2
0
mirror of https://github.com/edk2-porting/linux-next.git synced 2024-12-28 15:13:55 +08:00
linux-next/net/rxrpc/rxkad.c

1406 lines
35 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-or-later
/* Kerberos-based RxRPC security
*
* Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
* Written by David Howells (dhowells@redhat.com)
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <crypto/skcipher.h>
#include <linux/module.h>
#include <linux/net.h>
#include <linux/skbuff.h>
#include <linux/udp.h>
#include <linux/scatterlist.h>
#include <linux/ctype.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/key-type.h>
#include <net/sock.h>
#include <net/af_rxrpc.h>
#include <keys/rxrpc-type.h>
#include "ar-internal.h"
#define RXKAD_VERSION 2
#define MAXKRB5TICKETLEN 1024
#define RXKAD_TKT_TYPE_KERBEROS_V5 256
#define ANAME_SZ 40 /* size of authentication name */
#define INST_SZ 40 /* size of principal's instance */
#define REALM_SZ 40 /* size of principal's auth domain */
#define SNAME_SZ 40 /* size of service name */
#define RXKAD_ALIGN 8
struct rxkad_level1_hdr {
__be32 data_size; /* true data size (excluding padding) */
};
struct rxkad_level2_hdr {
__be32 data_size; /* true data size (excluding padding) */
__be32 checksum; /* decrypted data checksum */
};
static int rxkad_prime_packet_security(struct rxrpc_connection *conn,
struct crypto_sync_skcipher *ci);
/*
* this holds a pinned cipher so that keventd doesn't get called by the cipher
* alloc routine, but since we have it to hand, we use it to decrypt RESPONSE
* packets
*/
static struct crypto_sync_skcipher *rxkad_ci;
static struct skcipher_request *rxkad_ci_req;
static DEFINE_MUTEX(rxkad_ci_mutex);
/*
* Parse the information from a server key
*
* The data should be the 8-byte secret key.
*/
static int rxkad_preparse_server_key(struct key_preparsed_payload *prep)
{
struct crypto_skcipher *ci;
if (prep->datalen != 8)
return -EINVAL;
memcpy(&prep->payload.data[2], prep->data, 8);
ci = crypto_alloc_skcipher("pcbc(des)", 0, CRYPTO_ALG_ASYNC);
if (IS_ERR(ci)) {
_leave(" = %ld", PTR_ERR(ci));
return PTR_ERR(ci);
}
if (crypto_skcipher_setkey(ci, prep->data, 8) < 0)
BUG();
prep->payload.data[0] = ci;
_leave(" = 0");
return 0;
}
static void rxkad_free_preparse_server_key(struct key_preparsed_payload *prep)
{
if (prep->payload.data[0])
crypto_free_skcipher(prep->payload.data[0]);
}
static void rxkad_destroy_server_key(struct key *key)
{
if (key->payload.data[0]) {
crypto_free_skcipher(key->payload.data[0]);
key->payload.data[0] = NULL;
}
}
/*
* initialise connection security
*/
static int rxkad_init_connection_security(struct rxrpc_connection *conn,
struct rxrpc_key_token *token)
{
struct crypto_sync_skcipher *ci;
int ret;
_enter("{%d},{%x}", conn->debug_id, key_serial(conn->params.key));
conn->security_ix = token->security_index;
ci = crypto_alloc_sync_skcipher("pcbc(fcrypt)", 0, 0);
if (IS_ERR(ci)) {
_debug("no cipher");
ret = PTR_ERR(ci);
goto error;
}
if (crypto_sync_skcipher_setkey(ci, token->kad->session_key,
sizeof(token->kad->session_key)) < 0)
BUG();
switch (conn->params.security_level) {
case RXRPC_SECURITY_PLAIN:
case RXRPC_SECURITY_AUTH:
case RXRPC_SECURITY_ENCRYPT:
break;
default:
ret = -EKEYREJECTED;
goto error;
}
ret = rxkad_prime_packet_security(conn, ci);
if (ret < 0)
goto error_ci;
conn->rxkad.cipher = ci;
return 0;
error_ci:
crypto_free_sync_skcipher(ci);
error:
_leave(" = %d", ret);
return ret;
}
/*
* Work out how much data we can put in a packet.
*/
static int rxkad_how_much_data(struct rxrpc_call *call, size_t remain,
size_t *_buf_size, size_t *_data_size, size_t *_offset)
{
size_t shdr, buf_size, chunk;
switch (call->conn->params.security_level) {
default:
buf_size = chunk = min_t(size_t, remain, RXRPC_JUMBO_DATALEN);
shdr = 0;
goto out;
case RXRPC_SECURITY_AUTH:
shdr = sizeof(struct rxkad_level1_hdr);
break;
case RXRPC_SECURITY_ENCRYPT:
shdr = sizeof(struct rxkad_level2_hdr);
break;
}
buf_size = round_down(RXRPC_JUMBO_DATALEN, RXKAD_ALIGN);
chunk = buf_size - shdr;
if (remain < chunk)
buf_size = round_up(shdr + remain, RXKAD_ALIGN);
out:
*_buf_size = buf_size;
*_data_size = chunk;
*_offset = shdr;
return 0;
}
/*
* prime the encryption state with the invariant parts of a connection's
* description
*/
static int rxkad_prime_packet_security(struct rxrpc_connection *conn,
struct crypto_sync_skcipher *ci)
{
struct skcipher_request *req;
struct rxrpc_key_token *token;
struct scatterlist sg;
struct rxrpc_crypt iv;
__be32 *tmpbuf;
size_t tmpsize = 4 * sizeof(__be32);
_enter("");
if (!conn->params.key)
return 0;
tmpbuf = kmalloc(tmpsize, GFP_KERNEL);
if (!tmpbuf)
return -ENOMEM;
req = skcipher_request_alloc(&ci->base, GFP_NOFS);
if (!req) {
kfree(tmpbuf);
return -ENOMEM;
}
token = conn->params.key->payload.data[0];
memcpy(&iv, token->kad->session_key, sizeof(iv));
tmpbuf[0] = htonl(conn->proto.epoch);
tmpbuf[1] = htonl(conn->proto.cid);
tmpbuf[2] = 0;
tmpbuf[3] = htonl(conn->security_ix);
sg_init_one(&sg, tmpbuf, tmpsize);
skcipher_request_set_sync_tfm(req, ci);
skcipher_request_set_callback(req, 0, NULL, NULL);
skcipher_request_set_crypt(req, &sg, &sg, tmpsize, iv.x);
crypto_skcipher_encrypt(req);
skcipher_request_free(req);
memcpy(&conn->rxkad.csum_iv, tmpbuf + 2, sizeof(conn->rxkad.csum_iv));
kfree(tmpbuf);
_leave(" = 0");
return 0;
}
/*
* Allocate and prepare the crypto request on a call. For any particular call,
* this is called serially for the packets, so no lock should be necessary.
*/
static struct skcipher_request *rxkad_get_call_crypto(struct rxrpc_call *call)
{
struct crypto_skcipher *tfm = &call->conn->rxkad.cipher->base;
struct skcipher_request *cipher_req = call->cipher_req;
if (!cipher_req) {
cipher_req = skcipher_request_alloc(tfm, GFP_NOFS);
if (!cipher_req)
return NULL;
call->cipher_req = cipher_req;
}
return cipher_req;
}
/*
* Clean up the crypto on a call.
