linux/net/ipv6/exthdrs.c

1409 lines
33 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-or-later
/*
* Extension Header handling for IPv6
* Linux INET6 implementation
*
* Authors:
* Pedro Roque <roque@di.fc.ul.pt>
* Andi Kleen <ak@muc.de>
* Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
*/
/* Changes:
* yoshfuji : ensure not to overrun while parsing
* tlv options.
* Mitsuru KANDA @USAGI and: Remove ipv6_parse_exthdrs().
* YOSHIFUJI Hideaki @USAGI Register inbound extension header
* handlers as inet6_protocol{}.
*/
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/netdevice.h>
#include <linux/in6.h>
#include <linux/icmpv6.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/export.h>
#include <net/dst.h>
#include <net/sock.h>
#include <net/snmp.h>
#include <net/ipv6.h>
#include <net/protocol.h>
#include <net/transp_v6.h>
#include <net/rawv6.h>
#include <net/ndisc.h>
#include <net/ip6_route.h>
#include <net/addrconf.h>
#include <net/calipso.h>
#if IS_ENABLED(CONFIG_IPV6_MIP6)
#include <net/xfrm.h>
#endif
#include <linux/seg6.h>
#include <net/seg6.h>
#ifdef CONFIG_IPV6_SEG6_HMAC
#include <net/seg6_hmac.h>
#endif
#include <net/rpl.h>
ipv6: ioam: Data plane support for Pre-allocated Trace Implement support for processing the IOAM Pre-allocated Trace with IPv6, see [1] and [2]. Introduce a new IPv6 Hop-by-Hop TLV option, see IANA [3]. A new per-interface sysctl is introduced. The value is a boolean to accept (=1) or ignore (=0, by default) IPv6 IOAM options on ingress for an interface: - net.ipv6.conf.XXX.ioam6_enabled Two other sysctls are introduced to define IOAM IDs, represented by an integer. They are respectively per-namespace and per-interface: - net.ipv6.ioam6_id - net.ipv6.conf.XXX.ioam6_id The value of the first one represents the IOAM ID of the node itself (u32; max and default value = U32_MAX>>8, due to hop limit concatenation) while the other represents the IOAM ID of an interface (u16; max and default value = U16_MAX). Each "ioam6_id" sysctl has a "_wide" equivalent: - net.ipv6.ioam6_id_wide - net.ipv6.conf.XXX.ioam6_id_wide The value of the first one represents the wide IOAM ID of the node itself (u64; max and default value = U64_MAX>>8, due to hop limit concatenation) while the other represents the wide IOAM ID of an interface (u32; max and default value = U32_MAX). The use of short and wide equivalents is not exclusive, a deployment could choose to leverage both. For example, net.ipv6.conf.XXX.ioam6_id (short format) could be an identifier for a physical interface, whereas net.ipv6.conf.XXX.ioam6_id_wide (wide format) could be an identifier for a logical sub-interface. Documentation about new sysctls is provided at the end of this patchset. Two relativistic hash tables are used: one for IOAM namespaces, the other for IOAM schemas. A namespace can only have a single active schema and a schema can only be attached to a single namespace (1:1 relationship). [1] https://tools.ietf.org/html/draft-ietf-ippm-ioam-ipv6-options [2] https://tools.ietf.org/html/draft-ietf-ippm-ioam-data [3] https://www.iana.org/assignments/ipv6-parameters/ipv6-parameters.xhtml#ipv6-parameters-2 Signed-off-by: Justin Iurman <justin.iurman@uliege.be> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-07-21 03:42:57 +08:00
#include <linux/ioam6.h>
#include <net/ioam6.h>
#include <net/dst_metadata.h>
#include <linux/uaccess.h>
/*********************
Generic functions
*********************/
/* An unknown option is detected, decide what to do */
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
static bool ip6_tlvopt_unknown(struct sk_buff *skb, int optoff,
bool disallow_unknowns)
{
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
if (disallow_unknowns) {
/* If unknown TLVs are disallowed by configuration
* then always silently drop packet. Note this also
* means no ICMP parameter problem is sent which
* could be a good property to mitigate a reflection DOS
* attack.
*/
goto drop;
}
switch ((skb_network_header(skb)[optoff] & 0xC0) >> 6) {
case 0: /* ignore */
return true;
case 1: /* drop packet */
break;
case 3: /* Send ICMP if not a multicast address and drop packet */
/* Actually, it is redundant check. icmp_send
will recheck in any case.
*/
if (ipv6_addr_is_multicast(&ipv6_hdr(skb)->daddr))
break;
fallthrough;
case 2: /* send ICMP PARM PROB regardless and drop packet */
icmpv6_param_prob_reason(skb, ICMPV6_UNK_OPTION, optoff,
SKB_DROP_REASON_UNHANDLED_PROTO);
return false;
}
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
drop:
kfree_skb_reason(skb, SKB_DROP_REASON_UNHANDLED_PROTO);
return false;
}
static bool ipv6_hop_ra(struct sk_buff *skb, int optoff);
static bool ipv6_hop_ioam(struct sk_buff *skb, int optoff);
static bool ipv6_hop_jumbo(struct sk_buff *skb, int optoff);
static bool ipv6_hop_calipso(struct sk_buff *skb, int optoff);
#if IS_ENABLED(CONFIG_IPV6_MIP6)
static bool ipv6_dest_hao(struct sk_buff *skb, int optoff);
#endif
/* Parse tlv encoded option header (hop-by-hop or destination) */
static bool ip6_parse_tlv(bool hopbyhop,
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
struct sk_buff *skb,
int max_count)
{
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
int len = (skb_transport_header(skb)[1] + 1) << 3;
const unsigned char *nh = skb_network_header(skb);
int off = skb_network_header_len(skb);
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
bool disallow_unknowns = false;
int tlv_count = 0;
int padlen = 0;
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
if (unlikely(max_count < 0)) {
disallow_unknowns = true;
max_count = -max_count;
}
if (skb_transport_offset(skb) + len > skb_headlen(skb))
goto bad;
off += 2;
len -= 2;
while (len > 0) {
int optlen, i;
if (nh[off] == IPV6_TLV_PAD1) {
padlen++;
if (padlen > 7)
goto bad;
off++;
len--;
continue;
}
if (len < 2)
goto bad;
optlen = nh[off + 1] + 2;
if (optlen > len)
goto bad;
if (nh[off] == IPV6_TLV_PADN) {
/* RFC 2460 states that the purpose of PadN is
* to align the containing header to multiples
* of 8. 7 is therefore the highest valid value.
* See also RFC 4942, Section 2.1.9.5.
*/
padlen += optlen;
if (padlen > 7)
goto bad;
/* RFC 4942 recommends receiving hosts to
* actively check PadN payload to contain
* only zeroes.
*/
for (i = 2; i < optlen; i++) {
if (nh[off + i] != 0)
goto bad;
}
} else {
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
tlv_count++;
if (tlv_count > max_count)
goto bad;
if (hopbyhop) {
switch (nh[off]) {
case IPV6_TLV_ROUTERALERT:
if (!ipv6_hop_ra(skb, off))
return false;
break;
case IPV6_TLV_IOAM:
if (!ipv6_hop_ioam(skb, off))
return false;
break;
case IPV6_TLV_JUMBO:
if (!ipv6_hop_jumbo(skb, off))
return false;
break;
case IPV6_TLV_CALIPSO:
if (!ipv6_hop_calipso(skb, off))
return false;
break;
default:
if (!ip6_tlvopt_unknown(skb, off,
disallow_unknowns))
return false;
break;
}
} else {
switch (nh[off]) {
#if IS_ENABLED(CONFIG_IPV6_MIP6)
case IPV6_TLV_HAO:
if (!ipv6_dest_hao(skb, off))
return false;
break;
#endif
default:
if (!ip6_tlvopt_unknown(skb, off,
disallow_unknowns))
return false;
break;
}
}
padlen = 0;
}
off += optlen;
len -= optlen;
}
if (len == 0)
return true;
bad:
kfree_skb_reason(skb, SKB_DROP_REASON_IP_INHDR);
return false;
}
/*****************************
Destination options header.
