *key)
{
u8 keylen = READ_ONCE(key->keylen); /* paired with WRITE_ONCE() in tcp_md5_do_add */
struct scatterlist sg;
sg_init_one(&sg, key->key, keylen);
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
ahash_request_set_crypt(hp->req, &sg, NULL, keylen);
net/tcp: Prepare tcp_md5sig_pool for TCP-AO TCP-AO, similarly to TCP-MD5, needs to allocate tfms on a slow-path, which is setsockopt() and use crypto ahash requests on fast paths, which are RX/TX softirqs. Also, it needs a temporary/scratch buffer for preparing the hash. Rework tcp_md5sig_pool in order to support other hashing algorithms than MD5. It will make it possible to share pre-allocated crypto_ahash descriptors and scratch area between all TCP hash users. Internally tcp_sigpool calls crypto_clone_ahash() API over pre-allocated crypto ahash tfm. Kudos to Herbert, who provided this new crypto API. I was a little concerned over GFP_ATOMIC allocations of ahash and crypto_request in RX/TX (see tcp_sigpool_start()), so I benchmarked both "backends" with different algorithms, using patched version of iperf3[2]. On my laptop with i7-7600U @ 2.80GHz: clone-tfm per-CPU-requests TCP-MD5 2.25 Gbits/sec 2.30 Gbits/sec TCP-AO(hmac(sha1)) 2.53 Gbits/sec 2.54 Gbits/sec TCP-AO(hmac(sha512)) 1.67 Gbits/sec 1.64 Gbits/sec TCP-AO(hmac(sha384)) 1.77 Gbits/sec 1.80 Gbits/sec TCP-AO(hmac(sha224)) 1.29 Gbits/sec 1.30 Gbits/sec TCP-AO(hmac(sha3-512)) 481 Mbits/sec 480 Mbits/sec TCP-AO(hmac(md5)) 2.07 Gbits/sec 2.12 Gbits/sec TCP-AO(hmac(rmd160)) 1.01 Gbits/sec 995 Mbits/sec TCP-AO(cmac(aes128)) [not supporetd yet] 2.11 Gbits/sec So, it seems that my concerns don't have strong grounds and per-CPU crypto_request allocation can be dropped/removed from tcp_sigpool once ciphers get crypto_clone_ahash() support. [1]: https://lore.kernel.org/all/ZDefxOq6Ax0JeTRH@gondor.apana.org.au/T/#u [2]: https://github.com/0x7f454c46/iperf/tree/tcp-md5-ao Signed-off-by: Dmitry Safonov <dima@arista.com> Reviewed-by: Steen Hegelund <Steen.Hegelund@microchip.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-10-24 03:21:53 +08:00
/* We use data_race() because tcp_md5_do_add() might change
* key->key under us
*/
return data_race(crypto_ahash_update(hp->req));
}
EXPORT_SYMBOL(tcp_md5_hash_key);
/* Called with rcu_read_lock() */
enum skb_drop_reason
tcp_inbound_md5_hash(const struct sock *sk, const struct sk_buff *skb,
const void *saddr, const void *daddr,
int family, int l3index, const __u8 *hash_location)
{
/* This gets called for each TCP segment that has TCP-MD5 option.
* We have 3 drop cases:
* o No MD5 hash and one expected.
* o MD5 hash and we're not expecting one.
* o MD5 hash and its wrong.
*/
const struct tcp_sock *tp = tcp_sk(sk);
struct tcp_md5sig_key *key;
u8 newhash[16];
int genhash;
key = tcp_md5_do_lookup(sk, l3index, saddr, family);
if (!key && hash_location) {
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5UNEXPECTED);
tcp_hash_fail("Unexpected MD5 Hash found", family, skb, "");
return SKB_DROP_REASON_TCP_MD5UNEXPECTED;
}
/* Check the signature.
* To support dual stack listeners, we need to handle
* IPv4-mapped case.
*/
if (family == AF_INET)
genhash = tcp_v4_md5_hash_skb(newhash, key, NULL, skb);
else
genhash = tp->af_specific->calc_md5_hash(newhash, key,
NULL, skb);
if (genhash || memcmp(hash_location, newhash, 16) != 0) {
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5FAILURE);
if (family == AF_INET) {
tcp_hash_fail("MD5 Hash failed", AF_INET, skb, "%s L3 index %d",
genhash ? "tcp_v4_calc_md5_hash failed"
: "", l3index);
} else {
if (genhash) {
tcp_hash_fail("MD5 Hash failed",
AF_INET6, skb, "L3 index %d",
l3index);
} else {
tcp_hash_fail("MD5 Hash mismatch",
AF_INET6, skb, "L3 index %d",
l3index);
}
}
return SKB_DROP_REASON_TCP_MD5FAILURE;
}
return SKB_NOT_DROPPED_YET;
}
EXPORT_SYMBOL(tcp_inbound_md5_hash);
#endif
void tcp_done(struct sock *sk)
{
struct request_sock *req;
/* We might be called with a new socket, after
* inet_csk_prepare_forced_close() has been called
* so we can not use lockdep_sock_is_held(sk)
*/
req = rcu_dereference_protected(tcp_sk(sk)->fastopen_rsk, 1);
if (sk->sk_state == TCP_SYN_SENT || sk->sk_state == TCP_SYN_RECV)
TCP_INC_STATS(sock_net(sk), TCP_MIB_ATTEMPTFAILS);
tcp_set_state(sk, TCP_CLOSE);
tcp_clear_xmit_timers(sk);
if (req)
reqsk_fastopen_remove(sk, req, false);
tcp: add annotations around sk->sk_shutdown accesses Now sk->sk_shutdown is no longer a bitfield, we can add standard READ_ONCE()/WRITE_ONCE() annotations to silence KCSAN reports like the following: BUG: KCSAN: data-race in tcp_disconnect / tcp_poll write to 0xffff88814588582c of 1 bytes by task 3404 on cpu 1: tcp_disconnect+0x4d6/0xdb0 net/ipv4/tcp.c:3121 __inet_stream_connect+0x5dd/0x6e0 net/ipv4/af_inet.c:715 inet_stream_connect+0x48/0x70 net/ipv4/af_inet.c:727 __sys_connect_file net/socket.c:2001 [inline] __sys_connect+0x19b/0x1b0 net/socket.c:2018 __do_sys_connect net/socket.c:2028 [inline] __se_sys_connect net/socket.c:2025 [inline] __x64_sys_connect+0x41/0x50 net/socket.c:2025 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x63/0xcd read to 0xffff88814588582c of 1 bytes by task 3374 on cpu 0: tcp_poll+0x2e6/0x7d0 net/ipv4/tcp.c:562 sock_poll+0x253/0x270 net/socket.c:1383 vfs_poll include/linux/poll.h:88 [inline] io_poll_check_events io_uring/poll.c:281 [inline] io_poll_task_func+0x15a/0x820 io_uring/poll.c:333 handle_tw_list io_uring/io_uring.c:1184 [inline] tctx_task_work+0x1fe/0x4d0 io_uring/io_uring.c:1246 task_work_run+0x123/0x160 kernel/task_work.c:179 get_signal+0xe64/0xff0 kernel/signal.c:2635 arch_do_signal_or_restart+0x89/0x2a0 arch/x86/kernel/signal.c:306 exit_to_user_mode_loop+0x6f/0xe0 kernel/entry/common.c:168 exit_to_user_mode_prepare+0x6c/0xb0 kernel/entry/common.c:204 __syscall_exit_to_user_mode_work kernel/entry/common.c:286 [inline] syscall_exit_to_user_mode+0x26/0x140 kernel/entry/common.c:297 do_syscall_64+0x4d/0xc0 arch/x86/entry/common.c:86 entry_SYSCALL_64_after_hwframe+0x63/0xcd value changed: 0x03 -> 0x00 Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2023-05-10 04:36:56 +08:00
WRITE_ONCE(sk->sk_shutdown, SHUTDOWN_MASK);
if (!sock_flag(sk, SOCK_DEAD))
sk->sk_state_change(sk);
else
inet_csk_destroy_sock(sk);
}
EXPORT_SYMBOL_GPL(tcp_done);
int tcp_abort(struct sock *sk, int err)