*/
static void rxkad_free_call_crypto(struct rxrpc_call *call)
{
if (call->cipher_req)
skcipher_request_free(call->cipher_req);
call->cipher_req = NULL;
}
/*
* partially encrypt a packet (level 1 security)
*/
static int rxkad_secure_packet_auth(const struct rxrpc_call *call,
struct sk_buff *skb, u32 data_size,
struct skcipher_request *req)
{
struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
struct rxkad_level1_hdr hdr;
struct rxrpc_crypt iv;
struct scatterlist sg;
size_t pad;
u16 check;
_enter("");
check = sp->hdr.seq ^ call->call_id;
data_size |= (u32)check << 16;
hdr.data_size = htonl(data_size);
memcpy(skb->head, &hdr, sizeof(hdr));
pad = sizeof(struct rxkad_level1_hdr) + data_size;
pad = RXKAD_ALIGN - pad;
pad &= RXKAD_ALIGN - 1;
if (pad)
skb_put_zero(skb, pad);
/* start the encryption afresh */
memset(&iv, 0, sizeof(iv));
sg_init_one(&sg, skb->head, 8);
skcipher_request_set_sync_tfm(req, call->conn->rxkad.cipher);
skcipher_request_set_callback(req, 0, NULL, NULL);
skcipher_request_set_crypt(req, &sg, &sg, 8, iv.x);
crypto_skcipher_encrypt(req);
skcipher_request_zero(req);
_leave(" = 0");
return 0;
}
/*
* wholly encrypt a packet (level 2 security)
*/
static int rxkad_secure_packet_encrypt(const struct rxrpc_call *call,
struct sk_buff *skb,
u32 data_size,
struct skcipher_request *req)
{
const struct rxrpc_key_token *token;
struct rxkad_level2_hdr rxkhdr;
struct rxrpc_skb_priv *sp;
struct rxrpc_crypt iv;
struct scatterlist sg[16];
unsigned int len;
size_t pad;
u16 check;
int err;
sp = rxrpc_skb(skb);
_enter("");
check = sp->hdr.seq ^ call->call_id;
rxkhdr.data_size = htonl(data_size | (u32)check << 16);
rxkhdr.checksum = 0;
memcpy(skb->head, &rxkhdr, sizeof(rxkhdr));
pad = sizeof(struct rxkad_level2_hdr) + data_size;
pad = RXKAD_ALIGN - pad;
pad &= RXKAD_ALIGN - 1;
if (pad)
skb_put_zero(skb, pad);
/* encrypt from the session key */
token = call->conn->params.key->payload.data[0];
memcpy(&iv, token->kad->session_key, sizeof(iv));
sg_init_one(&sg[0], skb->head, sizeof(rxkhdr));
skcipher_request_set_sync_tfm(req, call->conn->rxkad.cipher);
skcipher_request_set_callback(req, 0, NULL, NULL);
skcipher_request_set_crypt(req, &sg[0], &sg[0], sizeof(rxkhdr), iv.x);
crypto_skcipher_encrypt(req);
/* we want to encrypt the skbuff in-place */
rxrpc: Use skb_unshare() rather than skb_cow_data() The in-place decryption routines in AF_RXRPC's rxkad security module currently call skb_cow_data() to make sure the data isn't shared and that the skb can be written over. This has a problem, however, as the softirq handler may be still holding a ref or the Rx ring may be holding multiple refs when skb_cow_data() is called in rxkad_verify_packet() - and so skb_shared() returns true and __pskb_pull_tail() dislikes that. If this occurs, something like the following report will be generated. kernel BUG at net/core/skbuff.c:1463! ... RIP: 0010:pskb_expand_head+0x253/0x2b0 ... Call Trace: __pskb_pull_tail+0x49/0x460 skb_cow_data+0x6f/0x300 rxkad_verify_packet+0x18b/0xb10 [rxrpc] rxrpc_recvmsg_data.isra.11+0x4a8/0xa10 [rxrpc] rxrpc_kernel_recv_data+0x126/0x240 [rxrpc] afs_extract_data+0x51/0x2d0 [kafs] afs_deliver_fs_fetch_data+0x188/0x400 [kafs] afs_deliver_to_call+0xac/0x430 [kafs] afs_wait_for_call_to_complete+0x22f/0x3d0 [kafs] afs_make_call+0x282/0x3f0 [kafs] afs_fs_fetch_data+0x164/0x300 [kafs] afs_fetch_data+0x54/0x130 [kafs] afs_readpages+0x20d/0x340 [kafs] read_pages+0x66/0x180 __do_page_cache_readahead+0x188/0x1a0 ondemand_readahead+0x17d/0x2e0 generic_file_read_iter+0x740/0xc10 __vfs_read+0x145/0x1a0 vfs_read+0x8c/0x140 ksys_read+0x4a/0xb0 do_syscall_64+0x43/0xf0 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Fix this by using skb_unshare() instead in the input path for DATA packets that have a security index != 0. Non-DATA packets don't need in-place encryption and neither do unencrypted DATA packets. Fixes: 248f219cb8bc ("rxrpc: Rewrite the data and ack handling code") Reported-by: Julian Wollrath <jwollrath@web.de> Signed-off-by: David Howells <dhowells@redhat.com>
2019-08-27 17:13:46 +08:00
err = -EMSGSIZE;
if (skb_shinfo(skb)->nr_frags > 16)
goto out;
len = round_up(data_size, RXKAD_ALIGN);
rxrpc: Use skb_unshare() rather than skb_cow_data() The in-place decryption routines in AF_RXRPC's rxkad security module currently call skb_cow_data() to make sure the data isn't shared and that the skb can be written over. This has a problem, however, as the softirq handler may be still holding a ref or the Rx ring may be holding multiple refs when skb_cow_data() is called in rxkad_verify_packet() - and so skb_shared() returns true and __pskb_pull_tail() dislikes that. If this occurs, something like the following report will be generated. kernel BUG at net/core/skbuff.c:1463! ... RIP: 0010:pskb_expand_head+0x253/0x2b0 ... Call Trace: __pskb_pull_tail+0x49/0x460 skb_cow_data+0x6f/0x300 rxkad_verify_packet+0x18b/0xb10 [rxrpc] rxrpc_recvmsg_data.isra.11+0x4a8/0xa10 [rxrpc] rxrpc_kernel_recv_data+0x126/0x240 [rxrpc] afs_extract_data+0x51/0x2d0 [kafs] afs_deliver_fs_fetch_data+0x188/0x400 [kafs] afs_deliver_to_call+0xac/0x430 [kafs] afs_wait_for_call_to_complete+0x22f/0x3d0 [kafs] afs_make_call+0x282/0x3f0 [kafs] afs_fs_fetch_data+0x164/0x300 [kafs] afs_fetch_data+0x54/0x130 [kafs] afs_readpages+0x20d/0x340 [kafs] read_pages+0x66/0x180 __do_page_cache_readahead+0x188/0x1a0 ondemand_readahead+0x17d/0x2e0 generic_file_read_iter+0x740/0xc10 __vfs_read+0x145/0x1a0 vfs_read+0x8c/0x140 ksys_read+0x4a/0xb0 do_syscall_64+0x43/0xf0 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Fix this by using skb_unshare() instead in the input path for DATA packets that have a security index != 0. Non-DATA packets don't need in-place encryption and neither do unencrypted DATA packets. Fixes: 248f219cb8bc ("rxrpc: Rewrite the data and ack handling code") Reported-by: Julian Wollrath <jwollrath@web.de> Signed-off-by: David Howells <dhowells@redhat.com>
2019-08-27 17:13:46 +08:00
sg_init_table(sg, ARRAY_SIZE(sg));
err = skb_to_sgvec(skb, sg, 8, len);
if (unlikely(err < 0))
goto out;
skcipher_request_set_crypt(req, sg, sg, len, iv.x);
crypto_skcipher_encrypt(req);
_leave(" = 0");
err = 0;
out:
skcipher_request_zero(req);
return err;
}
/*
* checksum an RxRPC packet header
*/
static int rxkad_secure_packet(struct rxrpc_call *call,
struct sk_buff *skb,
size_t data_size)
{
struct rxrpc_skb_priv *sp;
struct skcipher_request *req;
struct rxrpc_crypt iv;
struct scatterlist sg;
u32 x, y;
int ret;
sp = rxrpc_skb(skb);
_enter("{%d{%x}},{#%u},%zu,",
call->debug_id, key_serial(call->conn->params.key),
sp->hdr.seq, data_size);
if (!call->conn->rxkad.cipher)
return 0;
ret = key_validate(call->conn->params.key);
if (ret < 0)
return ret;
req = rxkad_get_call_crypto(call);
if (!req)
return -ENOMEM;
/* continue encrypting from where we left off */
memcpy(&iv, call->conn->rxkad.csum_iv.x, sizeof(iv));
/* calculate the security checksum */
x = (call->cid & RXRPC_CHANNELMASK) << (32 - RXRPC_CIDSHIFT);
x |= sp->hdr.seq & 0x3fffffff;
call->crypto_buf[0] = htonl(call->call_id);
call->crypto_buf[1] = htonl(x);
sg_init_one(&sg, call->crypto_buf, 8);