*****************************/
#if IS_ENABLED(CONFIG_IPV6_MIP6)
static bool ipv6_dest_hao(struct sk_buff *skb, int optoff)
{
struct ipv6_destopt_hao *hao;
struct inet6_skb_parm *opt = IP6CB(skb);
struct ipv6hdr *ipv6h = ipv6_hdr(skb);
SKB_DR(reason);
int ret;
if (opt->dsthao) {
net_dbg_ratelimited("hao duplicated\n");
goto discard;
}
opt->dsthao = opt->dst1;
opt->dst1 = 0;
hao = (struct ipv6_destopt_hao *)(skb_network_header(skb) + optoff);
if (hao->length != 16) {
net_dbg_ratelimited("hao invalid option length = %d\n",
hao->length);
SKB_DR_SET(reason, IP_INHDR);
goto discard;
}
if (!(ipv6_addr_type(&hao->addr) & IPV6_ADDR_UNICAST)) {
net_dbg_ratelimited("hao is not an unicast addr: %pI6\n",
&hao->addr);
SKB_DR_SET(reason, INVALID_PROTO);
goto discard;
}
ret = xfrm6_input_addr(skb, (xfrm_address_t *)&ipv6h->daddr,
(xfrm_address_t *)&hao->addr, IPPROTO_DSTOPTS);
if (unlikely(ret < 0)) {
SKB_DR_SET(reason, XFRM_POLICY);
goto discard;
}
if (skb_cloned(skb)) {
if (pskb_expand_head(skb, 0, 0, GFP_ATOMIC))
goto discard;
/* update all variable using below by copied skbuff */
hao = (struct ipv6_destopt_hao *)(skb_network_header(skb) +
optoff);
ipv6h = ipv6_hdr(skb);
}
if (skb->ip_summed == CHECKSUM_COMPLETE)
skb->ip_summed = CHECKSUM_NONE;
swap(ipv6h->saddr, hao->addr);
if (skb->tstamp == 0)
__net_timestamp(skb);
return true;
discard:
kfree_skb_reason(skb, reason);
return false;
}
#endif
static int ipv6_destopt_rcv(struct sk_buff *skb)
{
struct inet6_dev *idev = __in6_dev_get(skb->dev);
struct inet6_skb_parm *opt = IP6CB(skb);
#if IS_ENABLED(CONFIG_IPV6_MIP6)
__u16 dstbuf;
#endif
struct dst_entry *dst = skb_dst(skb);
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
struct net *net = dev_net(skb->dev);
int extlen;
if (!pskb_may_pull(skb, skb_transport_offset(skb) + 8) ||
!pskb_may_pull(skb, (skb_transport_offset(skb) +
((skb_transport_header(skb)[1] + 1) << 3)))) {
__IP6_INC_STATS(dev_net(dst->dev), idev,
IPSTATS_MIB_INHDRERRORS);
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
fail_and_free:
kfree_skb(skb);
return -1;
}
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
extlen = (skb_transport_header(skb)[1] + 1) << 3;
if (extlen > net->ipv6.sysctl.max_dst_opts_len)
goto fail_and_free;
opt->lastopt = opt->dst1 = skb_network_header_len(skb);
#if IS_ENABLED(CONFIG_IPV6_MIP6)
dstbuf = opt->dst1;
#endif
if (ip6_parse_tlv(false, skb, net->ipv6.sysctl.max_dst_opts_cnt)) {
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
skb->transport_header += extlen;
opt = IP6CB(skb);
#if IS_ENABLED(CONFIG_IPV6_MIP6)
opt->nhoff = dstbuf;
#else
opt->nhoff = opt->dst1;
#endif
return 1;
}
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INHDRERRORS);
return -1;
}
static void seg6_update_csum(struct sk_buff *skb)
{
struct ipv6_sr_hdr *hdr;
struct in6_addr *addr;
__be32 from, to;
/* srh is at transport offset and seg_left is already decremented
* but daddr is not yet updated with next segment
*/
hdr = (struct ipv6_sr_hdr *)skb_transport_header(skb);
addr = hdr->segments + hdr->segments_left;
hdr->segments_left++;
from = *(__be32 *)hdr;
hdr->segments_left--;
to = *(__be32 *)hdr;
/* update skb csum with diff resulting from seg_left decrement */
update_csum_diff4(skb, from, to);
/* compute csum diff between current and next segment and update */
update_csum_diff16(skb, (__be32 *)(&ipv6_hdr(skb)->daddr),
(__be32 *)addr);
}
static int ipv6_srh_rcv(struct sk_buff *skb)
{
struct inet6_skb_parm *opt = IP6CB(skb);
struct net *net = dev_net(skb->dev);
struct ipv6_sr_hdr *hdr;
struct inet6_dev *idev;
struct in6_addr *addr;
int accept_seg6;
hdr = (struct ipv6_sr_hdr *)skb_transport_header(skb);
idev = __in6_dev_get(skb->dev);
accept_seg6 = net->ipv6.devconf_all->seg6_enabled;
if (accept_seg6 > idev->cnf.seg6_enabled)
accept_seg6 = idev->cnf.seg6_enabled;
if (!accept_seg6) {
kfree_skb(skb);
return -1;
}
#ifdef CONFIG_IPV6_SEG6_HMAC
if (!seg6_hmac_validate_skb(skb)) {
kfree_skb(skb);
return -1;
}
#endif
looped_back:
if (hdr->segments_left == 0) {
if (hdr->nexthdr == NEXTHDR_IPV6 || hdr->nexthdr == NEXTHDR_IPV4) {
int offset = (hdr->hdrlen + 1) << 3;
skb_postpull_rcsum(skb, skb_network_header(skb),
skb_network_header_len(skb));
if (!pskb_pull(skb, offset)) {
kfree_skb(skb);
return -1;
}
skb_postpull_rcsum(skb, skb_transport_header(skb),
offset);
skb_reset_network_header(skb);
skb_reset_transport_header(skb);
skb->encapsulation = 0;
if (hdr->nexthdr == NEXTHDR_IPV4)
skb->protocol = htons(ETH_P_IP);
__skb_tunnel_rx(skb, skb->dev, net);
netif_rx(skb);
return -1;
}
opt->srcrt = skb_network_header_len(skb);
opt->lastopt = opt->srcrt;
skb->transport_header += (hdr->hdrlen + 1) << 3;
opt->nhoff = (&hdr->nexthdr) - skb_network_header(skb);
return 1;
}
if (hdr->segments_left >= (hdr->hdrlen >> 1)) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INHDRERRORS);
icmpv6_param_prob(skb, ICMPV6_HDR_FIELD,
((&hdr->segments_left) -
skb_network_header(skb)));
return -1;
}
if (skb_cloned(skb)) {
if (pskb_expand_head(skb, 0, 0, GFP_ATOMIC)) {
__IP6_INC_STATS(net, ip6_dst_idev(skb_dst(skb)),
IPSTATS_MIB_OUTDISCARDS);
kfree_skb(skb);
return -1;
}
}
hdr = (struct ipv6_sr_hdr *)skb_transport_header(skb);
hdr->segments_left--;
addr = hdr->segments + hdr->segments_left;
skb_push(skb, sizeof(struct ipv6hdr));
if (skb->ip_summed == CHECKSUM_COMPLETE)
seg6_update_csum(skb);
ipv6_hdr(skb)->daddr = *addr;
skb_dst_drop(skb);
ip6_route_input(skb);
if (skb_dst(skb)->error) {
dst_input(skb);
return -1;
}
if (skb_dst(skb)->dev->flags & IFF_LOOPBACK) {
if (ipv6_hdr(skb)->hop_limit <= 1) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INHDRERRORS);
icmpv6_send(skb, ICMPV6_TIME_EXCEED,
ICMPV6_EXC_HOPLIMIT, 0);
kfree_skb(skb);
return -1;
}
ipv6_hdr(skb)->hop_limit--;
skb_pull(skb, sizeof(struct ipv6hdr));
goto looped_back;
}
dst_input(skb);
return -1;
}
static int ipv6_rpl_srh_rcv(struct sk_buff *skb)
{
struct ipv6_rpl_sr_hdr *hdr, *ohdr, *chdr;
struct inet6_skb_parm *opt = IP6CB(skb);
struct net *net = dev_net(skb->dev);
struct inet6_dev *idev;
struct ipv6hdr *oldhdr;
unsigned char *buf;
int accept_rpl_seg;
int i, err;
u64 n = 0;
u32 r;
idev = __in6_dev_get(skb->dev);
accept_rpl_seg = net->ipv6.devconf_all->rpl_seg_enabled;
if (accept_rpl_seg > idev->cnf.rpl_seg_enabled)
accept_rpl_seg = idev->cnf.rpl_seg_enabled;
if (!accept_rpl_seg) {
kfree_skb(skb);
return -1;
}
looped_back:
hdr = (struct ipv6_rpl_sr_hdr *)skb_transport_header(skb);
if (hdr->segments_left == 0) {
if (hdr->nexthdr == NEXTHDR_IPV6) {
int offset = (hdr->hdrlen + 1) << 3;
skb_postpull_rcsum(skb, skb_network_header(skb),
skb_network_header_len(skb));
if (!pskb_pull(skb, offset)) {
kfree_skb(skb);
return -1;
}
skb_postpull_rcsum(skb, skb_transport_header(skb),
offset);
skb_reset_network_header(skb);
skb_reset_transport_header(skb);
skb->encapsulation = 0;
__skb_tunnel_rx(skb, skb->dev, net);
netif_rx(skb);
return -1;
}
opt->srcrt = skb_network_header_len(skb);
opt->lastopt = opt->srcrt;
skb->transport_header += (hdr->hdrlen + 1) << 3;
opt->nhoff = (&hdr->nexthdr) - skb_network_header(skb);
return 1;
}
if (!pskb_may_pull(skb, sizeof(*hdr))) {
kfree_skb(skb);
return -1;
}
n = (hdr->hdrlen << 3) - hdr->pad - (16 - hdr->cmpre);
r = do_div(n, (16 - hdr->cmpri));
/* checks if calculation was without remainder and n fits into
* unsigned char which is segments_left field. Should not be
* higher than that.