{
int state = inet_sk_state_load(sk);
if (state == TCP_NEW_SYN_RECV) {
struct request_sock *req = inet_reqsk(sk);
local_bh_disable();
inet_csk_reqsk_queue_drop(req->rsk_listener, req);
local_bh_enable();
return 0;
}
if (state == TCP_TIME_WAIT) {
struct inet_timewait_sock *tw = inet_twsk(sk);
refcount_inc(&tw->tw_refcnt);
local_bh_disable();
inet_twsk_deschedule_put(tw);
local_bh_enable();
return 0;
}
bpf: Add bpf_sock_destroy kfunc The socket destroy kfunc is used to forcefully terminate sockets from certain BPF contexts. We plan to use the capability in Cilium load-balancing to terminate client sockets that continue to connect to deleted backends. The other use case is on-the-fly policy enforcement where existing socket connections prevented by policies need to be forcefully terminated. The kfunc also allows terminating sockets that may or may not be actively sending traffic. The kfunc can currently be called only from BPF TCP and UDP iterators where users can filter, and terminate selected sockets. More specifically, it can only be called from BPF contexts that ensure socket locking in order to allow synchronous execution of protocol specific `diag_destroy` handlers. The previous commit that batches UDP sockets during iteration facilitated a synchronous invocation of the UDP destroy callback from BPF context by skipping socket locks in `udp_abort`. TCP iterator already supported batching of sockets being iterated. To that end, `tracing_iter_filter` callback filter is added so that verifier can restrict the kfunc to programs with `BPF_TRACE_ITER` attach type, and reject other programs. The kfunc takes `sock_common` type argument, even though it expects, and casts them to a `sock` pointer. This enables the verifier to allow the sock_destroy kfunc to be called for TCP with `sock_common` and UDP with `sock` structs. Furthermore, as `sock_common` only has a subset of certain fields of `sock`, casting pointer to the latter type might not always be safe for certain sockets like request sockets, but these have a special handling in the diag_destroy handlers. Additionally, the kfunc is defined with `KF_TRUSTED_ARGS` flag to avoid the cases where a `PTR_TO_BTF_ID` sk is obtained by following another pointer. eg. getting a sk pointer (may be even NULL) by following another sk pointer. The pointer socket argument passed in TCP and UDP iterators is tagged as `PTR_TRUSTED` in {tcp,udp}_reg_info. The TRUSTED arg changes are contributed by Martin KaFai Lau <martin.lau@kernel.org>. Signed-off-by: Aditi Ghag <aditi.ghag@isovalent.com> Link: https://lore.kernel.org/r/20230519225157.760788-8-aditi.ghag@isovalent.com Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
2023-05-20 06:51:55 +08:00
/* BPF context ensures sock locking. */
if (!has_current_bpf_ctx())
/* Don't race with userspace socket closes such as tcp_close. */
lock_sock(sk);
if (sk->sk_state == TCP_LISTEN) {
tcp_set_state(sk, TCP_CLOSE);
inet_csk_listen_stop(sk);
}
/* Don't race with BH socket closes such as inet_csk_listen_stop. */
local_bh_disable();
bh_lock_sock(sk);
if (!sock_flag(sk, SOCK_DEAD)) {
WRITE_ONCE(sk->sk_err, err);
/* This barrier is coupled with smp_rmb() in tcp_poll() */
smp_wmb();
sk_error_report(sk);
if (tcp_need_reset(sk->sk_state))
tcp_send_active_reset(sk, GFP_ATOMIC);
tcp_done(sk);
}
bh_unlock_sock(sk);
local_bh_enable();
tcp_write_queue_purge(sk);
bpf: Add bpf_sock_destroy kfunc The socket destroy kfunc is used to forcefully terminate sockets from certain BPF contexts. We plan to use the capability in Cilium load-balancing to terminate client sockets that continue to connect to deleted backends. The other use case is on-the-fly policy enforcement where existing socket connections prevented by policies need to be forcefully terminated. The kfunc also allows terminating sockets that may or may not be actively sending traffic. The kfunc can currently be called only from BPF TCP and UDP iterators where users can filter, and terminate selected sockets. More specifically, it can only be called from BPF contexts that ensure socket locking in order to allow synchronous execution of protocol specific `diag_destroy` handlers. The previous commit that batches UDP sockets during iteration facilitated a synchronous invocation of the UDP destroy callback from BPF context by skipping socket locks in `udp_abort`. TCP iterator already supported batching of sockets being iterated. To that end, `tracing_iter_filter` callback filter is added so that verifier can restrict the kfunc to programs with `BPF_TRACE_ITER` attach type, and reject other programs. The kfunc takes `sock_common` type argument, even though it expects, and casts them to a `sock` pointer. This enables the verifier to allow the sock_destroy kfunc to be called for TCP with `sock_common` and UDP with `sock` structs. Furthermore, as `sock_common` only has a subset of certain fields of `sock`, casting pointer to the latter type might not always be safe for certain sockets like request sockets, but these have a special handling in the diag_destroy handlers. Additionally, the kfunc is defined with `KF_TRUSTED_ARGS` flag to avoid the cases where a `PTR_TO_BTF_ID` sk is obtained by following another pointer. eg. getting a sk pointer (may be even NULL) by following another sk pointer. The pointer socket argument passed in TCP and UDP iterators is tagged as `PTR_TRUSTED` in {tcp,udp}_reg_info. The TRUSTED arg changes are contributed by Martin KaFai Lau <martin.lau@kernel.org>. Signed-off-by: Aditi Ghag <aditi.ghag@isovalent.com> Link: https://lore.kernel.org/r/20230519225157.760788-8-aditi.ghag@isovalent.com Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
2023-05-20 06:51:55 +08:00
if (!has_current_bpf_ctx())
release_sock(sk);
return 0;
}
EXPORT_SYMBOL_GPL(tcp_abort);
extern struct tcp_congestion_ops tcp_reno;
static __initdata unsigned long thash_entries;
static int __init set_thash_entries(char *str)
{
ssize_t ret;
if (!str)
return 0;
ret = kstrtoul(str, 0, &thash_entries);
if (ret)
return 0;
return 1;
}
__setup("thash_entries=", set_thash_entries);
static void __init tcp_init_mem(void)
{
unsigned long limit = nr_free_buffer_pages() / 16;
limit = max(limit, 128UL);
sysctl_tcp_mem[0] = limit / 4 * 3; /* 4.68 % */