skcipher_request_set_sync_tfm(req, call->conn->rxkad.cipher);
skcipher_request_set_callback(req, 0, NULL, NULL);
skcipher_request_set_crypt(req, &sg, &sg, 8, iv.x);
crypto_skcipher_encrypt(req);
skcipher_request_zero(req);
y = ntohl(call->crypto_buf[1]);
y = (y >> 16) & 0xffff;
if (y == 0)
y = 1; /* zero checksums are not permitted */
sp->hdr.cksum = y;
switch (call->conn->params.security_level) {
case RXRPC_SECURITY_PLAIN:
ret = 0;
break;
case RXRPC_SECURITY_AUTH:
ret = rxkad_secure_packet_auth(call, skb, data_size, req);
break;
case RXRPC_SECURITY_ENCRYPT:
ret = rxkad_secure_packet_encrypt(call, skb, data_size, req);
break;
default:
ret = -EPERM;
break;
}
_leave(" = %d [set %hx]", ret, y);
return ret;
}
/*
* decrypt partial encryption on a packet (level 1 security)
*/
static int rxkad_verify_packet_1(struct rxrpc_call *call, struct sk_buff *skb,
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
unsigned int offset, unsigned int len,
rxrpc_seq_t seq,
struct skcipher_request *req)
{
struct rxkad_level1_hdr sechdr;
struct rxrpc_crypt iv;
struct scatterlist sg[16];
bool aborted;
u32 data_size, buf;
u16 check;
rxrpc: Use skb_unshare() rather than skb_cow_data() The in-place decryption routines in AF_RXRPC's rxkad security module currently call skb_cow_data() to make sure the data isn't shared and that the skb can be written over. This has a problem, however, as the softirq handler may be still holding a ref or the Rx ring may be holding multiple refs when skb_cow_data() is called in rxkad_verify_packet() - and so skb_shared() returns true and __pskb_pull_tail() dislikes that. If this occurs, something like the following report will be generated. kernel BUG at net/core/skbuff.c:1463! ... RIP: 0010:pskb_expand_head+0x253/0x2b0 ... Call Trace: __pskb_pull_tail+0x49/0x460 skb_cow_data+0x6f/0x300 rxkad_verify_packet+0x18b/0xb10 [rxrpc] rxrpc_recvmsg_data.isra.11+0x4a8/0xa10 [rxrpc] rxrpc_kernel_recv_data+0x126/0x240 [rxrpc] afs_extract_data+0x51/0x2d0 [kafs] afs_deliver_fs_fetch_data+0x188/0x400 [kafs] afs_deliver_to_call+0xac/0x430 [kafs] afs_wait_for_call_to_complete+0x22f/0x3d0 [kafs] afs_make_call+0x282/0x3f0 [kafs] afs_fs_fetch_data+0x164/0x300 [kafs] afs_fetch_data+0x54/0x130 [kafs] afs_readpages+0x20d/0x340 [kafs] read_pages+0x66/0x180 __do_page_cache_readahead+0x188/0x1a0 ondemand_readahead+0x17d/0x2e0 generic_file_read_iter+0x740/0xc10 __vfs_read+0x145/0x1a0 vfs_read+0x8c/0x140 ksys_read+0x4a/0xb0 do_syscall_64+0x43/0xf0 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Fix this by using skb_unshare() instead in the input path for DATA packets that have a security index != 0. Non-DATA packets don't need in-place encryption and neither do unencrypted DATA packets. Fixes: 248f219cb8bc ("rxrpc: Rewrite the data and ack handling code") Reported-by: Julian Wollrath <jwollrath@web.de> Signed-off-by: David Howells <dhowells@redhat.com>
2019-08-27 17:13:46 +08:00
int ret;
_enter("");
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
if (len < 8) {
aborted = rxrpc_abort_eproto(call, skb, "rxkad_1_hdr", "V1H",
RXKADSEALEDINCON);
goto protocol_error;
}
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
/* Decrypt the skbuff in-place. TODO: We really want to decrypt
* directly into the target buffer.
*/
rxrpc: Use skb_unshare() rather than skb_cow_data() The in-place decryption routines in AF_RXRPC's rxkad security module currently call skb_cow_data() to make sure the data isn't shared and that the skb can be written over. This has a problem, however, as the softirq handler may be still holding a ref or the Rx ring may be holding multiple refs when skb_cow_data() is called in rxkad_verify_packet() - and so skb_shared() returns true and __pskb_pull_tail() dislikes that. If this occurs, something like the following report will be generated. kernel BUG at net/core/skbuff.c:1463! ... RIP: 0010:pskb_expand_head+0x253/0x2b0 ... Call Trace: __pskb_pull_tail+0x49/0x460 skb_cow_data+0x6f/0x300 rxkad_verify_packet+0x18b/0xb10 [rxrpc] rxrpc_recvmsg_data.isra.11+0x4a8/0xa10 [rxrpc] rxrpc_kernel_recv_data+0x126/0x240 [rxrpc] afs_extract_data+0x51/0x2d0 [kafs] afs_deliver_fs_fetch_data+0x188/0x400 [kafs] afs_deliver_to_call+0xac/0x430 [kafs] afs_wait_for_call_to_complete+0x22f/0x3d0 [kafs] afs_make_call+0x282/0x3f0 [kafs] afs_fs_fetch_data+0x164/0x300 [kafs] afs_fetch_data+0x54/0x130 [kafs] afs_readpages+0x20d/0x340 [kafs] read_pages+0x66/0x180 __do_page_cache_readahead+0x188/0x1a0 ondemand_readahead+0x17d/0x2e0 generic_file_read_iter+0x740/0xc10 __vfs_read+0x145/0x1a0 vfs_read+0x8c/0x140 ksys_read+0x4a/0xb0 do_syscall_64+0x43/0xf0 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Fix this by using skb_unshare() instead in the input path for DATA packets that have a security index != 0. Non-DATA packets don't need in-place encryption and neither do unencrypted DATA packets. Fixes: 248f219cb8bc ("rxrpc: Rewrite the data and ack handling code") Reported-by: Julian Wollrath <jwollrath@web.de> Signed-off-by: David Howells <dhowells@redhat.com>
2019-08-27 17:13:46 +08:00
sg_init_table(sg, ARRAY_SIZE(sg));
ret = skb_to_sgvec(skb, sg, offset, 8);
if (unlikely(ret < 0))
return ret;
/* start the decryption afresh */
memset(&iv, 0, sizeof(iv));
skcipher_request_set_sync_tfm(req, call->conn->rxkad.cipher);
skcipher_request_set_callback(req, 0, NULL, NULL);
skcipher_request_set_crypt(req, sg, sg, 8, iv.x);
crypto_skcipher_decrypt(req);
skcipher_request_zero(req);
/* Extract the decrypted packet length */
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
if (skb_copy_bits(skb, offset, &sechdr, sizeof(sechdr)) < 0) {
aborted = rxrpc_abort_eproto(call, skb, "rxkad_1_len", "XV1",
RXKADDATALEN);
goto protocol_error;
}
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
len -= sizeof(sechdr);
buf = ntohl(sechdr.data_size);
data_size = buf & 0xffff;
check = buf >> 16;
check ^= seq ^ call->call_id;
check &= 0xffff;
if (check != 0) {
aborted = rxrpc_abort_eproto(call, skb, "rxkad_1_check", "V1C",
RXKADSEALEDINCON);
goto protocol_error;
}
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
if (data_size > len) {
aborted = rxrpc_abort_eproto(call, skb, "rxkad_1_datalen", "V1L",
RXKADDATALEN);
goto protocol_error;
}
_leave(" = 0 [dlen=%x]", data_size);
return 0;
protocol_error:
if (aborted)
rxrpc_send_abort_packet(call);
return -EPROTO;
}
/*
* wholly decrypt a packet (level 2 security)
*/
static int rxkad_verify_packet_2(struct rxrpc_call *call, struct sk_buff *skb,
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
unsigned int offset, unsigned int len,
rxrpc_seq_t seq,
struct skcipher_request *req)
{
const struct rxrpc_key_token *token;
struct rxkad_level2_hdr sechdr;
struct rxrpc_crypt iv;
struct scatterlist _sg[4], *sg;
bool aborted;
u32 data_size, buf;
u16 check;
int nsg, ret;
_enter(",{%d}", skb->len);
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
if (len < 8) {
aborted = rxrpc_abort_eproto(call, skb, "rxkad_2_hdr", "V2H",
RXKADSEALEDINCON);
goto protocol_error;
}
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
/* Decrypt the skbuff in-place. TODO: We really want to decrypt
* directly into the target buffer.