*/
if (r || (n + 1) > 255) {
kfree_skb(skb);
return -1;
}
if (hdr->segments_left > n + 1) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INHDRERRORS);
icmpv6_param_prob(skb, ICMPV6_HDR_FIELD,
((&hdr->segments_left) -
skb_network_header(skb)));
return -1;
}
if (skb_cloned(skb)) {
if (pskb_expand_head(skb, IPV6_RPL_SRH_WORST_SWAP_SIZE, 0,
GFP_ATOMIC)) {
__IP6_INC_STATS(net, ip6_dst_idev(skb_dst(skb)),
IPSTATS_MIB_OUTDISCARDS);
kfree_skb(skb);
return -1;
}
} else {
err = skb_cow_head(skb, IPV6_RPL_SRH_WORST_SWAP_SIZE);
if (unlikely(err)) {
kfree_skb(skb);
return -1;
}
}
hdr = (struct ipv6_rpl_sr_hdr *)skb_transport_header(skb);
if (!pskb_may_pull(skb, ipv6_rpl_srh_size(n, hdr->cmpri,
hdr->cmpre))) {
kfree_skb(skb);
return -1;
}
hdr->segments_left--;
i = n - hdr->segments_left;
buf = kcalloc(struct_size(hdr, segments.addr, n + 2), 2, GFP_ATOMIC);
if (unlikely(!buf)) {
kfree_skb(skb);
return -1;
}
ohdr = (struct ipv6_rpl_sr_hdr *)buf;
ipv6_rpl_srh_decompress(ohdr, hdr, &ipv6_hdr(skb)->daddr, n);
chdr = (struct ipv6_rpl_sr_hdr *)(buf + ((ohdr->hdrlen + 1) << 3));
if ((ipv6_addr_type(&ipv6_hdr(skb)->daddr) & IPV6_ADDR_MULTICAST) ||
(ipv6_addr_type(&ohdr->rpl_segaddr[i]) & IPV6_ADDR_MULTICAST)) {
kfree_skb(skb);
kfree(buf);
return -1;
}
err = ipv6_chk_rpl_srh_loop(net, ohdr->rpl_segaddr, n + 1);
if (err) {
icmpv6_send(skb, ICMPV6_PARAMPROB, 0, 0);
kfree_skb(skb);
kfree(buf);
return -1;
}
swap(ipv6_hdr(skb)->daddr, ohdr->rpl_segaddr[i]);
ipv6_rpl_srh_compress(chdr, ohdr, &ipv6_hdr(skb)->daddr, n);
oldhdr = ipv6_hdr(skb);
skb_pull(skb, ((hdr->hdrlen + 1) << 3));
skb_postpull_rcsum(skb, oldhdr,
sizeof(struct ipv6hdr) + ((hdr->hdrlen + 1) << 3));
skb_push(skb, ((chdr->hdrlen + 1) << 3) + sizeof(struct ipv6hdr));
skb_reset_network_header(skb);
skb_mac_header_rebuild(skb);
skb_set_transport_header(skb, sizeof(struct ipv6hdr));
memmove(ipv6_hdr(skb), oldhdr, sizeof(struct ipv6hdr));
memcpy(skb_transport_header(skb), chdr, (chdr->hdrlen + 1) << 3);
ipv6_hdr(skb)->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
skb_postpush_rcsum(skb, ipv6_hdr(skb),
sizeof(struct ipv6hdr) + ((chdr->hdrlen + 1) << 3));
kfree(buf);
skb_dst_drop(skb);
ip6_route_input(skb);
if (skb_dst(skb)->error) {
dst_input(skb);
return -1;
}
if (skb_dst(skb)->dev->flags & IFF_LOOPBACK) {
if (ipv6_hdr(skb)->hop_limit <= 1) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INHDRERRORS);
icmpv6_send(skb, ICMPV6_TIME_EXCEED,
ICMPV6_EXC_HOPLIMIT, 0);
kfree_skb(skb);
return -1;
}
ipv6_hdr(skb)->hop_limit--;
skb_pull(skb, sizeof(struct ipv6hdr));
goto looped_back;
}
dst_input(skb);
return -1;
}
/********************************
Routing header.
********************************/
/* called with rcu_read_lock() */
static int ipv6_rthdr_rcv(struct sk_buff *skb)
{
struct inet6_dev *idev = __in6_dev_get(skb->dev);
struct inet6_skb_parm *opt = IP6CB(skb);
struct in6_addr *addr = NULL;
struct in6_addr daddr;
int n, i;
struct ipv6_rt_hdr *hdr;
struct rt0_hdr *rthdr;
struct net *net = dev_net(skb->dev);
int accept_source_route = net->ipv6.devconf_all->accept_source_route;
if (idev && accept_source_route > idev->cnf.accept_source_route)
accept_source_route = idev->cnf.accept_source_route;
if (!pskb_may_pull(skb, skb_transport_offset(skb) + 8) ||
!pskb_may_pull(skb, (skb_transport_offset(skb) +
((skb_transport_header(skb)[1] + 1) << 3)))) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INHDRERRORS);
kfree_skb(skb);
return -1;
}
hdr = (struct ipv6_rt_hdr *)skb_transport_header(skb);
if (ipv6_addr_is_multicast(&ipv6_hdr(skb)->daddr) ||
skb->pkt_type != PACKET_HOST) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INADDRERRORS);
kfree_skb(skb);
return -1;
}
switch (hdr->type) {
case IPV6_SRCRT_TYPE_4:
/* segment routing */
return ipv6_srh_rcv(skb);
case IPV6_SRCRT_TYPE_3:
/* rpl segment routing */
return ipv6_rpl_srh_rcv(skb);
default:
break;
}
looped_back:
if (hdr->segments_left == 0) {
switch (hdr->type) {
#if IS_ENABLED(CONFIG_IPV6_MIP6)
case IPV6_SRCRT_TYPE_2:
/* Silently discard type 2 header unless it was
* processed by own
*/
if (!addr) {
__IP6_INC_STATS(net, idev,
IPSTATS_MIB_INADDRERRORS);
kfree_skb(skb);
return -1;
}
break;
#endif
default:
break;
}
opt->lastopt = opt->srcrt = skb_network_header_len(skb);
skb->transport_header += (hdr->hdrlen + 1) << 3;
opt->dst0 = opt->dst1;
opt->dst1 = 0;
opt->nhoff = (&hdr->nexthdr) - skb_network_header(skb);
return 1;
}
switch (hdr->type) {
#if IS_ENABLED(CONFIG_IPV6_MIP6)
case IPV6_SRCRT_TYPE_2:
if (accept_source_route < 0)
goto unknown_rh;
/* Silently discard invalid RTH type 2 */
if (hdr->hdrlen != 2 || hdr->segments_left != 1) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INHDRERRORS);
kfree_skb(skb);
return -1;
}
break;
#endif
default:
goto unknown_rh;
}
/*
* This is the routing header forwarding algorithm from
* RFC 2460, page 16.
*/
n = hdr->hdrlen >> 1;
if (hdr->segments_left > n) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INHDRERRORS);
icmpv6_param_prob(skb, ICMPV6_HDR_FIELD,
((&hdr->segments_left) -
skb_network_header(skb)));
return -1;
}
/* We are about to mangle packet header. Be careful!
Do not damage packets queued somewhere.