sysctl_tcp_mem[1] = limit; /* 6.25 % */
sysctl_tcp_mem[2] = sysctl_tcp_mem[0] * 2; /* 9.37 % */
}
static void __init tcp_struct_check(void)
{
/* TX read-mostly hotpath cache lines */
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_tx, max_window);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_tx, rcv_ssthresh);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_tx, reordering);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_tx, notsent_lowat);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_tx, gso_segs);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_tx, lost_skb_hint);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_tx, retransmit_skb_hint);
CACHELINE_ASSERT_GROUP_SIZE(struct tcp_sock, tcp_sock_read_tx, 40);
/* TXRX read-mostly hotpath cache lines */
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_txrx, tsoffset);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_txrx, snd_wnd);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_txrx, mss_cache);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_txrx, snd_cwnd);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_txrx, prr_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_txrx, lost_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_txrx, sacked_out);
CACHELINE_ASSERT_GROUP_SIZE(struct tcp_sock, tcp_sock_read_txrx, 31);
/* RX read-mostly hotpath cache lines */
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, copied_seq);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, rcv_tstamp);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, snd_wl1);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, tlp_high_seq);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, rttvar_us);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, retrans_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, advmss);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, urg_data);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, lost);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, rtt_min);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, out_of_order_queue);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_read_rx, snd_ssthresh);
CACHELINE_ASSERT_GROUP_SIZE(struct tcp_sock, tcp_sock_read_rx, 69);
/* TX read-write hotpath cache lines */
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, segs_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, data_segs_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, bytes_sent);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, snd_sml);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, chrono_start);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, chrono_stat);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, write_seq);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, pushed_seq);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, lsndtime);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, mdev_us);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, tcp_wstamp_ns);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, tcp_clock_cache);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, tcp_mstamp);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, rtt_seq);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, tsorted_sent_queue);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, highest_sack);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_tx, ecn_flags);
CACHELINE_ASSERT_GROUP_SIZE(struct tcp_sock, tcp_sock_write_tx, 113);
/* TXRX read-write hotpath cache lines */
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, pred_flags);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, rcv_nxt);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, snd_nxt);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, snd_una);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, window_clamp);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, srtt_us);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, packets_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, snd_up);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, delivered);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, delivered_ce);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, app_limited);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, rcv_wnd);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, rx_opt);
CACHELINE_ASSERT_GROUP_SIZE(struct tcp_sock, tcp_sock_write_txrx, 76);
/* RX read-write hotpath cache lines */
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, bytes_received);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, segs_in);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, data_segs_in);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rcv_wup);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, max_packets_out);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, cwnd_usage_seq);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rate_delivered);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rate_interval_us);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rcv_rtt_last_tsecr);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, first_tx_mstamp);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, delivered_mstamp);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, bytes_acked);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rcv_rtt_est);
CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, rcvq_space);
CACHELINE_ASSERT_GROUP_SIZE(struct tcp_sock, tcp_sock_write_rx, 99);
}
void __init tcp_init(void)
{