*/
sg = _sg;
rxrpc: Use skb_unshare() rather than skb_cow_data() The in-place decryption routines in AF_RXRPC's rxkad security module currently call skb_cow_data() to make sure the data isn't shared and that the skb can be written over. This has a problem, however, as the softirq handler may be still holding a ref or the Rx ring may be holding multiple refs when skb_cow_data() is called in rxkad_verify_packet() - and so skb_shared() returns true and __pskb_pull_tail() dislikes that. If this occurs, something like the following report will be generated. kernel BUG at net/core/skbuff.c:1463! ... RIP: 0010:pskb_expand_head+0x253/0x2b0 ... Call Trace: __pskb_pull_tail+0x49/0x460 skb_cow_data+0x6f/0x300 rxkad_verify_packet+0x18b/0xb10 [rxrpc] rxrpc_recvmsg_data.isra.11+0x4a8/0xa10 [rxrpc] rxrpc_kernel_recv_data+0x126/0x240 [rxrpc] afs_extract_data+0x51/0x2d0 [kafs] afs_deliver_fs_fetch_data+0x188/0x400 [kafs] afs_deliver_to_call+0xac/0x430 [kafs] afs_wait_for_call_to_complete+0x22f/0x3d0 [kafs] afs_make_call+0x282/0x3f0 [kafs] afs_fs_fetch_data+0x164/0x300 [kafs] afs_fetch_data+0x54/0x130 [kafs] afs_readpages+0x20d/0x340 [kafs] read_pages+0x66/0x180 __do_page_cache_readahead+0x188/0x1a0 ondemand_readahead+0x17d/0x2e0 generic_file_read_iter+0x740/0xc10 __vfs_read+0x145/0x1a0 vfs_read+0x8c/0x140 ksys_read+0x4a/0xb0 do_syscall_64+0x43/0xf0 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Fix this by using skb_unshare() instead in the input path for DATA packets that have a security index != 0. Non-DATA packets don't need in-place encryption and neither do unencrypted DATA packets. Fixes: 248f219cb8bc ("rxrpc: Rewrite the data and ack handling code") Reported-by: Julian Wollrath <jwollrath@web.de> Signed-off-by: David Howells <dhowells@redhat.com>
2019-08-27 17:13:46 +08:00
nsg = skb_shinfo(skb)->nr_frags;
if (nsg <= 4) {
nsg = 4;
} else {
treewide: kmalloc() -> kmalloc_array() The kmalloc() function has a 2-factor argument form, kmalloc_array(). This patch replaces cases of: kmalloc(a * b, gfp) with: kmalloc_array(a * b, gfp) as well as handling cases of: kmalloc(a * b * c, gfp) with: kmalloc(array3_size(a, b, c), gfp) as it's slightly less ugly than: kmalloc_array(array_size(a, b), c, gfp) This does, however, attempt to ignore constant size factors like: kmalloc(4 * 1024, gfp) though any constants defined via macros get caught up in the conversion. Any factors with a sizeof() of "unsigned char", "char", and "u8" were dropped, since they're redundant. The tools/ directory was manually excluded, since it has its own implementation of kmalloc(). The Coccinelle script used for this was: // Fix redundant parens around sizeof(). @@ type TYPE; expression THING, E; @@ ( kmalloc( - (sizeof(TYPE)) * E + sizeof(TYPE) * E , ...) | kmalloc( - (sizeof(THING)) * E + sizeof(THING) * E , ...) ) // Drop single-byte sizes and redundant parens. @@ expression COUNT; typedef u8; typedef __u8; @@ ( kmalloc( - sizeof(u8) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(__u8) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(char) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(unsigned char) * (COUNT) + COUNT , ...) | kmalloc( - sizeof(u8) * COUNT + COUNT , ...) | kmalloc( - sizeof(__u8) * COUNT + COUNT , ...) | kmalloc( - sizeof(char) * COUNT + COUNT , ...) | kmalloc( - sizeof(unsigned char) * COUNT + COUNT , ...) ) // 2-factor product with sizeof(type/expression) and identifier or constant. @@ type TYPE; expression THING; identifier COUNT_ID; constant COUNT_CONST; @@ ( - kmalloc + kmalloc_array ( - sizeof(TYPE) * (COUNT_ID) + COUNT_ID, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * COUNT_ID + COUNT_ID, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * (COUNT_CONST) + COUNT_CONST, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * COUNT_CONST + COUNT_CONST, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * (COUNT_ID) + COUNT_ID, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * COUNT_ID + COUNT_ID, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * (COUNT_CONST) + COUNT_CONST, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * COUNT_CONST + COUNT_CONST, sizeof(THING) , ...) ) // 2-factor product, only identifiers. @@ identifier SIZE, COUNT; @@ - kmalloc + kmalloc_array ( - SIZE * COUNT + COUNT, SIZE , ...) // 3-factor product with 1 sizeof(type) or sizeof(expression), with // redundant parens removed. @@ expression THING; identifier STRIDE, COUNT; type TYPE; @@ ( kmalloc( - sizeof(TYPE) * (COUNT) * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(TYPE) * (COUNT) * STRIDE + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(TYPE) * COUNT * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(TYPE) * COUNT * STRIDE + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kmalloc( - sizeof(THING) * (COUNT) * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kmalloc( - sizeof(THING) * (COUNT) * STRIDE + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kmalloc( - sizeof(THING) * COUNT * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kmalloc( - sizeof(THING) * COUNT * STRIDE + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) ) // 3-factor product with 2 sizeof(variable), with redundant parens removed. @@ expression THING1, THING2; identifier COUNT; type TYPE1, TYPE2; @@ ( kmalloc( - sizeof(TYPE1) * sizeof(TYPE2) * COUNT + array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2)) , ...) | kmalloc( - sizeof(TYPE1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2)) , ...) | kmalloc( - sizeof(THING1) * sizeof(THING2) * COUNT + array3_size(COUNT, sizeof(THING1), sizeof(THING2)) , ...) | kmalloc( - sizeof(THING1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(THING1), sizeof(THING2)) , ...) | kmalloc( - sizeof(TYPE1) * sizeof(THING2) * COUNT + array3_size(COUNT, sizeof(TYPE1), sizeof(THING2)) , ...) | kmalloc( - sizeof(TYPE1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(TYPE1), sizeof(THING2)) , ...) ) // 3-factor product, only identifiers, with redundant parens removed. @@ identifier STRIDE, SIZE, COUNT; @@ ( kmalloc( - (COUNT) * STRIDE * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * (STRIDE) * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * STRIDE * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - (COUNT) * (STRIDE) * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * (STRIDE) * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - (COUNT) * STRIDE * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - (COUNT) * (STRIDE) * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kmalloc( - COUNT * STRIDE * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) ) // Any remaining multi-factor products, first at least 3-factor products, // when they're not all constants... @@ expression E1, E2, E3; constant C1, C2, C3; @@ ( kmalloc(C1 * C2 * C3, ...) | kmalloc( - (E1) * E2 * E3 + array3_size(E1, E2, E3) , ...) | kmalloc( - (E1) * (E2) * E3 + array3_size(E1, E2, E3) , ...) | kmalloc( - (E1) * (E2) * (E3) + array3_size(E1, E2, E3) , ...) | kmalloc( - E1 * E2 * E3 + array3_size(E1, E2, E3) , ...) ) // And then all remaining 2 factors products when they're not all constants, // keeping sizeof() as the second factor argument. @@ expression THING, E1, E2; type TYPE; constant C1, C2, C3; @@ ( kmalloc(sizeof(THING) * C2, ...) | kmalloc(sizeof(TYPE) * C2, ...) | kmalloc(C1 * C2 * C3, ...) | kmalloc(C1 * C2, ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * (E2) + E2, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(TYPE) * E2 + E2, sizeof(TYPE) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * (E2) + E2, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - sizeof(THING) * E2 + E2, sizeof(THING) , ...) | - kmalloc + kmalloc_array ( - (E1) * E2 + E1, E2 , ...) | - kmalloc + kmalloc_array ( - (E1) * (E2) + E1, E2 , ...) | - kmalloc + kmalloc_array ( - E1 * E2 + E1, E2 , ...) ) Signed-off-by: Kees Cook <keescook@chromium.org>
2018-06-13 04:55:00 +08:00
sg = kmalloc_array(nsg, sizeof(*sg), GFP_NOIO);
if (!sg)
goto nomem;
}
sg_init_table(sg, nsg);
ret = skb_to_sgvec(skb, sg, offset, len);
if (unlikely(ret < 0)) {
if (sg != _sg)
kfree(sg);
return ret;
}
/* decrypt from the session key */
token = call->conn->params.key->payload.data[0];
memcpy(&iv, token->kad->session_key, sizeof(iv));
skcipher_request_set_sync_tfm(req, call->conn->rxkad.cipher);
skcipher_request_set_callback(req, 0, NULL, NULL);
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
skcipher_request_set_crypt(req, sg, sg, len, iv.x);
crypto_skcipher_decrypt(req);
skcipher_request_zero(req);
if (sg != _sg)
kfree(sg);
/* Extract the decrypted packet length */
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
if (skb_copy_bits(skb, offset, &sechdr, sizeof(sechdr)) < 0) {
aborted = rxrpc_abort_eproto(call, skb, "rxkad_2_len", "XV2",
RXKADDATALEN);
goto protocol_error;
}
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
len -= sizeof(sechdr);
buf = ntohl(sechdr.data_size);
data_size = buf & 0xffff;
check = buf >> 16;
check ^= seq ^ call->call_id;
check &= 0xffff;
if (check != 0) {
aborted = rxrpc_abort_eproto(call, skb, "rxkad_2_check", "V2C",
RXKADSEALEDINCON);
goto protocol_error;
}
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
if (data_size > len) {
aborted = rxrpc_abort_eproto(call, skb, "rxkad_2_datalen", "V2L",
RXKADDATALEN);
goto protocol_error;
}
_leave(" = 0 [dlen=%x]", data_size);
return 0;
protocol_error:
if (aborted)
rxrpc_send_abort_packet(call);
return -EPROTO;
nomem:
_leave(" = -ENOMEM");
return -ENOMEM;
}
/*
* Verify the security on a received packet or subpacket (if part of a
* jumbo packet).