*/
if (skb_cloned(skb)) {
/* the copy is a forwarded packet */
if (pskb_expand_head(skb, 0, 0, GFP_ATOMIC)) {
__IP6_INC_STATS(net, ip6_dst_idev(skb_dst(skb)),
IPSTATS_MIB_OUTDISCARDS);
kfree_skb(skb);
return -1;
}
hdr = (struct ipv6_rt_hdr *)skb_transport_header(skb);
}
if (skb->ip_summed == CHECKSUM_COMPLETE)
skb->ip_summed = CHECKSUM_NONE;
i = n - --hdr->segments_left;
rthdr = (struct rt0_hdr *) hdr;
addr = rthdr->addr;
addr += i - 1;
switch (hdr->type) {
#if IS_ENABLED(CONFIG_IPV6_MIP6)
case IPV6_SRCRT_TYPE_2:
if (xfrm6_input_addr(skb, (xfrm_address_t *)addr,
(xfrm_address_t *)&ipv6_hdr(skb)->saddr,
IPPROTO_ROUTING) < 0) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INADDRERRORS);
kfree_skb(skb);
return -1;
}
if (!ipv6_chk_home_addr(dev_net(skb_dst(skb)->dev), addr)) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INADDRERRORS);
kfree_skb(skb);
return -1;
}
break;
#endif
default:
break;
}
if (ipv6_addr_is_multicast(addr)) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INADDRERRORS);
kfree_skb(skb);
return -1;
}
daddr = *addr;
*addr = ipv6_hdr(skb)->daddr;
ipv6_hdr(skb)->daddr = daddr;
skb_dst_drop(skb);
ip6_route_input(skb);
if (skb_dst(skb)->error) {
skb_push(skb, skb->data - skb_network_header(skb));
dst_input(skb);
return -1;
}
if (skb_dst(skb)->dev->flags&IFF_LOOPBACK) {
if (ipv6_hdr(skb)->hop_limit <= 1) {
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INHDRERRORS);
icmpv6_send(skb, ICMPV6_TIME_EXCEED, ICMPV6_EXC_HOPLIMIT,
0);
kfree_skb(skb);
return -1;
}
ipv6_hdr(skb)->hop_limit--;
goto looped_back;
}
skb_push(skb, skb->data - skb_network_header(skb));
dst_input(skb);
return -1;
unknown_rh:
__IP6_INC_STATS(net, idev, IPSTATS_MIB_INHDRERRORS);
icmpv6_param_prob(skb, ICMPV6_HDR_FIELD,
(&hdr->type) - skb_network_header(skb));
return -1;
}
static const struct inet6_protocol rthdr_protocol = {
.handler = ipv6_rthdr_rcv,
.flags = INET6_PROTO_NOPOLICY,
};
static const struct inet6_protocol destopt_protocol = {
.handler = ipv6_destopt_rcv,
.flags = INET6_PROTO_NOPOLICY,
};
static const struct inet6_protocol nodata_protocol = {
.handler = dst_discard,
.flags = INET6_PROTO_NOPOLICY,
};
int __init ipv6_exthdrs_init(void)
{
int ret;
ret = inet6_add_protocol(&rthdr_protocol, IPPROTO_ROUTING);
if (ret)
goto out;
ret = inet6_add_protocol(&destopt_protocol, IPPROTO_DSTOPTS);
if (ret)
goto out_rthdr;
ret = inet6_add_protocol(&nodata_protocol, IPPROTO_NONE);
if (ret)
goto out_destopt;
out:
return ret;
out_destopt:
inet6_del_protocol(&destopt_protocol, IPPROTO_DSTOPTS);
out_rthdr:
inet6_del_protocol(&rthdr_protocol, IPPROTO_ROUTING);
goto out;
};
void ipv6_exthdrs_exit(void)
{
inet6_del_protocol(&nodata_protocol, IPPROTO_NONE);
inet6_del_protocol(&destopt_protocol, IPPROTO_DSTOPTS);
inet6_del_protocol(&rthdr_protocol, IPPROTO_ROUTING);
}
/**********************************
Hop-by-hop options.
**********************************/
/*
* Note: we cannot rely on skb_dst(skb) before we assign it in ip6_route_input().
*/
static inline struct net *ipv6_skb_net(struct sk_buff *skb)
{
return skb_dst(skb) ? dev_net(skb_dst(skb)->dev) : dev_net(skb->dev);
}
/* Router Alert as of RFC 2711 */
static bool ipv6_hop_ra(struct sk_buff *skb, int optoff)
{
const unsigned char *nh = skb_network_header(skb);
if (nh[optoff + 1] == 2) {
IP6CB(skb)->flags |= IP6SKB_ROUTERALERT;
memcpy(&IP6CB(skb)->ra, nh + optoff + 2, sizeof(IP6CB(skb)->ra));
return true;
}
net_dbg_ratelimited("ipv6_hop_ra: wrong RA length %d\n",
nh[optoff + 1]);
kfree_skb_reason(skb, SKB_DROP_REASON_IP_INHDR);
return false;
}
ipv6: ioam: Data plane support for Pre-allocated Trace Implement support for processing the IOAM Pre-allocated Trace with IPv6, see [1] and [2]. Introduce a new IPv6 Hop-by-Hop TLV option, see IANA [3]. A new per-interface sysctl is introduced. The value is a boolean to accept (=1) or ignore (=0, by default) IPv6 IOAM options on ingress for an interface: - net.ipv6.conf.XXX.ioam6_enabled Two other sysctls are introduced to define IOAM IDs, represented by an integer. They are respectively per-namespace and per-interface: - net.ipv6.ioam6_id - net.ipv6.conf.XXX.ioam6_id The value of the first one represents the IOAM ID of the node itself (u32; max and default value = U32_MAX>>8, due to hop limit concatenation) while the other represents the IOAM ID of an interface (u16; max and default value = U16_MAX). Each "ioam6_id" sysctl has a "_wide" equivalent: - net.ipv6.ioam6_id_wide - net.ipv6.conf.XXX.ioam6_id_wide The value of the first one represents the wide IOAM ID of the node itself (u64; max and default value = U64_MAX>>8, due to hop limit concatenation) while the other represents the wide IOAM ID of an interface (u32; max and default value = U32_MAX). The use of short and wide equivalents is not exclusive, a deployment could choose to leverage both. For example, net.ipv6.conf.XXX.ioam6_id (short format) could be an identifier for a physical interface, whereas net.ipv6.conf.XXX.ioam6_id_wide (wide format) could be an identifier for a logical sub-interface. Documentation about new sysctls is provided at the end of this patchset. Two relativistic hash tables are used: one for IOAM namespaces, the other for IOAM schemas. A namespace can only have a single active schema and a schema can only be attached to a single namespace (1:1 relationship). [1] https://tools.ietf.org/html/draft-ietf-ippm-ioam-ipv6-options [2] https://tools.ietf.org/html/draft-ietf-ippm-ioam-data [3] https://www.iana.org/assignments/ipv6-parameters/ipv6-parameters.xhtml#ipv6-parameters-2 Signed-off-by: Justin Iurman <justin.iurman@uliege.be> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-07-21 03:42:57 +08:00
/* IOAM */
static bool ipv6_hop_ioam(struct sk_buff *skb, int optoff)
{
struct ioam6_trace_hdr *trace;
struct ioam6_namespace *ns;
struct ioam6_hdr *hdr;
/* Bad alignment (must be 4n-aligned) */
if (optoff & 3)
goto drop;
/* Ignore if IOAM is not enabled on ingress */
if (!__in6_dev_get(skb->dev)->cnf.ioam6_enabled)
goto ignore;
/* Truncated Option header */
hdr = (struct ioam6_hdr *)(skb_network_header(skb) + optoff);
if (hdr->opt_len < 2)
goto drop;
switch (hdr->type) {
case IOAM6_TYPE_PREALLOC:
/* Truncated Pre-allocated Trace header */
if (hdr->opt_len < 2 + sizeof(*trace))
goto drop;
/* Malformed Pre-allocated Trace header */
trace = (struct ioam6_trace_hdr *)((u8 *)hdr + sizeof(*hdr));
if (hdr->opt_len < 2 + sizeof(*trace) + trace->remlen * 4)
goto drop;
/* Ignore if the IOAM namespace is unknown */
ns = ioam6_namespace(ipv6_skb_net(skb), trace->namespace_id);
if (!ns)
goto ignore;
if (!skb_valid_dst(skb))
ip6_route_input(skb);
ioam6_fill_trace_data(skb, ns, trace, true);
ipv6: ioam: Data plane support for Pre-allocated Trace Implement support for processing the IOAM Pre-allocated Trace with IPv6, see [1] and [2]. Introduce a new IPv6 Hop-by-Hop TLV option, see IANA [3]. A new per-interface sysctl is introduced. The value is a boolean to accept (=1) or ignore (=0, by default) IPv6 IOAM options on ingress for an interface: - net.ipv6.conf.XXX.ioam6_enabled Two other sysctls are introduced to define IOAM IDs, represented by an integer. They are respectively per-namespace and per-interface: - net.ipv6.ioam6_id - net.ipv6.conf.XXX.ioam6_id The value of the first one represents the IOAM ID of the node itself (u32; max and default value = U32_MAX>>8, due to hop limit concatenation) while the other represents the IOAM ID of an interface (u16; max and default value = U16_MAX). Each "ioam6_id" sysctl has a "_wide" equivalent: - net.ipv6.ioam6_id_wide - net.ipv6.conf.XXX.ioam6_id_wide The value of the first one represents the wide IOAM ID of the node itself (u64; max and default value = U64_MAX>>8, due to hop limit concatenation) while the other represents the wide IOAM ID of an interface (u32; max and default value = U32_MAX). The use of short and wide equivalents is not exclusive, a deployment could choose to leverage both. For example, net.ipv6.conf.XXX.ioam6_id (short format) could be an identifier for a physical interface, whereas net.ipv6.conf.XXX.ioam6_id_wide (wide format) could be an identifier for a logical sub-interface. Documentation about new sysctls is provided at the end of this patchset. Two relativistic hash tables are used: one for IOAM namespaces, the other for IOAM schemas. A namespace can only have a single active schema and a schema can only be attached to a single namespace (1:1 relationship). [1] https://tools.ietf.org/html/draft-ietf-ippm-ioam-ipv6-options [2] https://tools.ietf.org/html/draft-ietf-ippm-ioam-data [3] https://www.iana.org/assignments/ipv6-parameters/ipv6-parameters.xhtml#ipv6-parameters-2 Signed-off-by: Justin Iurman <justin.iurman@uliege.be> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-07-21 03:42:57 +08:00