tcp: change tcp_adv_win_scale and tcp_rmem[2] tcp_adv_win_scale default value is 2, meaning we expect a good citizen skb to have skb->len / skb->truesize ratio of 75% (3/4) In 2.6 kernels we (mis)accounted for typical MSS=1460 frame : 1536 + 64 + 256 = 1856 'estimated truesize', and 1856 * 3/4 = 1392. So these skbs were considered as not bloated. With recent truesize fixes, a typical MSS=1460 frame truesize is now the more precise : 2048 + 256 = 2304. But 2304 * 3/4 = 1728. So these skb are not good citizen anymore, because 1460 < 1728 (GRO can escape this problem because it build skbs with a too low truesize.) This also means tcp advertises a too optimistic window for a given allocated rcvspace : When receiving frames, sk_rmem_alloc can hit sk_rcvbuf limit and we call tcp_prune_queue()/tcp_collapse() too often, especially when application is slow to drain its receive queue or in case of losses (netperf is fast, scp is slow). This is a major latency source. We should adjust the len/truesize ratio to 50% instead of 75% This patch : 1) changes tcp_adv_win_scale default to 1 instead of 2 2) increase tcp_rmem[2] limit from 4MB to 6MB to take into account better truesize tracking and to allow autotuning tcp receive window to reach same value than before. Note that same amount of kernel memory is consumed compared to 2.6 kernels. Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Tom Herbert <therbert@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-05-02 10:28:41 +08:00
int max_rshare, max_wshare, cnt;
unsigned long limit;
unsigned int i;
BUILD_BUG_ON(TCP_MIN_SND_MSS <= MAX_TCP_OPTION_SPACE);
BUILD_BUG_ON(sizeof(struct tcp_skb_cb) >
sizeof_field(struct sk_buff, cb));
tcp_struct_check();
percpu_counter_init(&tcp_sockets_allocated, 0, GFP_KERNEL);
tcp: switch orphan_count to bare per-cpu counters Use of percpu_counter structure to track count of orphaned sockets is causing problems on modern hosts with 256 cpus or more. Stefan Bach reported a serious spinlock contention in real workloads, that I was able to reproduce with a netfilter rule dropping incoming FIN packets. 53.56% server [kernel.kallsyms] [k] queued_spin_lock_slowpath | ---queued_spin_lock_slowpath | --53.51%--_raw_spin_lock_irqsave | --53.51%--__percpu_counter_sum tcp_check_oom | |--39.03%--__tcp_close | tcp_close | inet_release | inet6_release | sock_close | __fput | ____fput | task_work_run | exit_to_usermode_loop | do_syscall_64 | entry_SYSCALL_64_after_hwframe | __GI___libc_close | --14.48%--tcp_out_of_resources tcp_write_timeout tcp_retransmit_timer tcp_write_timer_handler tcp_write_timer call_timer_fn expire_timers __run_timers run_timer_softirq __softirqentry_text_start As explained in commit cf86a086a180 ("net/dst: use a smaller percpu_counter batch for dst entries accounting"), default batch size is too big for the default value of tcp_max_orphans (262144). But even if we reduce batch sizes, there would still be cases where the estimated count of orphans is beyond the limit, and where tcp_too_many_orphans() has to call the expensive percpu_counter_sum_positive(). One solution is to use plain per-cpu counters, and have a timer to periodically refresh this cache. Updating this cache every 100ms seems about right, tcp pressure state is not radically changing over shorter periods. percpu_counter was nice 15 years ago while hosts had less than 16 cpus, not anymore by current standards. v2: Fix the build issue for CONFIG_CRYPTO_DEV_CHELSIO_TLS=m, reported by kernel test robot <lkp@intel.com> Remove unused socket argument from tcp_too_many_orphans() Fixes: dd24c00191d5 ("net: Use a percpu_counter for orphan_count") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Stefan Bach <sfb@google.com> Cc: Neal Cardwell <ncardwell@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-10-14 21:41:26 +08:00
timer_setup(&tcp_orphan_timer, tcp_orphan_update, TIMER_DEFERRABLE);
mod_timer(&tcp_orphan_timer, jiffies + TCP_ORPHAN_TIMER_PERIOD);
inet_hashinfo2_init(&tcp_hashinfo, "tcp_listen_portaddr_hash",
thash_entries, 21, /* one slot per 2 MB*/
0, 64 * 1024);
tcp_hashinfo.bind_bucket_cachep =
kmem_cache_create("tcp_bind_bucket",
sizeof(struct inet_bind_bucket), 0,
SLAB_HWCACHE_ALIGN | SLAB_PANIC |
SLAB_ACCOUNT,
NULL);
net: Add a bhash2 table hashed by port and address The current bind hashtable (bhash) is hashed by port only. In the socket bind path, we have to check for bind conflicts by traversing the specified port's inet_bind_bucket while holding the hashbucket's spinlock (see inet_csk_get_port() and inet_csk_bind_conflict()). In instances where there are tons of sockets hashed to the same port at different addresses, the bind conflict check is time-intensive and can cause softirq cpu lockups, as well as stops new tcp connections since __inet_inherit_port() also contests for the spinlock. This patch adds a second bind table, bhash2, that hashes by port and sk->sk_rcv_saddr (ipv4) and sk->sk_v6_rcv_saddr (ipv6). Searching the bhash2 table leads to significantly faster conflict resolution and less time holding the hashbucket spinlock. Please note a few things: * There can be the case where the a socket's address changes after it has been bound. There are two cases where this happens: 1) The case where there is a bind() call on INADDR_ANY (ipv4) or IPV6_ADDR_ANY (ipv6) and then a connect() call. The kernel will assign the socket an address when it handles the connect() 2) In inet_sk_reselect_saddr(), which is called when rebuilding the sk header and a few pre-conditions are met (eg rerouting fails). In these two cases, we need to update the bhash2 table by removing the entry for the old address, and add a new entry reflecting the updated address. * The bhash2 table must have its own lock, even though concurrent accesses on the same port are protected by the bhash lock. Bhash2 must have its own lock to protect against cases where sockets on different ports hash to different bhash hashbuckets but to the same bhash2 hashbucket. This brings up a few stipulations: 1) When acquiring both the bhash and the bhash2 lock, the bhash2 lock will always be acquired after the bhash lock and released before the bhash lock is released. 2) There are no nested bhash2 hashbucket locks. A bhash2 lock is always acquired+released before another bhash2 lock is acquired+released. * The bhash table cannot be superseded by the bhash2 table because for bind requests on INADDR_ANY (ipv4) or IPV6_ADDR_ANY (ipv6), every socket bound to that port must be checked for a potential conflict. The bhash table is the only source of port->socket associations. Signed-off-by: Joanne Koong <joannelkoong@gmail.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2022-08-23 02:10:21 +08:00
tcp_hashinfo.bind2_bucket_cachep =
kmem_cache_create("tcp_bind2_bucket",
sizeof(struct inet_bind2_bucket), 0,
SLAB_HWCACHE_ALIGN | SLAB_PANIC |
SLAB_ACCOUNT,
NULL);
/* Size and allocate the main established and bind bucket
* hash tables.