*/
static int rxkad_verify_packet(struct rxrpc_call *call, struct sk_buff *skb,
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
unsigned int offset, unsigned int len,
rxrpc_seq_t seq, u16 expected_cksum)
{
struct skcipher_request *req;
struct rxrpc_crypt iv;
struct scatterlist sg;
bool aborted;
u16 cksum;
u32 x, y;
_enter("{%d{%x}},{#%u}",
call->debug_id, key_serial(call->conn->params.key), seq);
if (!call->conn->rxkad.cipher)
return 0;
req = rxkad_get_call_crypto(call);
if (!req)
return -ENOMEM;
/* continue encrypting from where we left off */
memcpy(&iv, call->conn->rxkad.csum_iv.x, sizeof(iv));
/* validate the security checksum */
x = (call->cid & RXRPC_CHANNELMASK) << (32 - RXRPC_CIDSHIFT);
x |= seq & 0x3fffffff;
call->crypto_buf[0] = htonl(call->call_id);
call->crypto_buf[1] = htonl(x);
sg_init_one(&sg, call->crypto_buf, 8);
skcipher_request_set_sync_tfm(req, call->conn->rxkad.cipher);
skcipher_request_set_callback(req, 0, NULL, NULL);
skcipher_request_set_crypt(req, &sg, &sg, 8, iv.x);
crypto_skcipher_encrypt(req);
skcipher_request_zero(req);
y = ntohl(call->crypto_buf[1]);
cksum = (y >> 16) & 0xffff;
if (cksum == 0)
cksum = 1; /* zero checksums are not permitted */
if (cksum != expected_cksum) {
aborted = rxrpc_abort_eproto(call, skb, "rxkad_csum", "VCK",
RXKADSEALEDINCON);
goto protocol_error;
}
switch (call->conn->params.security_level) {
case RXRPC_SECURITY_PLAIN:
return 0;
case RXRPC_SECURITY_AUTH:
return rxkad_verify_packet_1(call, skb, offset, len, seq, req);
case RXRPC_SECURITY_ENCRYPT:
return rxkad_verify_packet_2(call, skb, offset, len, seq, req);
default:
return -ENOANO;
}
protocol_error:
if (aborted)
rxrpc_send_abort_packet(call);
return -EPROTO;
}
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
/*
* Locate the data contained in a packet that was partially encrypted.
*/
static void rxkad_locate_data_1(struct rxrpc_call *call, struct sk_buff *skb,
unsigned int *_offset, unsigned int *_len)
{
struct rxkad_level1_hdr sechdr;
if (skb_copy_bits(skb, *_offset, &sechdr, sizeof(sechdr)) < 0)
BUG();
*_offset += sizeof(sechdr);
*_len = ntohl(sechdr.data_size) & 0xffff;
}
/*
* Locate the data contained in a packet that was completely encrypted.
*/
static void rxkad_locate_data_2(struct rxrpc_call *call, struct sk_buff *skb,
unsigned int *_offset, unsigned int *_len)
{
struct rxkad_level2_hdr sechdr;
if (skb_copy_bits(skb, *_offset, &sechdr, sizeof(sechdr)) < 0)
BUG();
*_offset += sizeof(sechdr);
*_len = ntohl(sechdr.data_size) & 0xffff;
}
/*
* Locate the data contained in an already decrypted packet.
*/
static void rxkad_locate_data(struct rxrpc_call *call, struct sk_buff *skb,
unsigned int *_offset, unsigned int *_len)
{
switch (call->conn->params.security_level) {
case RXRPC_SECURITY_AUTH:
rxkad_locate_data_1(call, skb, _offset, _len);
return;
case RXRPC_SECURITY_ENCRYPT:
rxkad_locate_data_2(call, skb, _offset, _len);
return;
default:
return;
}
}
/*
* issue a challenge
*/
static int rxkad_issue_challenge(struct rxrpc_connection *conn)
{
struct rxkad_challenge challenge;
struct rxrpc_wire_header whdr;
struct msghdr msg;
struct kvec iov[2];
size_t len;
u32 serial;
int ret;
_enter("{%d}", conn->debug_id);
get_random_bytes(&conn->rxkad.nonce, sizeof(conn->rxkad.nonce));
challenge.version = htonl(2);
challenge.nonce = htonl(conn->rxkad.nonce);
challenge.min_level = htonl(0);
challenge.__padding = 0;
msg.msg_name = &conn->params.peer->srx.transport;
msg.msg_namelen = conn->params.peer->srx.transport_len;
msg.msg_control = NULL;
msg.msg_controllen = 0;
msg.msg_flags = 0;
whdr.epoch = htonl(conn->proto.epoch);
whdr.cid = htonl(conn->proto.cid);
whdr.callNumber = 0;
whdr.seq = 0;
whdr.type = RXRPC_PACKET_TYPE_CHALLENGE;
whdr.flags = conn->out_clientflag;
whdr.userStatus = 0;
whdr.securityIndex = conn->security_ix;
whdr._rsvd = 0;
whdr.serviceId = htons(conn->service_id);
iov[0].iov_base = &whdr;
iov[0].iov_len = sizeof(whdr);
iov[1].iov_base = &challenge;
iov[1].iov_len = sizeof(challenge);
len = iov[0].iov_len + iov[1].iov_len;
serial = atomic_inc_return(&conn->serial);
whdr.serial = htonl(serial);
_proto("Tx CHALLENGE %%%u", serial);
ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 2, len);
if (ret < 0) {
trace_rxrpc_tx_fail(conn->debug_id, serial, ret,
rxrpc_tx_point_rxkad_challenge);
return -EAGAIN;
}
rxrpc: Fix the keepalive generator [ver #2] AF_RXRPC has a keepalive message generator that generates a message for a peer ~20s after the last transmission to that peer to keep firewall ports open. The implementation is incorrect in the following ways: (1) It mixes up ktime_t and time64_t types. (2) It uses ktime_get_real(), the output of which may jump forward or backward due to adjustments to the time of day. (3) If the current time jumps forward too much or jumps backwards, the generator function will crank the base of the time ring round one slot at a time (ie. a 1s period) until it catches up, spewing out VERSION packets as it goes. Fix the problem by: (1) Only using time64_t. There's no need for sub-second resolution. (2) Use ktime_get_seconds() rather than ktime_get_real() so that time isn't perceived to go backwards. (3) Simplifying rxrpc_peer_keepalive_worker() by splitting it into two parts: (a) The "worker" function that manages the buckets and the timer. (b) The "dispatch" function that takes the pending peers and potentially transmits a keepalive packet before putting them back in the ring into the slot appropriate to the revised last-Tx time. (4) Taking everything that's pending out of the ring and splicing it into a temporary collector list for processing. In the case that there's been a significant jump forward, the ring gets entirely emptied and then the time base can be warped forward before the peers are processed. The warping can't happen if the ring isn't empty because the slot a peer is in is keepalive-time dependent, relative to the base time. (5) Limit the number of iterations of the bucket array when scanning it. (6) Set the timer to skip any empty slots as there's no point waking up if there's nothing to do yet. This can be triggered by an incoming call from a server after a reboot with AF_RXRPC and AFS built into the kernel causing a peer record to be set up before userspace is started. The system clock is then adjusted by userspace, thereby potentially causing the keepalive generator to have a meltdown - which leads to a message like: watchdog: BUG: soft lockup - CPU#0 stuck for 23s! [kworker/0:1:23] ... Workqueue: krxrpcd rxrpc_peer_keepalive_worker EIP: lock_acquire+0x69/0x80 ... Call Trace: ? rxrpc_peer_keepalive_worker+0x5e/0x350 ? _raw_spin_lock_bh+0x29/0x60 ? rxrpc_peer_keepalive_worker+0x5e/0x350 ? rxrpc_peer_keepalive_worker+0x5e/0x350 ? __lock_acquire+0x3d3/0x870 ? process_one_work+0x110/0x340 ? process_one_work+0x166/0x340 ? process_one_work+0x110/0x340 ? worker_thread+0x39/0x3c0 ? kthread+0xdb/0x110 ? cancel_delayed_work+0x90/0x90 ? kthread_stop+0x70/0x70 ? ret_from_fork+0x19/0x24 Fixes: ace45bec6d77 ("rxrpc: Fix firewall route keepalive") Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-08-08 18:30:02 +08:00
conn->params.peer->last_tx_at = ktime_get_seconds();
trace_rxrpc_tx_packet(conn->debug_id, &whdr,
rxrpc_tx_point_rxkad_challenge);
_leave(" = 0");
return 0;
}
/*
* send a Kerberos security response
*/
static int rxkad_send_response(struct rxrpc_connection *conn,
struct rxrpc_host_header *hdr,
struct rxkad_response *resp,
const struct rxkad_key *s2)
{
struct rxrpc_wire_header whdr;
struct msghdr msg;
struct kvec iov[3];
size_t len;
u32 serial;
int ret;
_enter("");
msg.msg_name = &conn->params.peer->srx.transport;
msg.msg_namelen = conn->params.peer->srx.transport_len;
msg.msg_control = NULL;
msg.msg_controllen = 0;
msg.msg_flags = 0;
memset(&whdr, 0, sizeof(whdr));
whdr.epoch = htonl(hdr->epoch);
whdr.cid = htonl(hdr->cid);
whdr.type = RXRPC_PACKET_TYPE_RESPONSE;