break;
default:
break;
}
ignore:
return true;
drop:
kfree_skb_reason(skb, SKB_DROP_REASON_IP_INHDR);
ipv6: ioam: Data plane support for Pre-allocated Trace Implement support for processing the IOAM Pre-allocated Trace with IPv6, see [1] and [2]. Introduce a new IPv6 Hop-by-Hop TLV option, see IANA [3]. A new per-interface sysctl is introduced. The value is a boolean to accept (=1) or ignore (=0, by default) IPv6 IOAM options on ingress for an interface: - net.ipv6.conf.XXX.ioam6_enabled Two other sysctls are introduced to define IOAM IDs, represented by an integer. They are respectively per-namespace and per-interface: - net.ipv6.ioam6_id - net.ipv6.conf.XXX.ioam6_id The value of the first one represents the IOAM ID of the node itself (u32; max and default value = U32_MAX>>8, due to hop limit concatenation) while the other represents the IOAM ID of an interface (u16; max and default value = U16_MAX). Each "ioam6_id" sysctl has a "_wide" equivalent: - net.ipv6.ioam6_id_wide - net.ipv6.conf.XXX.ioam6_id_wide The value of the first one represents the wide IOAM ID of the node itself (u64; max and default value = U64_MAX>>8, due to hop limit concatenation) while the other represents the wide IOAM ID of an interface (u32; max and default value = U32_MAX). The use of short and wide equivalents is not exclusive, a deployment could choose to leverage both. For example, net.ipv6.conf.XXX.ioam6_id (short format) could be an identifier for a physical interface, whereas net.ipv6.conf.XXX.ioam6_id_wide (wide format) could be an identifier for a logical sub-interface. Documentation about new sysctls is provided at the end of this patchset. Two relativistic hash tables are used: one for IOAM namespaces, the other for IOAM schemas. A namespace can only have a single active schema and a schema can only be attached to a single namespace (1:1 relationship). [1] https://tools.ietf.org/html/draft-ietf-ippm-ioam-ipv6-options [2] https://tools.ietf.org/html/draft-ietf-ippm-ioam-data [3] https://www.iana.org/assignments/ipv6-parameters/ipv6-parameters.xhtml#ipv6-parameters-2 Signed-off-by: Justin Iurman <justin.iurman@uliege.be> Signed-off-by: David S. Miller <davem@davemloft.net>
2021-07-21 03:42:57 +08:00
return false;
}
/* Jumbo payload */
static bool ipv6_hop_jumbo(struct sk_buff *skb, int optoff)
{
const unsigned char *nh = skb_network_header(skb);
SKB_DR(reason);
u32 pkt_len;
if (nh[optoff + 1] != 4 || (optoff & 3) != 2) {
net_dbg_ratelimited("ipv6_hop_jumbo: wrong jumbo opt length/alignment %d\n",
nh[optoff+1]);
SKB_DR_SET(reason, IP_INHDR);
goto drop;
}
pkt_len = ntohl(*(__be32 *)(nh + optoff + 2));
if (pkt_len <= IPV6_MAXPLEN) {
icmpv6_param_prob_reason(skb, ICMPV6_HDR_FIELD, optoff + 2,
SKB_DROP_REASON_IP_INHDR);
return false;
}
if (ipv6_hdr(skb)->payload_len) {
icmpv6_param_prob_reason(skb, ICMPV6_HDR_FIELD, optoff,
SKB_DROP_REASON_IP_INHDR);
return false;
}
if (pkt_len > skb->len - sizeof(struct ipv6hdr)) {
SKB_DR_SET(reason, PKT_TOO_SMALL);
goto drop;
}
if (pskb_trim_rcsum(skb, pkt_len + sizeof(struct ipv6hdr)))
goto drop;
IP6CB(skb)->flags |= IP6SKB_JUMBOGRAM;
return true;
drop:
kfree_skb_reason(skb, reason);
return false;
}
/* CALIPSO RFC 5570 */
static bool ipv6_hop_calipso(struct sk_buff *skb, int optoff)
{
const unsigned char *nh = skb_network_header(skb);
if (nh[optoff + 1] < 8)
goto drop;
if (nh[optoff + 6] * 4 + 8 > nh[optoff + 1])
goto drop;
if (!calipso_validate(skb, nh + optoff))
goto drop;
return true;
drop:
kfree_skb_reason(skb, SKB_DROP_REASON_IP_INHDR);
return false;
}
int ipv6_parse_hopopts(struct sk_buff *skb)
{
struct inet6_skb_parm *opt = IP6CB(skb);
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
struct net *net = dev_net(skb->dev);
int extlen;
/*
* skb_network_header(skb) is equal to skb->data, and
* skb_network_header_len(skb) is always equal to
* sizeof(struct ipv6hdr) by definition of
* hop-by-hop options.
*/
if (!pskb_may_pull(skb, sizeof(struct ipv6hdr) + 8) ||
!pskb_may_pull(skb, (sizeof(struct ipv6hdr) +
((skb_transport_header(skb)[1] + 1) << 3)))) {
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
fail_and_free:
kfree_skb(skb);
return -1;
}
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
extlen = (skb_transport_header(skb)[1] + 1) << 3;
if (extlen > net->ipv6.sysctl.max_hbh_opts_len)
goto fail_and_free;
opt->flags |= IP6SKB_HOPBYHOP;
if (ip6_parse_tlv(true, skb, net->ipv6.sysctl.max_hbh_opts_cnt)) {
ipv6: Implement limits on Hop-by-Hop and Destination options RFC 8200 (IPv6) defines Hop-by-Hop options and Destination options extension headers. Both of these carry a list of TLVs which is only limited by the maximum length of the extension header (2048 bytes). By the spec a host must process all the TLVs in these options, however these could be used as a fairly obvious denial of service attack. I think this could in fact be a significant DOS vector on the Internet, one mitigating factor might be that many FWs drop all packets with EH (and obviously this is only IPv6) so an Internet wide attack might not be so effective (yet!). By my calculation, the worse case packet with TLVs in a standard 1500 byte MTU packet that would be processed by the stack contains 1282 invidual TLVs (including pad TLVS) or 724 two byte TLVs. I wrote a quick test program that floods a whole bunch of these packets to a host and sure enough there is substantial time spent in ip6_parse_tlv. These packets contain nothing but unknown TLVS (that are ignored), TLV padding, and bogus UDP header with zero payload length. 25.38% [kernel] [k] __fib6_clean_all 21.63% [kernel] [k] ip6_parse_tlv 4.21% [kernel] [k] __local_bh_enable_ip 2.18% [kernel] [k] ip6_pol_route.isra.39 1.98% [kernel] [k] fib6_walk_continue 1.88% [kernel] [k] _raw_write_lock_bh 1.65% [kernel] [k] dst_release This patch adds configurable limits to Destination and Hop-by-Hop options. There are three limits that may be set: - Limit the number of options in a Hop-by-Hop or Destination options extension header. - Limit the byte length of a Hop-by-Hop or Destination options extension header. - Disallow unrecognized options in a Hop-by-Hop or Destination options extension header. The limits are set in corresponding sysctls: ipv6.sysctl.max_dst_opts_cnt ipv6.sysctl.max_hbh_opts_cnt ipv6.sysctl.max_dst_opts_len ipv6.sysctl.max_hbh_opts_len If a max_*_opts_cnt is less than zero then unknown TLVs are disallowed. The number of known TLVs that are allowed is the absolute value of this number. If a limit is exceeded when processing an extension header the packet is dropped. Default values are set to 8 for options counts, and set to INT_MAX for maximum length. Note the choice to limit options to 8 is an arbitrary guess (roughly based on the fact that the stack supports three HBH options and just one destination option). These limits have being proposed in draft-ietf-6man-rfc6434-bis. Tested (by Martin Lau) I tested out 1 thread (i.e. one raw_udp process). I changed the net.ipv6.max_dst_(opts|hbh)_number between 8 to 2048. With sysctls setting to 2048, the softirq% is packed to 100%. With 8, the softirq% is almost unnoticable from mpstat. v2; - Code and documention cleanup. - Change references of RFC2460 to be RFC8200. - Add reference to RFC6434-bis where the limits will be in standard. Signed-off-by: Tom Herbert <tom@quantonium.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-31 05:16:00 +08:00
skb->transport_header += extlen;
opt = IP6CB(skb);
opt->nhoff = sizeof(struct ipv6hdr);
return 1;
}
return -1;
}
/*
* Creating outbound headers.