*
* The methodology is similar to that of the buffer cache.
*/
tcp_hashinfo.ehash =
alloc_large_system_hash("TCP established",
sizeof(struct inet_ehash_bucket),
thash_entries,
17, /* one slot per 128 KB of memory */
0,
NULL,
&tcp_hashinfo.ehash_mask,
0,
[TCP]: Saner thash_entries default with much memory. On systems with a very large amount of memory, the heuristics in alloc_large_system_hash() result in a very large TCP established hash table: 16 millions of entries for a 128 GB ia64 system. This makes reading from /proc/net/tcp pretty slow (well over a second) and as a result netstat is slow on these machines. I know that /proc/net/tcp is deprecated in favor of tcp_diag, however at the moment netstat only knows of the former. I am skeptical that such a large TCP established hash is often needed. Just because a system has a lot of memory doesn't imply that it will have several millions of concurrent TCP connections. Thus I believe that we should put an arbitrary high limit to the size of the TCP established hash by default. Users who really need a bigger hash can always use the thash_entries boot parameter to get more. I propose 2 millions of entries as the arbitrary high limit. This makes /proc/net/tcp reasonably fast on the system in question (0.2 s) while being still large enough for me to be confident that network performance won't suffer. This is just one way to limit the hash size, there are others; I am not familiar enough with the TCP code to decide which is best. Thus, I would welcome the proposals of alternatives. [ 2 million is still too large, thus I've modified the limit in the change to be '512 * 1024'. -DaveM ] Signed-off-by: Jean Delvare <jdelvare@suse.de> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-30 15:59:25 +08:00
thash_entries ? 0 : 512 * 1024);
tcp/dccp: remove twchain TCP listener refactoring, part 3 : Our goal is to hash SYN_RECV sockets into main ehash for fast lookup, and parallel SYN processing. Current inet_ehash_bucket contains two chains, one for ESTABLISH (and friend states) sockets, another for TIME_WAIT sockets only. As the hash table is sized to get at most one socket per bucket, it makes little sense to have separate twchain, as it makes the lookup slightly more complicated, and doubles hash table memory usage. If we make sure all socket types have the lookup keys at the same offsets, we can use a generic and faster lookup. It turns out TIME_WAIT and ESTABLISHED sockets already have common lookup fields for IPv4. [ INET_TW_MATCH() is no longer needed ] I'll provide a follow-up to factorize IPv6 lookup as well, to remove INET6_TW_MATCH() This way, SYN_RECV pseudo sockets will be supported the same. A new sock_gen_put() helper is added, doing either a sock_put() or inet_twsk_put() [ and will support SYN_RECV later ]. Note this helper should only be called in real slow path, when rcu lookup found a socket that was moved to another identity (freed/reused immediately), but could eventually be used in other contexts, like sock_edemux() Before patch : dmesg | grep "TCP established" TCP established hash table entries: 524288 (order: 11, 8388608 bytes) After patch : TCP established hash table entries: 524288 (order: 10, 4194304 bytes) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03 15:22:02 +08:00
for (i = 0; i <= tcp_hashinfo.ehash_mask; i++)
INIT_HLIST_NULLS_HEAD(&tcp_hashinfo.ehash[i].chain, i);
tcp/dccp: remove twchain TCP listener refactoring, part 3 : Our goal is to hash SYN_RECV sockets into main ehash for fast lookup, and parallel SYN processing. Current inet_ehash_bucket contains two chains, one for ESTABLISH (and friend states) sockets, another for TIME_WAIT sockets only. As the hash table is sized to get at most one socket per bucket, it makes little sense to have separate twchain, as it makes the lookup slightly more complicated, and doubles hash table memory usage. If we make sure all socket types have the lookup keys at the same offsets, we can use a generic and faster lookup. It turns out TIME_WAIT and ESTABLISHED sockets already have common lookup fields for IPv4. [ INET_TW_MATCH() is no longer needed ] I'll provide a follow-up to factorize IPv6 lookup as well, to remove INET6_TW_MATCH() This way, SYN_RECV pseudo sockets will be supported the same. A new sock_gen_put() helper is added, doing either a sock_put() or inet_twsk_put() [ and will support SYN_RECV later ]. Note this helper should only be called in real slow path, when rcu lookup found a socket that was moved to another identity (freed/reused immediately), but could eventually be used in other contexts, like sock_edemux() Before patch : dmesg | grep "TCP established" TCP established hash table entries: 524288 (order: 11, 8388608 bytes) After patch : TCP established hash table entries: 524288 (order: 10, 4194304 bytes) Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-10-03 15:22:02 +08:00
if (inet_ehash_locks_alloc(&tcp_hashinfo))
panic("TCP: failed to alloc ehash_locks");
tcp_hashinfo.bhash =
alloc_large_system_hash("TCP bind",
net: Add a bhash2 table hashed by port and address The current bind hashtable (bhash) is hashed by port only. In the socket bind path, we have to check for bind conflicts by traversing the specified port's inet_bind_bucket while holding the hashbucket's spinlock (see inet_csk_get_port() and inet_csk_bind_conflict()). In instances where there are tons of sockets hashed to the same port at different addresses, the bind conflict check is time-intensive and can cause softirq cpu lockups, as well as stops new tcp connections since __inet_inherit_port() also contests for the spinlock. This patch adds a second bind table, bhash2, that hashes by port and sk->sk_rcv_saddr (ipv4) and sk->sk_v6_rcv_saddr (ipv6). Searching the bhash2 table leads to significantly faster conflict resolution and less time holding the hashbucket spinlock. Please note a few things: * There can be the case where the a socket's address changes after it