whdr.flags = conn->out_clientflag;
whdr.securityIndex = hdr->securityIndex;
whdr.serviceId = htons(hdr->serviceId);
iov[0].iov_base = &whdr;
iov[0].iov_len = sizeof(whdr);
iov[1].iov_base = resp;
iov[1].iov_len = sizeof(*resp);
iov[2].iov_base = (void *)s2->ticket;
iov[2].iov_len = s2->ticket_len;
len = iov[0].iov_len + iov[1].iov_len + iov[2].iov_len;
serial = atomic_inc_return(&conn->serial);
whdr.serial = htonl(serial);
_proto("Tx RESPONSE %%%u", serial);
ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 3, len);
if (ret < 0) {
trace_rxrpc_tx_fail(conn->debug_id, serial, ret,
rxrpc_tx_point_rxkad_response);
return -EAGAIN;
}
rxrpc: Fix the keepalive generator [ver #2] AF_RXRPC has a keepalive message generator that generates a message for a peer ~20s after the last transmission to that peer to keep firewall ports open. The implementation is incorrect in the following ways: (1) It mixes up ktime_t and time64_t types. (2) It uses ktime_get_real(), the output of which may jump forward or backward due to adjustments to the time of day. (3) If the current time jumps forward too much or jumps backwards, the generator function will crank the base of the time ring round one slot at a time (ie. a 1s period) until it catches up, spewing out VERSION packets as it goes. Fix the problem by: (1) Only using time64_t. There's no need for sub-second resolution. (2) Use ktime_get_seconds() rather than ktime_get_real() so that time isn't perceived to go backwards. (3) Simplifying rxrpc_peer_keepalive_worker() by splitting it into two parts: (a) The "worker" function that manages the buckets and the timer. (b) The "dispatch" function that takes the pending peers and potentially transmits a keepalive packet before putting them back in the ring into the slot appropriate to the revised last-Tx time. (4) Taking everything that's pending out of the ring and splicing it into a temporary collector list for processing. In the case that there's been a significant jump forward, the ring gets entirely emptied and then the time base can be warped forward before the peers are processed. The warping can't happen if the ring isn't empty because the slot a peer is in is keepalive-time dependent, relative to the base time. (5) Limit the number of iterations of the bucket array when scanning it. (6) Set the timer to skip any empty slots as there's no point waking up if there's nothing to do yet. This can be triggered by an incoming call from a server after a reboot with AF_RXRPC and AFS built into the kernel causing a peer record to be set up before userspace is started. The system clock is then adjusted by userspace, thereby potentially causing the keepalive generator to have a meltdown - which leads to a message like: watchdog: BUG: soft lockup - CPU#0 stuck for 23s! [kworker/0:1:23] ... Workqueue: krxrpcd rxrpc_peer_keepalive_worker EIP: lock_acquire+0x69/0x80 ... Call Trace: ? rxrpc_peer_keepalive_worker+0x5e/0x350 ? _raw_spin_lock_bh+0x29/0x60 ? rxrpc_peer_keepalive_worker+0x5e/0x350 ? rxrpc_peer_keepalive_worker+0x5e/0x350 ? __lock_acquire+0x3d3/0x870 ? process_one_work+0x110/0x340 ? process_one_work+0x166/0x340 ? process_one_work+0x110/0x340 ? worker_thread+0x39/0x3c0 ? kthread+0xdb/0x110 ? cancel_delayed_work+0x90/0x90 ? kthread_stop+0x70/0x70 ? ret_from_fork+0x19/0x24 Fixes: ace45bec6d77 ("rxrpc: Fix firewall route keepalive") Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-08-08 18:30:02 +08:00
conn->params.peer->last_tx_at = ktime_get_seconds();
_leave(" = 0");
return 0;
}
/*
* calculate the response checksum
*/
static void rxkad_calc_response_checksum(struct rxkad_response *response)
{
u32 csum = 1000003;
int loop;
u8 *p = (u8 *) response;
for (loop = sizeof(*response); loop > 0; loop--)
csum = csum * 0x10204081 + *p++;
response->encrypted.checksum = htonl(csum);
}
/*
* encrypt the response packet
*/
static int rxkad_encrypt_response(struct rxrpc_connection *conn,
struct rxkad_response *resp,
const struct rxkad_key *s2)
{
struct skcipher_request *req;
struct rxrpc_crypt iv;
struct scatterlist sg[1];
req = skcipher_request_alloc(&conn->rxkad.cipher->base, GFP_NOFS);
if (!req)
return -ENOMEM;
/* continue encrypting from where we left off */
memcpy(&iv, s2->session_key, sizeof(iv));
sg_init_table(sg, 1);
sg_set_buf(sg, &resp->encrypted, sizeof(resp->encrypted));
skcipher_request_set_sync_tfm(req, conn->rxkad.cipher);
skcipher_request_set_callback(req, 0, NULL, NULL);
skcipher_request_set_crypt(req, sg, sg, sizeof(resp->encrypted), iv.x);
crypto_skcipher_encrypt(req);
skcipher_request_free(req);
return 0;
}
/*
* respond to a challenge packet
*/
static int rxkad_respond_to_challenge(struct rxrpc_connection *conn,
struct sk_buff *skb,
u32 *_abort_code)
{
const struct rxrpc_key_token *token;
struct rxkad_challenge challenge;
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
struct rxkad_response *resp;
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
const char *eproto;
u32 version, nonce, min_level, abort_code;
int ret;
_enter("{%d,%x}", conn->debug_id, key_serial(conn->params.key));
eproto = tracepoint_string("chall_no_key");
abort_code = RX_PROTOCOL_ERROR;
if (!conn->params.key)
goto protocol_error;
abort_code = RXKADEXPIRED;
ret = key_validate(conn->params.key);
if (ret < 0)
goto other_error;
eproto = tracepoint_string("chall_short");
abort_code = RXKADPACKETSHORT;
if (skb_copy_bits(skb, sizeof(struct rxrpc_wire_header),
&challenge, sizeof(challenge)) < 0)
goto protocol_error;
version = ntohl(challenge.version);
nonce = ntohl(challenge.nonce);
min_level = ntohl(challenge.min_level);
_proto("Rx CHALLENGE %%%u { v=%u n=%u ml=%u }",
sp->hdr.serial, version, nonce, min_level);
eproto = tracepoint_string("chall_ver");
abort_code = RXKADINCONSISTENCY;
if (version != RXKAD_VERSION)
goto protocol_error;
abort_code = RXKADLEVELFAIL;
ret = -EACCES;
if (conn->params.security_level < min_level)
goto other_error;
token = conn->params.key->payload.data[0];
/* build the response packet */
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
resp = kzalloc(sizeof(struct rxkad_response), GFP_NOFS);
if (!resp)
return -ENOMEM;
resp->version = htonl(RXKAD_VERSION);
resp->encrypted.epoch = htonl(conn->proto.epoch);
resp->encrypted.cid = htonl(conn->proto.cid);
resp->encrypted.securityIndex = htonl(conn->security_ix);
resp->encrypted.inc_nonce = htonl(nonce + 1);
resp->encrypted.level = htonl(conn->params.security_level);
resp->kvno = htonl(token->kad->kvno);
resp->ticket_len = htonl(token->kad->ticket_len);
resp->encrypted.call_id[0] = htonl(conn->channels[0].call_counter);
resp->encrypted.call_id[1] = htonl(conn->channels[1].call_counter);
resp->encrypted.call_id[2] = htonl(conn->channels[2].call_counter);
resp->encrypted.call_id[3] = htonl(conn->channels[3].call_counter);
/* calculate the response checksum and then do the encryption */
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
rxkad_calc_response_checksum(resp);
ret = rxkad_encrypt_response(conn, resp, token->kad);
if (ret == 0)
ret = rxkad_send_response(conn, &sp->hdr, resp, token->kad);
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
kfree(resp);
return ret;
protocol_error:
trace_rxrpc_rx_eproto(NULL, sp->hdr.serial, eproto);
ret = -EPROTO;
other_error:
*_abort_code = abort_code;
return ret;
}
/*
* decrypt the kerberos IV ticket in the response
*/
static int rxkad_decrypt_ticket(struct rxrpc_connection *conn,
struct key *server_key,
struct sk_buff *skb,
void *ticket, size_t ticket_len,
struct rxrpc_crypt *_session_key,
time64_t *_expiry,
u32 *_abort_code)
{
struct skcipher_request *req;
struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
struct rxrpc_crypt iv, key;
struct scatterlist sg[1];
struct in_addr addr;
unsigned int life;
const char *eproto;
time64_t issue, now;
bool little_endian;
int ret;
u32 abort_code;
u8 *p, *q, *name, *end;