*
* "build" functions work when skb is filled from head to tail (datagram)
* "push" functions work when headers are added from tail to head (tcp)
*
* In both cases we assume, that caller reserved enough room
* for headers.
*/
static void ipv6_push_rthdr0(struct sk_buff *skb, u8 *proto,
struct ipv6_rt_hdr *opt,
struct in6_addr **addr_p, struct in6_addr *saddr)
{
struct rt0_hdr *phdr, *ihdr;
int hops;
ihdr = (struct rt0_hdr *) opt;
phdr = skb_push(skb, (ihdr->rt_hdr.hdrlen + 1) << 3);
memcpy(phdr, ihdr, sizeof(struct rt0_hdr));
hops = ihdr->rt_hdr.hdrlen >> 1;
if (hops > 1)
memcpy(phdr->addr, ihdr->addr + 1,
(hops - 1) * sizeof(struct in6_addr));
phdr->addr[hops - 1] = **addr_p;
*addr_p = ihdr->addr;
phdr->rt_hdr.nexthdr = *proto;
*proto = NEXTHDR_ROUTING;
}
static void ipv6_push_rthdr4(struct sk_buff *skb, u8 *proto,
struct ipv6_rt_hdr *opt,
struct in6_addr **addr_p, struct in6_addr *saddr)
{
struct ipv6_sr_hdr *sr_phdr, *sr_ihdr;
int plen, hops;
sr_ihdr = (struct ipv6_sr_hdr *)opt;
plen = (sr_ihdr->hdrlen + 1) << 3;
sr_phdr = skb_push(skb, plen);
memcpy(sr_phdr, sr_ihdr, sizeof(struct ipv6_sr_hdr));
hops = sr_ihdr->first_segment + 1;
memcpy(sr_phdr->segments + 1, sr_ihdr->segments + 1,
(hops - 1) * sizeof(struct in6_addr));
sr_phdr->segments[0] = **addr_p;
*addr_p = &sr_ihdr->segments[sr_ihdr->segments_left];
if (sr_ihdr->hdrlen > hops * 2) {
int tlvs_offset, tlvs_length;
tlvs_offset = (1 + hops * 2) << 3;
tlvs_length = (sr_ihdr->hdrlen - hops * 2) << 3;
memcpy((char *)sr_phdr + tlvs_offset,
(char *)sr_ihdr + tlvs_offset, tlvs_length);
}
#ifdef CONFIG_IPV6_SEG6_HMAC
if (sr_has_hmac(sr_phdr)) {
struct net *net = NULL;
if (skb->dev)
net = dev_net(skb->dev);
else if (skb->sk)
net = sock_net(skb->sk);
WARN_ON(!net);
if (net)
seg6_push_hmac(net, saddr, sr_phdr);
}
#endif
sr_phdr->nexthdr = *proto;
*proto = NEXTHDR_ROUTING;
}
static void ipv6_push_rthdr(struct sk_buff *skb, u8 *proto,
struct ipv6_rt_hdr *opt,
struct in6_addr **addr_p, struct in6_addr *saddr)
{
switch (opt->type) {
case IPV6_SRCRT_TYPE_0:
case IPV6_SRCRT_STRICT:
case IPV6_SRCRT_TYPE_2:
ipv6_push_rthdr0(skb, proto, opt, addr_p, saddr);
break;
case IPV6_SRCRT_TYPE_4:
ipv6_push_rthdr4(skb, proto, opt, addr_p, saddr);
break;
default:
break;
}
}
static void ipv6_push_exthdr(struct sk_buff *skb, u8 *proto, u8 type, struct ipv6_opt_hdr *opt)
{
struct ipv6_opt_hdr *h = skb_push(skb, ipv6_optlen(opt));
memcpy(h, opt, ipv6_optlen(opt));
h->nexthdr = *proto;
*proto = type;
}
void ipv6_push_nfrag_opts(struct sk_buff *skb, struct ipv6_txoptions *opt,
u8 *proto,
struct in6_addr **daddr, struct in6_addr *saddr)
{
if (opt->srcrt) {
ipv6_push_rthdr(skb, proto, opt->srcrt, daddr, saddr);
/*
* IPV6_RTHDRDSTOPTS is ignored
* unless IPV6_RTHDR is set (RFC3542).
*/
if (opt->dst0opt)
ipv6_push_exthdr(skb, proto, NEXTHDR_DEST, opt->dst0opt);
}
if (opt->hopopt)
ipv6_push_exthdr(skb, proto, NEXTHDR_HOP, opt->hopopt);
}
void ipv6_push_frag_opts(struct sk_buff *skb, struct ipv6_txoptions *opt, u8 *proto)
{
if (opt->dst1opt)
ipv6_push_exthdr(skb, proto, NEXTHDR_DEST, opt->dst1opt);
}
EXPORT_SYMBOL(ipv6_push_frag_opts);
struct ipv6_txoptions *
ipv6_dup_options(struct sock *sk, struct ipv6_txoptions *opt)
{
struct ipv6_txoptions *opt2;
opt2 = sock_kmalloc(sk, opt->tot_len, GFP_ATOMIC);
if (opt2) {
long dif = (char *)opt2 - (char *)opt;
memcpy(opt2, opt, opt->tot_len);
if (opt2->hopopt)
*((char **)&opt2->hopopt) += dif;
if (opt2->dst0opt)
*((char **)&opt2->dst0opt) += dif;
if (opt2->dst1opt)
*((char **)&opt2->dst1opt) += dif;
if (opt2->srcrt)
*((char **)&opt2->srcrt) += dif;
refcount_set(&opt2->refcnt, 1);
}
return opt2;
}
EXPORT_SYMBOL_GPL(ipv6_dup_options);
ipv6: make ipv6_renew_options() interrupt/kernel safe At present the ipv6_renew_options_kern() function ends up calling into access_ok() which is problematic if done from inside an interrupt as access_ok() calls WARN_ON_IN_IRQ() on some (all?) architectures (x86-64 is affected). Example warning/backtrace is shown below: WARNING: CPU: 1 PID: 3144 at lib/usercopy.c:11 _copy_from_user+0x85/0x90 ... Call Trace: <IRQ> ipv6_renew_option+0xb2/0xf0 ipv6_renew_options+0x26a/0x340 ipv6_renew_options_kern+0x2c/0x40 calipso_req_setattr+0x72/0xe0 netlbl_req_setattr+0x126/0x1b0 selinux_netlbl_inet_conn_request+0x80/0x100 selinux_inet_conn_request+0x6d/0xb0 security_inet_conn_request+0x32/0x50 tcp_conn_request+0x35f/0xe00 ? __lock_acquire+0x250/0x16c0 ? selinux_socket_sock_rcv_skb+0x1ae/0x210 ? tcp_rcv_state_process+0x289/0x106b tcp_rcv_state_process+0x289/0x106b ? tcp_v6_do_rcv+0x1a7/0x3c0 tcp_v6_do_rcv+0x1a7/0x3c0 tcp_v6_rcv+0xc82/0xcf0 ip6_input_finish+0x10d/0x690 ip6_input+0x45/0x1e0 ? ip6_rcv_finish+0x1d0/0x1d0 ipv6_rcv+0x32b/0x880 ? ip6_make_skb+0x1e0/0x1e0 __netif_receive_skb_core+0x6f2/0xdf0 ? process_backlog+0x85/0x250 ? process_backlog+0x85/0x250 ? process_backlog+0xec/0x250 process_backlog+0xec/0x250 net_rx_action+0x153/0x480 __do_softirq+0xd9/0x4f7 do_softirq_own_stack+0x2a/0x40 </IRQ> ... While not present in the backtrace, ipv6_renew_option() ends up calling access_ok() via the following chain: access_ok() _copy_from_user() copy_from_user() ipv6_renew_option() The fix presented in this patch is to perform the userspace copy earlier in the call chain such that it is only called when the option data is actually coming from userspace; that place is do_ipv6_setsockopt(). Not only does this solve the problem seen in the backtrace above, it also allows us to simplify the code quite a bit by removing ipv6_renew_options_kern() completely. We also take this opportunity to cleanup ipv6_renew_options()/ipv6_renew_option() a small amount as well. This patch is heavily based on a rough patch by Al Viro. I've taken his original patch, converted a kmemdup() call in do_ipv6_setsockopt() to a memdup_user() call, made better use of the e_inval jump target in the same function, and cleaned up the use ipv6_renew_option() by ipv6_renew_options(). CC: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Paul Moore <paul@paul-moore.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-04 21:58:05 +08:00
static void ipv6_renew_option(int renewtype,
struct ipv6_opt_hdr **dest,
struct ipv6_opt_hdr *old,
struct ipv6_opt_hdr *new,
int newtype, char **p)
{