has been bound. There are two cases where this happens: 1) The case where there is a bind() call on INADDR_ANY (ipv4) or IPV6_ADDR_ANY (ipv6) and then a connect() call. The kernel will assign the socket an address when it handles the connect() 2) In inet_sk_reselect_saddr(), which is called when rebuilding the sk header and a few pre-conditions are met (eg rerouting fails). In these two cases, we need to update the bhash2 table by removing the entry for the old address, and add a new entry reflecting the updated address. * The bhash2 table must have its own lock, even though concurrent accesses on the same port are protected by the bhash lock. Bhash2 must have its own lock to protect against cases where sockets on different ports hash to different bhash hashbuckets but to the same bhash2 hashbucket. This brings up a few stipulations: 1) When acquiring both the bhash and the bhash2 lock, the bhash2 lock will always be acquired after the bhash lock and released before the bhash lock is released. 2) There are no nested bhash2 hashbucket locks. A bhash2 lock is always acquired+released before another bhash2 lock is acquired+released. * The bhash table cannot be superseded by the bhash2 table because for bind requests on INADDR_ANY (ipv4) or IPV6_ADDR_ANY (ipv6), every socket bound to that port must be checked for a potential conflict. The bhash table is the only source of port->socket associations. Signed-off-by: Joanne Koong <joannelkoong@gmail.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2022-08-23 02:10:21 +08:00
2 * sizeof(struct inet_bind_hashbucket),
tcp_hashinfo.ehash_mask + 1,
17, /* one slot per 128 KB of memory */
0,
&tcp_hashinfo.bhash_size,
NULL,
0,
64 * 1024);
tcp_hashinfo.bhash_size = 1U << tcp_hashinfo.bhash_size;
net: Add a bhash2 table hashed by port and address The current bind hashtable (bhash) is hashed by port only. In the socket bind path, we have to check for bind conflicts by traversing the specified port's inet_bind_bucket while holding the hashbucket's spinlock (see inet_csk_get_port() and inet_csk_bind_conflict()). In instances where there are tons of sockets hashed to the same port at different addresses, the bind conflict check is time-intensive and can cause softirq cpu lockups, as well as stops new tcp connections since __inet_inherit_port() also contests for the spinlock. This patch adds a second bind table, bhash2, that hashes by port and sk->sk_rcv_saddr (ipv4) and sk->sk_v6_rcv_saddr (ipv6). Searching the bhash2 table leads to significantly faster conflict resolution and less time holding the hashbucket spinlock. Please note a few things: * There can be the case where the a socket's address changes after it has been bound. There are two cases where this happens: 1) The case where there is a bind() call on INADDR_ANY (ipv4) or IPV6_ADDR_ANY (ipv6) and then a connect() call. The kernel will assign the socket an address when it handles the connect() 2) In inet_sk_reselect_saddr(), which is called when rebuilding the sk header and a few pre-conditions are met (eg rerouting fails). In these two cases, we need to update the bhash2 table by removing the entry for the old address, and add a new entry reflecting the updated address. * The bhash2 table must have its own lock, even though concurrent accesses on the same port are protected by the bhash lock. Bhash2 must have its own lock to protect against cases where sockets on different ports hash to different bhash hashbuckets but to the same bhash2 hashbucket. This brings up a few stipulations: 1) When acquiring both the bhash and the bhash2 lock, the bhash2 lock will always be acquired after the bhash lock and released before the bhash lock is released. 2) There are no nested bhash2 hashbucket locks. A bhash2 lock is always acquired+released before another bhash2 lock is acquired+released. * The bhash table cannot be superseded by the bhash2 table because for bind requests on INADDR_ANY (ipv4) or IPV6_ADDR_ANY (ipv6), every socket bound to that port must be checked for a potential conflict. The bhash table is the only source of port->socket associations. Signed-off-by: Joanne Koong <joannelkoong@gmail.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2022-08-23 02:10:21 +08:00
tcp_hashinfo.bhash2 = tcp_hashinfo.bhash + tcp_hashinfo.bhash_size;
for (i = 0; i < tcp_hashinfo.bhash_size; i++) {
spin_lock_init(&tcp_hashinfo.bhash[i].lock);
INIT_HLIST_HEAD(&tcp_hashinfo.bhash[i].chain);
net: Add a bhash2 table hashed by port and address The current bind hashtable (bhash) is hashed by port only. In the socket bind path, we have to check for bind conflicts by traversing the specified port's inet_bind_bucket while holding the hashbucket's spinlock (see inet_csk_get_port() and inet_csk_bind_conflict()). In instances where there are tons of sockets hashed to the same port at different addresses, the bind conflict check is time-intensive and can cause softirq cpu lockups, as well as stops new tcp connections since __inet_inherit_port() also contests for the spinlock. This patch adds a second bind table, bhash2, that hashes by port and sk->sk_rcv_saddr (ipv4) and sk->sk_v6_rcv_saddr (ipv6). Searching the bhash2 table leads to significantly faster conflict resolution and less time holding the hashbucket spinlock. Please note a few things: * There can be the case where the a socket's address changes after it has been bound. There are two cases where this happens: 1) The case where there is a bind() call on INADDR_ANY (ipv4) or IPV6_ADDR_ANY (ipv6) and then a connect() call. The kernel will assign the socket an address when it handles the connect() 2) In inet_sk_reselect_saddr(), which is called when rebuilding the sk header and a few pre-conditions are met (eg rerouting fails). In these two cases, we need to update the bhash2 table by removing the entry for the old address, and add a new entry reflecting the updated address. * The bhash2 table must have its own lock, even though concurrent accesses on the same port are protected by the bhash lock. Bhash2 must have its own lock to protect against cases where sockets on different ports hash to different bhash hashbuckets but to the same bhash2 hashbucket. This brings up a few stipulations: 1) When acquiring both the bhash and the bhash2 lock, the bhash2 lock will always be acquired after the bhash lock and released before the bhash lock is released. 