_enter("{%d},{%x}", conn->debug_id, key_serial(server_key));
*_expiry = 0;
ASSERT(server_key->payload.data[0] != NULL);
ASSERTCMP((unsigned long) ticket & 7UL, ==, 0);
memcpy(&iv, &server_key->payload.data[2], sizeof(iv));
ret = -ENOMEM;
req = skcipher_request_alloc(server_key->payload.data[0], GFP_NOFS);
if (!req)
goto temporary_error;
sg_init_one(&sg[0], ticket, ticket_len);
skcipher_request_set_callback(req, 0, NULL, NULL);
skcipher_request_set_crypt(req, sg, sg, ticket_len, iv.x);
crypto_skcipher_decrypt(req);
skcipher_request_free(req);
p = ticket;
end = p + ticket_len;
#define Z(field) \
({ \
u8 *__str = p; \
eproto = tracepoint_string("rxkad_bad_"#field); \
q = memchr(p, 0, end - p); \
if (!q || q - p > (field##_SZ)) \
goto bad_ticket; \
for (; p < q; p++) \
if (!isprint(*p)) \
goto bad_ticket; \
p++; \
__str; \
})
/* extract the ticket flags */
_debug("KIV FLAGS: %x", *p);
little_endian = *p & 1;
p++;
/* extract the authentication name */
name = Z(ANAME);
_debug("KIV ANAME: %s", name);
/* extract the principal's instance */
name = Z(INST);
_debug("KIV INST : %s", name);
/* extract the principal's authentication domain */
name = Z(REALM);
_debug("KIV REALM: %s", name);
eproto = tracepoint_string("rxkad_bad_len");
if (end - p < 4 + 8 + 4 + 2)
goto bad_ticket;
/* get the IPv4 address of the entity that requested the ticket */
memcpy(&addr, p, sizeof(addr));
p += 4;
_debug("KIV ADDR : %pI4", &addr);
/* get the session key from the ticket */
memcpy(&key, p, sizeof(key));
p += 8;
_debug("KIV KEY : %08x %08x", ntohl(key.n[0]), ntohl(key.n[1]));
memcpy(_session_key, &key, sizeof(key));
/* get the ticket's lifetime */
life = *p++ * 5 * 60;
_debug("KIV LIFE : %u", life);
/* get the issue time of the ticket */
if (little_endian) {
__le32 stamp;
memcpy(&stamp, p, 4);
issue = rxrpc_u32_to_time64(le32_to_cpu(stamp));
} else {
__be32 stamp;
memcpy(&stamp, p, 4);
issue = rxrpc_u32_to_time64(be32_to_cpu(stamp));
}
p += 4;
now = ktime_get_real_seconds();
_debug("KIV ISSUE: %llx [%llx]", issue, now);
/* check the ticket is in date */
if (issue > now) {
abort_code = RXKADNOAUTH;
ret = -EKEYREJECTED;
goto other_error;
}
if (issue < now - life) {
abort_code = RXKADEXPIRED;
ret = -EKEYEXPIRED;
goto other_error;
}
*_expiry = issue + life;
/* get the service name */
name = Z(SNAME);
_debug("KIV SNAME: %s", name);
/* get the service instance name */
name = Z(INST);
_debug("KIV SINST: %s", name);
return 0;
bad_ticket:
trace_rxrpc_rx_eproto(NULL, sp->hdr.serial, eproto);
abort_code = RXKADBADTICKET;
ret = -EPROTO;
other_error:
*_abort_code = abort_code;
return ret;
temporary_error:
return ret;
}
/*
* decrypt the response packet
*/
static void rxkad_decrypt_response(struct rxrpc_connection *conn,
struct rxkad_response *resp,
const struct rxrpc_crypt *session_key)
{
struct skcipher_request *req = rxkad_ci_req;
struct scatterlist sg[1];
struct rxrpc_crypt iv;
_enter(",,%08x%08x",
ntohl(session_key->n[0]), ntohl(session_key->n[1]));
mutex_lock(&rxkad_ci_mutex);
if (crypto_sync_skcipher_setkey(rxkad_ci, session_key->x,
sizeof(*session_key)) < 0)
BUG();
memcpy(&iv, session_key, sizeof(iv));
sg_init_table(sg, 1);
sg_set_buf(sg, &resp->encrypted, sizeof(resp->encrypted));
skcipher_request_set_sync_tfm(req, rxkad_ci);
skcipher_request_set_callback(req, 0, NULL, NULL);
skcipher_request_set_crypt(req, sg, sg, sizeof(resp->encrypted), iv.x);
crypto_skcipher_decrypt(req);
skcipher_request_zero(req);
mutex_unlock(&rxkad_ci_mutex);
_leave("");
}
/*
* verify a response
*/
static int rxkad_verify_response(struct rxrpc_connection *conn,
struct sk_buff *skb,
u32 *_abort_code)
{
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
struct rxkad_response *response;
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
struct rxrpc_crypt session_key;
struct key *server_key;
const char *eproto;
time64_t expiry;
void *ticket;
u32 abort_code, version, kvno, ticket_len, level;
__be32 csum;
rxrpc: Call channels should have separate call number spaces Each channel on a connection has a separate, independent number space from which to allocate callNumber values. It is entirely possible, for example, to have a connection with four active calls, each with call number 1. Note that the callNumber values for any particular channel don't have to start at 1, but they are supposed to increment monotonically for that channel from a client's perspective and may not be reused once the call number is transmitted (until the epoch cycles all the way back round). Currently, however, call numbers are allocated on a per-connection basis and, further, are held in an rb-tree. The rb-tree is redundant as the four channel pointers in the rxrpc_connection struct are entirely capable of pointing to all the calls currently in progress on a connection. To this end, make the following changes: (1) Handle call number allocation independently per channel. (2) Get rid of the conn->calls rb-tree. This is overkill as a connection may have a maximum of four calls in progress at any one time. Use the pointers in the channels[] array instead, indexed by the channel number from the packet. (3) For each channel, save the result of the last call that was in progress on that channel in conn->channels[] so that the final ACK or ABORT packet can be replayed if necessary. Any call earlier than that is just ignored. If we've seen the next call number in a packet, the last one is most definitely defunct. (4) When generating a RESPONSE packet for a connection, the call number counter for each channel must be included in it. (5) When parsing a RESPONSE packet for a connection, the call number counters contained therein should be used to set the minimum expected call numbers on each channel. To do in future commits: (1) Replay terminal packets based on the last call stored in conn->channels[]. (2) Connections should be retired before the callNumber space on any channel runs out. (3) A server is expected to disregard or reject any new incoming call that has a call number less than the current call number counter. The call number counter for that channel must be advanced to the new call number. Note that the server cannot just require that the next call that it sees on a channel be exactly the call number counter + 1 because then there's a scenario that could cause a problem: The client transmits a packet to initiate a connection, the network goes out, the server sends an ACK (which gets lost), the client sends an ABORT (which also gets lost); the network then reconnects, the client then reuses the call number for the next call (it doesn't know the server already saw the call number), but the server thinks it already has the first packet of this call (it doesn't know that the client doesn't know that it saw the call number the first time). Signed-off-by: David Howells <dhowells@redhat.com>
2016-06-27 21:39:44 +08:00
int ret, i;
_enter("{%d}", conn->debug_id);
server_key = rxrpc_look_up_server_security(conn, skb, 0, 0);
if (IS_ERR(server_key)) {
switch (PTR_ERR(server_key)) {
case -ENOKEY:
abort_code = RXKADUNKNOWNKEY;
break;
case -EKEYEXPIRED:
abort_code = RXKADEXPIRED;
break;
default:
abort_code = RXKADNOAUTH;
break;
}
trace_rxrpc_abort(0, "SVK",
sp->hdr.cid, sp->hdr.callNumber, sp->hdr.seq,
abort_code, PTR_ERR(server_key));
*_abort_code = abort_code;
return -EPROTO;
}
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
ret = -ENOMEM;
response = kzalloc(sizeof(struct rxkad_response), GFP_NOFS);
if (!response)
goto temporary_error;
eproto = tracepoint_string("rxkad_rsp_short");
abort_code = RXKADPACKETSHORT;
if (skb_copy_bits(skb, sizeof(struct rxrpc_wire_header),
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
response, sizeof(*response)) < 0)
goto protocol_error;
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
version = ntohl(response->version);
ticket_len = ntohl(response->ticket_len);
kvno = ntohl(response->kvno);