ipv6: make ipv6_renew_options() interrupt/kernel safe At present the ipv6_renew_options_kern() function ends up calling into access_ok() which is problematic if done from inside an interrupt as access_ok() calls WARN_ON_IN_IRQ() on some (all?) architectures (x86-64 is affected). Example warning/backtrace is shown below: WARNING: CPU: 1 PID: 3144 at lib/usercopy.c:11 _copy_from_user+0x85/0x90 ... Call Trace: <IRQ> ipv6_renew_option+0xb2/0xf0 ipv6_renew_options+0x26a/0x340 ipv6_renew_options_kern+0x2c/0x40 calipso_req_setattr+0x72/0xe0 netlbl_req_setattr+0x126/0x1b0 selinux_netlbl_inet_conn_request+0x80/0x100 selinux_inet_conn_request+0x6d/0xb0 security_inet_conn_request+0x32/0x50 tcp_conn_request+0x35f/0xe00 ? __lock_acquire+0x250/0x16c0 ? selinux_socket_sock_rcv_skb+0x1ae/0x210 ? tcp_rcv_state_process+0x289/0x106b tcp_rcv_state_process+0x289/0x106b ? tcp_v6_do_rcv+0x1a7/0x3c0 tcp_v6_do_rcv+0x1a7/0x3c0 tcp_v6_rcv+0xc82/0xcf0 ip6_input_finish+0x10d/0x690 ip6_input+0x45/0x1e0 ? ip6_rcv_finish+0x1d0/0x1d0 ipv6_rcv+0x32b/0x880 ? ip6_make_skb+0x1e0/0x1e0 __netif_receive_skb_core+0x6f2/0xdf0 ? process_backlog+0x85/0x250 ? process_backlog+0x85/0x250 ? process_backlog+0xec/0x250 process_backlog+0xec/0x250 net_rx_action+0x153/0x480 __do_softirq+0xd9/0x4f7 do_softirq_own_stack+0x2a/0x40 </IRQ> ... While not present in the backtrace, ipv6_renew_option() ends up calling access_ok() via the following chain: access_ok() _copy_from_user() copy_from_user() ipv6_renew_option() The fix presented in this patch is to perform the userspace copy earlier in the call chain such that it is only called when the option data is actually coming from userspace; that place is do_ipv6_setsockopt(). Not only does this solve the problem seen in the backtrace above, it also allows us to simplify the code quite a bit by removing ipv6_renew_options_kern() completely. We also take this opportunity to cleanup ipv6_renew_options()/ipv6_renew_option() a small amount as well. This patch is heavily based on a rough patch by Al Viro. I've taken his original patch, converted a kmemdup() call in do_ipv6_setsockopt() to a memdup_user() call, made better use of the e_inval jump target in the same function, and cleaned up the use ipv6_renew_option() by ipv6_renew_options(). CC: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Paul Moore <paul@paul-moore.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-04 21:58:05 +08:00
struct ipv6_opt_hdr *src;
src = (renewtype == newtype ? new : old);
if (!src)
return;
memcpy(*p, src, ipv6_optlen(src));
*dest = (struct ipv6_opt_hdr *)*p;
*p += CMSG_ALIGN(ipv6_optlen(*dest));
}
/**
* ipv6_renew_options - replace a specific ext hdr with a new one.
*
* @sk: sock from which to allocate memory
* @opt: original options
* @newtype: option type to replace in @opt
* @newopt: new option of type @newtype to replace (user-mem)
*
* Returns a new set of options which is a copy of @opt with the
* option type @newtype replaced with @newopt.
*
* @opt may be NULL, in which case a new set of options is returned
* containing just @newopt.
*
* @newopt may be NULL, in which case the specified option type is
* not copied into the new set of options.
*
* The new set of options is allocated from the socket option memory
* buffer of @sk.
*/
struct ipv6_txoptions *
ipv6_renew_options(struct sock *sk, struct ipv6_txoptions *opt,
ipv6: make ipv6_renew_options() interrupt/kernel safe At present the ipv6_renew_options_kern() function ends up calling into access_ok() which is problematic if done from inside an interrupt as access_ok() calls WARN_ON_IN_IRQ() on some (all?) architectures (x86-64 is affected). Example warning/backtrace is shown below: WARNING: CPU: 1 PID: 3144 at lib/usercopy.c:11 _copy_from_user+0x85/0x90 ... Call Trace: <IRQ> ipv6_renew_option+0xb2/0xf0 ipv6_renew_options+0x26a/0x340 ipv6_renew_options_kern+0x2c/0x40 calipso_req_setattr+0x72/0xe0 netlbl_req_setattr+0x126/0x1b0 selinux_netlbl_inet_conn_request+0x80/0x100 selinux_inet_conn_request+0x6d/0xb0 security_inet_conn_request+0x32/0x50 tcp_conn_request+0x35f/0xe00 ? __lock_acquire+0x250/0x16c0 ? selinux_socket_sock_rcv_skb+0x1ae/0x210 ? tcp_rcv_state_process+0x289/0x106b tcp_rcv_state_process+0x289/0x106b ? tcp_v6_do_rcv+0x1a7/0x3c0 tcp_v6_do_rcv+0x1a7/0x3c0 tcp_v6_rcv+0xc82/0xcf0 ip6_input_finish+0x10d/0x690 ip6_input+0x45/0x1e0 ? ip6_rcv_finish+0x1d0/0x1d0 ipv6_rcv+0x32b/0x880 ? ip6_make_skb+0x1e0/0x1e0 __netif_receive_skb_core+0x6f2/0xdf0 ? process_backlog+0x85/0x250 ? process_backlog+0x85/0x250 ? process_backlog+0xec/0x250 process_backlog+0xec/0x250 net_rx_action+0x153/0x480 __do_softirq+0xd9/0x4f7 do_softirq_own_stack+0x2a/0x40 </IRQ> ... While not present in the backtrace, ipv6_renew_option() ends up calling access_ok() via the following chain: access_ok() _copy_from_user() copy_from_user() ipv6_renew_option() The fix presented in this patch is to perform the userspace copy earlier in the call chain such that it is only called when the option data is actually coming from userspace; that place is do_ipv6_setsockopt(). Not only does this solve the problem seen in the backtrace above, it also allows us to simplify the code quite a bit by removing ipv6_renew_options_kern() completely. We also take this opportunity to cleanup ipv6_renew_options()/ipv6_renew_option() a small amount as well. This patch is heavily based on a rough patch by Al Viro. I've taken his original patch, converted a kmemdup() call in do_ipv6_setsockopt() to a memdup_user() call, made better use of the e_inval jump target in the same function, and cleaned up the use ipv6_renew_option() by ipv6_renew_options(). CC: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Paul Moore <paul@paul-moore.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-04 21:58:05 +08:00
int newtype, struct ipv6_opt_hdr *newopt)
{
int tot_len = 0;
char *p;
struct ipv6_txoptions *opt2;
if (opt) {
if (newtype != IPV6_HOPOPTS && opt->hopopt)
tot_len += CMSG_ALIGN(ipv6_optlen(opt->hopopt));
if (newtype != IPV6_RTHDRDSTOPTS && opt->dst0opt)
tot_len += CMSG_ALIGN(ipv6_optlen(opt->dst0opt));
if (newtype != IPV6_RTHDR && opt->srcrt)
tot_len += CMSG_ALIGN(ipv6_optlen(opt->srcrt));
if (newtype != IPV6_DSTOPTS && opt->dst1opt)
tot_len += CMSG_ALIGN(ipv6_optlen(opt->dst1opt));
}