2) There are no nested bhash2 hashbucket locks. A bhash2 lock is always acquired+released before another bhash2 lock is acquired+released. * The bhash table cannot be superseded by the bhash2 table because for bind requests on INADDR_ANY (ipv4) or IPV6_ADDR_ANY (ipv6), every socket bound to that port must be checked for a potential conflict. The bhash table is the only source of port->socket associations. Signed-off-by: Joanne Koong <joannelkoong@gmail.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2022-08-23 02:10:21 +08:00
spin_lock_init(&tcp_hashinfo.bhash2[i].lock);
INIT_HLIST_HEAD(&tcp_hashinfo.bhash2[i].chain);
}
tcp: Introduce optional per-netns ehash. The more sockets we have in the hash table, the longer we spend looking up the socket. While running a number of small workloads on the same host, they penalise each other and cause performance degradation. The root cause might be a single workload that consumes much more resources than the others. It often happens on a cloud service where different workloads share the same computing resource. On EC2 c5.24xlarge instance (196 GiB memory and 524288 (1Mi / 2) ehash entries), after running iperf3 in different netns, creating 24Mi sockets without data transfer in the root netns causes about 10% performance regression for the iperf3's connection. thash_entries sockets length Gbps 524288 1 1 50.7 24Mi 48 45.1 It is basically related to the length of the list of each hash bucket. For testing purposes to see how performance drops along the length, I set 131072 (1Mi / 8) to thash_entries, and here's the result. thash_entries sockets length Gbps 131072 1 1 50.7 1Mi 8 49.9 2Mi 16 48.9 4Mi 32 47.3 8Mi 64 44.6 16Mi 128 40.6 24Mi 192 36.3 32Mi 256 32.5 40Mi 320 27.0 48Mi 384 25.0 To resolve the socket lookup degradation, we introduce an optional per-netns hash table for TCP, but it's just ehash, and we still share the global bhash, bhash2 and lhash2. With a smaller ehash, we can look up non-listener sockets faster and isolate such noisy neighbours. In addition, we can reduce lock contention. We can control the ehash size by a new sysctl knob. However, depending on workloads, it will require very sensitive tuning, so we disable the feature by default (net.ipv4.tcp_child_ehash_entries == 0). Moreover, we can fall back to using the global ehash in case we fail to allocate enough memory for a new ehash. The maximum size is 16Mi, which is large enough that even if we have 48Mi sockets, the average list length is 3, and regression would be less than 1%. We can check the current ehash size by another read-only sysctl knob, net.ipv4.tcp_ehash_entries. A negative value means the netns shares the global ehash (per-netns ehash is disabled or failed to allocate memory). # dmesg | cut -d ' ' -f 5- | grep "established hash" TCP established hash table entries: 524288 (order: 10, 4194304 bytes, vmalloc hugepage) # sysctl net.ipv4.tcp_ehash_entries net.ipv4.tcp_ehash_entries = 524288 # can be changed by thash_entries # sysctl net.ipv4.tcp_child_ehash_entries net.ipv4.tcp_child_ehash_entries = 0 # disabled by default # ip netns add test1 # ip netns exec test1 sysctl net.ipv4.tcp_ehash_entries net.ipv4.tcp_ehash_entries = -524288 # share the global ehash # sysctl -w net.ipv4.tcp_child_ehash_entries=100 net.ipv4.tcp_child_ehash_entries = 100 # ip netns add test2 # ip netns exec test2 sysctl net.ipv4.tcp_ehash_entries net.ipv4.tcp_ehash_entries = 128 # own a per-netns ehash with 2^n buckets When more than two processes in the same netns create per-netns ehash concurrently with different sizes, we need to guarantee the size in one of the following ways: 1) Share the global ehash and create per-netns ehash First, unshare() with tcp_child_ehash_entries==0. It creates dedicated netns sysctl knobs where we can safely change tcp_child_ehash_entries and clone()/unshare() to create a per-netns ehash. 2) Control write on sysctl by BPF We can use BPF_PROG_TYPE_CGROUP_SYSCTL to allow/deny read/write on sysctl knobs. Note that the global ehash allocated at the boot time is spread over available NUMA nodes, but inet_pernet_hashinfo_alloc() will allocate pages for each per-netns ehash depending on the current process's NUMA policy. By default, the allocation is done in the local node only, so the per-netns hash table could fully reside on a random node. Thus, depending on the NUMA policy the netns is created with and the CPU the current thread is running on, we could see some performance differences for highly optimised networking applications. Note also that the default values of two sysctl knobs depend on the ehash size and should be tuned carefully: tcp_max_tw_buckets : tcp_child_ehash_entries / 2 tcp_max_syn_backlog : max(128, tcp_child_ehash_entries / 128) As a bonus, we can dismantle netns faster. Currently, while destroying netns, we call inet_twsk_purge(), which walks through the global ehash. It can be potentially big because it can have many sockets other than TIME_WAIT in all netns. Splitting ehash changes that situation, where it's only necessary for inet_twsk_purge() to clean up TIME_WAIT sockets in each netns. With regard to this, we do not free the per-netns ehash in inet_twsk_kill() to avoid UAF while iterating the per-netns ehash in inet_twsk_purge(). Instead, we do it in tcp_sk_exit_batch() after calling tcp_twsk_purge() to keep it protocol-family-independent. In the future, we could optimise ehash lookup/iteration further by removing netns comparison for the per-netns ehash. Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2022-09-08 09:10:22 +08:00