ipv6: make ipv6_renew_options() interrupt/kernel safe At present the ipv6_renew_options_kern() function ends up calling into access_ok() which is problematic if done from inside an interrupt as access_ok() calls WARN_ON_IN_IRQ() on some (all?) architectures (x86-64 is affected). Example warning/backtrace is shown below: WARNING: CPU: 1 PID: 3144 at lib/usercopy.c:11 _copy_from_user+0x85/0x90 ... Call Trace: <IRQ> ipv6_renew_option+0xb2/0xf0 ipv6_renew_options+0x26a/0x340 ipv6_renew_options_kern+0x2c/0x40 calipso_req_setattr+0x72/0xe0 netlbl_req_setattr+0x126/0x1b0 selinux_netlbl_inet_conn_request+0x80/0x100 selinux_inet_conn_request+0x6d/0xb0 security_inet_conn_request+0x32/0x50 tcp_conn_request+0x35f/0xe00 ? __lock_acquire+0x250/0x16c0 ? selinux_socket_sock_rcv_skb+0x1ae/0x210 ? tcp_rcv_state_process+0x289/0x106b tcp_rcv_state_process+0x289/0x106b ? tcp_v6_do_rcv+0x1a7/0x3c0 tcp_v6_do_rcv+0x1a7/0x3c0 tcp_v6_rcv+0xc82/0xcf0 ip6_input_finish+0x10d/0x690 ip6_input+0x45/0x1e0 ? ip6_rcv_finish+0x1d0/0x1d0 ipv6_rcv+0x32b/0x880 ? ip6_make_skb+0x1e0/0x1e0 __netif_receive_skb_core+0x6f2/0xdf0 ? process_backlog+0x85/0x250 ? process_backlog+0x85/0x250 ? process_backlog+0xec/0x250 process_backlog+0xec/0x250 net_rx_action+0x153/0x480 __do_softirq+0xd9/0x4f7 do_softirq_own_stack+0x2a/0x40 </IRQ> ... While not present in the backtrace, ipv6_renew_option() ends up calling access_ok() via the following chain: access_ok() _copy_from_user() copy_from_user() ipv6_renew_option() The fix presented in this patch is to perform the userspace copy earlier in the call chain such that it is only called when the option data is actually coming from userspace; that place is do_ipv6_setsockopt(). Not only does this solve the problem seen in the backtrace above, it also allows us to simplify the code quite a bit by removing ipv6_renew_options_kern() completely. We also take this opportunity to cleanup ipv6_renew_options()/ipv6_renew_option() a small amount as well. This patch is heavily based on a rough patch by Al Viro. I've taken his original patch, converted a kmemdup() call in do_ipv6_setsockopt() to a memdup_user() call, made better use of the e_inval jump target in the same function, and cleaned up the use ipv6_renew_option() by ipv6_renew_options(). CC: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Paul Moore <paul@paul-moore.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-04 21:58:05 +08:00
if (newopt)
tot_len += CMSG_ALIGN(ipv6_optlen(newopt));
if (!tot_len)
return NULL;
tot_len += sizeof(*opt2);
opt2 = sock_kmalloc(sk, tot_len, GFP_ATOMIC);
if (!opt2)
return ERR_PTR(-ENOBUFS);
memset(opt2, 0, tot_len);
refcount_set(&opt2->refcnt, 1);
opt2->tot_len = tot_len;
p = (char *)(opt2 + 1);
ipv6: make ipv6_renew_options() interrupt/kernel safe At present the ipv6_renew_options_kern() function ends up calling into access_ok() which is problematic if done from inside an interrupt as access_ok() calls WARN_ON_IN_IRQ() on some (all?) architectures (x86-64 is affected). Example warning/backtrace is shown below: WARNING: CPU: 1 PID: 3144 at lib/usercopy.c:11 _copy_from_user+0x85/0x90 ... Call Trace: <IRQ> ipv6_renew_option+0xb2/0xf0 ipv6_renew_options+0x26a/0x340 ipv6_renew_options_kern+0x2c/0x40 calipso_req_setattr+0x72/0xe0 netlbl_req_setattr+0x126/0x1b0 selinux_netlbl_inet_conn_request+0x80/0x100 selinux_inet_conn_request+0x6d/0xb0 security_inet_conn_request+0x32/0x50 tcp_conn_request+0x35f/0xe00 ? __lock_acquire+0x250/0x16c0 ? selinux_socket_sock_rcv_skb+0x1ae/0x210 ? tcp_rcv_state_process+0x289/0x106b tcp_rcv_state_process+0x289/0x106b ? tcp_v6_do_rcv+0x1a7/0x3c0 tcp_v6_do_rcv+0x1a7/0x3c0 tcp_v6_rcv+0xc82/0xcf0 ip6_input_finish+0x10d/0x690 ip6_input+0x45/0x1e0 ? ip6_rcv_finish+0x1d0/0x1d0 ipv6_rcv+0x32b/0x880 ? ip6_make_skb+0x1e0/0x1e0 __netif_receive_skb_core+0x6f2/0xdf0 ? process_backlog+0x85/0x250 ? process_backlog+0x85/0x250 ? process_backlog+0xec/0x250 process_backlog+0xec/0x250 net_rx_action+0x153/0x480 __do_softirq+0xd9/0x4f7 do_softirq_own_stack+0x2a/0x40 </IRQ> ... While not present in the backtrace, ipv6_renew_option() ends up calling access_ok() via the following chain: access_ok() _copy_from_user() copy_from_user() ipv6_renew_option() The fix presented in this patch is to perform the userspace copy earlier in the call chain such that it is only called when the option data is actually coming from userspace; that place is do_ipv6_setsockopt(). Not only does this solve the problem seen in the backtrace above, it also allows us to simplify the code quite a bit by removing ipv6_renew_options_kern() completely. We also take this opportunity to cleanup ipv6_renew_options()/ipv6_renew_option() a small amount as well. This patch is heavily based on a rough patch by Al Viro. I've taken his original patch, converted a kmemdup() call in do_ipv6_setsockopt() to a memdup_user() call, made better use of the e_inval jump target in the same function, and cleaned up the use ipv6_renew_option() by ipv6_renew_options(). CC: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Paul Moore <paul@paul-moore.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-07-04 21:58:05 +08:00
ipv6_renew_option(IPV6_HOPOPTS, &opt2->hopopt,
(opt ? opt->hopopt : NULL),
newopt, newtype, &p);
ipv6_renew_option(IPV6_RTHDRDSTOPTS, &opt2->dst0opt,
(opt ? opt->dst0opt : NULL),
newopt, newtype, &p);
ipv6_renew_option(IPV6_RTHDR,
(struct ipv6_opt_hdr **)&opt2->srcrt,
(opt ? (struct ipv6_opt_hdr *)opt->srcrt : NULL),
newopt, newtype, &p);
ipv6_renew_option(IPV6_DSTOPTS, &opt2->dst1opt,
(opt ? opt->dst1opt : NULL),
newopt, newtype, &p);
opt2->opt_nflen = (opt2->hopopt ? ipv6_optlen(opt2->hopopt) : 0) +
(opt2->dst0opt ? ipv6_optlen(opt2->dst0opt) : 0) +
(opt2->srcrt ? ipv6_optlen(opt2->srcrt) : 0);
opt2->opt_flen = (opt2->dst1opt ? ipv6_optlen(opt2->dst1opt) : 0);
return opt2;
}
struct ipv6_txoptions *__ipv6_fixup_options(struct ipv6_txoptions *opt_space,
struct ipv6_txoptions *opt)
{
/*
* ignore the dest before srcrt unless srcrt is being included.
* --yoshfuji
*/
if (opt->dst0opt && !opt->srcrt) {
if (opt_space != opt) {
memcpy(opt_space, opt, sizeof(*opt_space));
opt = opt_space;
}
opt->opt_nflen -= ipv6_optlen(opt->dst0opt);
opt->dst0opt = NULL;
}
return opt;
}
EXPORT_SYMBOL_GPL(__ipv6_fixup_options);
/**
* fl6_update_dst - update flowi destination address with info given
* by srcrt option, if any.
*
* @fl6: flowi6 for which daddr is to be updated
* @opt: struct ipv6_txoptions in which to look for srcrt opt
* @orig: copy of original daddr address if modified
*
* Returns NULL if no txoptions or no srcrt, otherwise returns orig
* and initial value of fl6->daddr set in orig
*/
struct in6_addr *fl6_update_dst(struct flowi6 *fl6,
const struct ipv6_txoptions *opt,
struct in6_addr *orig)
{
if (!opt || !opt->srcrt)
return NULL;
*orig = fl6->daddr;
switch (opt->srcrt->type) {
case IPV6_SRCRT_TYPE_0:
case IPV6_SRCRT_STRICT:
case IPV6_SRCRT_TYPE_2:
fl6->daddr = *((struct rt0_hdr *)opt->srcrt)->addr;
break;
case IPV6_SRCRT_TYPE_4:
{
struct ipv6_sr_hdr *srh = (struct ipv6_sr_hdr *)opt->srcrt;
fl6->daddr = srh->segments[srh->segments_left];
break;
}
default:
return NULL;
}
return orig;
}
EXPORT_SYMBOL_GPL(fl6_update_dst);