tcp_hashinfo.pernet = false;
cnt = tcp_hashinfo.ehash_mask + 1;
sysctl_tcp_max_orphans = cnt / 2;
tcp_init_mem();
/* Set per-socket limits to no more than 1/128 the pressure threshold */
limit = nr_free_buffer_pages() << (PAGE_SHIFT - 7);
tcp: change tcp_adv_win_scale and tcp_rmem[2] tcp_adv_win_scale default value is 2, meaning we expect a good citizen skb to have skb->len / skb->truesize ratio of 75% (3/4) In 2.6 kernels we (mis)accounted for typical MSS=1460 frame : 1536 + 64 + 256 = 1856 'estimated truesize', and 1856 * 3/4 = 1392. So these skbs were considered as not bloated. With recent truesize fixes, a typical MSS=1460 frame truesize is now the more precise : 2048 + 256 = 2304. But 2304 * 3/4 = 1728. So these skb are not good citizen anymore, because 1460 < 1728 (GRO can escape this problem because it build skbs with a too low truesize.) This also means tcp advertises a too optimistic window for a given allocated rcvspace : When receiving frames, sk_rmem_alloc can hit sk_rcvbuf limit and we call tcp_prune_queue()/tcp_collapse() too often, especially when application is slow to drain its receive queue or in case of losses (netperf is fast, scp is slow). This is a major latency source. We should adjust the len/truesize ratio to 50% instead of 75% This patch : 1) changes tcp_adv_win_scale default to 1 instead of 2 2) increase tcp_rmem[2] limit from 4MB to 6MB to take into account better truesize tracking and to allow autotuning tcp receive window to reach same value than before. Note that same amount of kernel memory is consumed compared to 2.6 kernels. Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Tom Herbert <therbert@google.com> Cc: Yuchung Cheng <ycheng@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-05-02 10:28:41 +08:00
max_wshare = min(4UL*1024*1024, limit);
max_rshare = min(6UL*1024*1024, limit);
init_net.ipv4.sysctl_tcp_wmem[0] = PAGE_SIZE;
init_net.ipv4.sysctl_tcp_wmem[1] = 16*1024;
init_net.ipv4.sysctl_tcp_wmem[2] = max(64*1024, max_wshare);
init_net.ipv4.sysctl_tcp_rmem[0] = PAGE_SIZE;
tcp: up initial rmem to 128KB and SYN rwin to around 64KB Previously TCP initial receive buffer is ~87KB by default and the initial receive window is ~29KB (20 MSS). This patch changes the two numbers to 128KB and ~64KB (rounding down to the multiples of MSS) respectively. The patch also simplifies the calculations s.t. the two numbers are directly controlled by sysctl tcp_rmem[1]: 1) Initial receiver buffer budget (sk_rcvbuf): while this should be configured via sysctl tcp_rmem[1], previously tcp_fixup_rcvbuf() always override and set a larger size when a new connection establishes. 2) Initial receive window in SYN: previously it is set to 20 packets if MSS <= 1460. The number 20 was based on the initial congestion window of 10: the receiver needs twice amount to avoid being limited by the receive window upon out-of-order delivery in the first window burst. But since this only applies if the receiving MSS <= 1460, connection using large MTU (e.g. to utilize receiver zero-copy) may be limited by the receive window. With this patch TCP memory configuration is more straight-forward and more properly sized to modern high-speed networks by default. Several popular stacks have been announcing 64KB rwin in SYNs as well. Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Wei Wang <weiwan@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-09-28 02:21:19 +08:00
init_net.ipv4.sysctl_tcp_rmem[1] = 131072;
init_net.ipv4.sysctl_tcp_rmem[2] = max(131072, max_rshare);
pr_info("Hash tables configured (established %u bind %u)\n",
tcp_hashinfo.ehash_mask + 1, tcp_hashinfo.bhash_size);
tcp_v4_init();
tcp_metrics_init();
BUG_ON(tcp_register_congestion_control(&tcp_reno) != 0);
tcp: TCP Small Queues This introduce TSQ (TCP Small Queues) TSQ goal is to reduce number of TCP packets in xmit queues (qdisc & device queues), to reduce RTT and cwnd bias, part of the bufferbloat problem. sk->sk_wmem_alloc not allowed to grow above a given limit, allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a given time. TSO packets are sized/capped to half the limit, so that we have two TSO packets in flight, allowing better bandwidth use. As a side effect, setting the limit to 40000 automatically reduces the standard gso max limit (65536) to 40000/2 : It can help to reduce latencies of high prio packets, having smaller TSO packets. This means we divert sock_wfree() to a tcp_wfree() handler, to queue/send following frames when skb_orphan() [2] is called for the already queued skbs. Results on my dev machines (tg3/ixgbe nics) are really impressive, using standard pfifo_fast, and with or without TSO/GSO. Without reduction of nominal bandwidth, we have reduction of buffering per bulk sender : < 1ms on Gbit (instead of 50ms with TSO) < 8ms on 100Mbit (instead of 132 ms) I no longer have 4 MBytes backlogged in qdisc by a single netperf session, and both side socket autotuning no longer use 4 Mbytes. As skb destructor cannot restart xmit itself ( as qdisc lock might be taken at this point ), we delegate the work to a tasklet. We use one tasklest per cpu for performance reasons. If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag. This flag is tested in a new protocol method called from release_sock(), to eventually send new segments. [1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable [2] skb_orphan() is usually called at TX completion time, but some drivers call it in their start_xmit() handler. These drivers should at least use BQL, or else a single TCP session can still fill the whole NIC TX ring, since TSQ will have no effect. Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Dave Taht <dave.taht@bufferbloat.net> Cc: Tom Herbert <therbert@google.com> Cc: Matt Mathis <mattmathis@google.com> Cc: Yuchung Cheng <ycheng@google.com> Cc: Nandita Dukkipati <nanditad@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
tcp_tasklet_init();
mptcp_init();
}