_proto("Rx RESPONSE %%%u { v=%u kv=%u tl=%u }",
sp->hdr.serial, version, kvno, ticket_len);
eproto = tracepoint_string("rxkad_rsp_ver");
abort_code = RXKADINCONSISTENCY;
if (version != RXKAD_VERSION)
goto protocol_error;
eproto = tracepoint_string("rxkad_rsp_tktlen");
abort_code = RXKADTICKETLEN;
if (ticket_len < 4 || ticket_len > MAXKRB5TICKETLEN)
goto protocol_error;
eproto = tracepoint_string("rxkad_rsp_unkkey");
abort_code = RXKADUNKNOWNKEY;
if (kvno >= RXKAD_TKT_TYPE_KERBEROS_V5)
goto protocol_error;
/* extract the kerberos ticket and decrypt and decode it */
ret = -ENOMEM;
ticket = kmalloc(ticket_len, GFP_NOFS);
if (!ticket)
goto temporary_error_free_resp;
eproto = tracepoint_string("rxkad_tkt_short");
abort_code = RXKADPACKETSHORT;
if (skb_copy_bits(skb, sizeof(struct rxrpc_wire_header) + sizeof(*response),
ticket, ticket_len) < 0)
goto protocol_error_free;
ret = rxkad_decrypt_ticket(conn, server_key, skb, ticket, ticket_len,
&session_key, &expiry, _abort_code);
if (ret < 0)
goto temporary_error_free_ticket;
/* use the session key from inside the ticket to decrypt the
* response */
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
rxkad_decrypt_response(conn, response, &session_key);
eproto = tracepoint_string("rxkad_rsp_param");
abort_code = RXKADSEALEDINCON;
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
if (ntohl(response->encrypted.epoch) != conn->proto.epoch)
goto protocol_error_free;
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
if (ntohl(response->encrypted.cid) != conn->proto.cid)
goto protocol_error_free;
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
if (ntohl(response->encrypted.securityIndex) != conn->security_ix)
goto protocol_error_free;
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
csum = response->encrypted.checksum;
response->encrypted.checksum = 0;
rxkad_calc_response_checksum(response);
eproto = tracepoint_string("rxkad_rsp_csum");
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
if (response->encrypted.checksum != csum)
goto protocol_error_free;
rxrpc: Rewrite the client connection manager Rewrite the rxrpc client connection manager so that it can support multiple connections for a given security key to a peer. The following changes are made: (1) For each open socket, the code currently maintains an rbtree with the connections placed into it, keyed by communications parameters. This is tricky to maintain as connections can be culled from the tree or replaced within it. Connections can require replacement for a number of reasons, e.g. their IDs span too great a range for the IDR data type to represent efficiently, the call ID numbers on that conn would overflow or the conn got aborted. This is changed so that there's now a connection bundle object placed in the tree, keyed on the same parameters. The bundle, however, does not need to be replaced. (2) An rxrpc_bundle object can now manage the available channels for a set of parallel connections. The lock that manages this is moved there from the rxrpc_connection struct (channel_lock). (3) There'a a dummy bundle for all incoming connections to share so that they have a channel_lock too. It might be better to give each incoming connection its own bundle. This bundle is not needed to manage which channels incoming calls are made on because that's the solely at whim of the client. (4) The restrictions on how many client connections are around are removed. Instead, a previous patch limits the number of client calls that can be allocated. Ordinarily, client connections are reaped after 2 minutes on the idle queue, but when more than a certain number of connections are in existence, the reaper starts reaping them after 2s of idleness instead to get the numbers back down. It could also be made such that new call allocations are forced to wait until the number of outstanding connections subsides. Signed-off-by: David Howells <dhowells@redhat.com>
2020-07-01 18:15:32 +08:00
spin_lock(&conn->bundle->channel_lock);
rxrpc: Call channels should have separate call number spaces Each channel on a connection has a separate, independent number space from which to allocate callNumber values. It is entirely possible, for example, to have a connection with four active calls, each with call number 1. Note that the callNumber values for any particular channel don't have to start at 1, but they are supposed to increment monotonically for that channel from a client's perspective and may not be reused once the call number is transmitted (until the epoch cycles all the way back round). Currently, however, call numbers are allocated on a per-connection basis and, further, are held in an rb-tree. The rb-tree is redundant as the four channel pointers in the rxrpc_connection struct are entirely capable of pointing to all the calls currently in progress on a connection. To this end, make the following changes: (1) Handle call number allocation independently per channel. (2) Get rid of the conn->calls rb-tree. This is overkill as a connection may have a maximum of four calls in progress at any one time. Use the pointers in the channels[] array instead, indexed by the channel number from the packet. (3) For each channel, save the result of the last call that was in progress on that channel in conn->channels[] so that the final ACK or ABORT packet can be replayed if necessary. Any call earlier than that is just ignored. If we've seen the next call number in a packet, the last one is most definitely defunct. (4) When generating a RESPONSE packet for a connection, the call number counter for each channel must be included in it. (5) When parsing a RESPONSE packet for a connection, the call number counters contained therein should be used to set the minimum expected call numbers on each channel. To do in future commits: (1) Replay terminal packets based on the last call stored in conn->channels[]. (2) Connections should be retired before the callNumber space on any channel runs out. (3) A server is expected to disregard or reject any new incoming call that has a call number less than the current call number counter. The call number counter for that channel must be advanced to the new call number. Note that the server cannot just require that the next call that it sees on a channel be exactly the call number counter + 1 because then there's a scenario that could cause a problem: The client transmits a packet to initiate a connection, the network goes out, the server sends an ACK (which gets lost), the client sends an ABORT (which also gets lost); the network then reconnects, the client then reuses the call number for the next call (it doesn't know the server already saw the call number), but the server thinks it already has the first packet of this call (it doesn't know that the client doesn't know that it saw the call number the first time). Signed-off-by: David Howells <dhowells@redhat.com>
2016-06-27 21:39:44 +08:00
for (i = 0; i < RXRPC_MAXCALLS; i++) {
struct rxrpc_call *call;
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
u32 call_id = ntohl(response->encrypted.call_id[i]);
rxrpc: Call channels should have separate call number spaces Each channel on a connection has a separate, independent number space from which to allocate callNumber values. It is entirely possible, for example, to have a connection with four active calls, each with call number 1. Note that the callNumber values for any particular channel don't have to start at 1, but they are supposed to increment monotonically for that channel from a client's perspective and may not be reused once the call number is transmitted (until the epoch cycles all the way back round). Currently, however, call numbers are allocated on a per-connection basis and, further, are held in an rb-tree. The rb-tree is redundant as the four channel pointers in the rxrpc_connection struct are entirely capable of pointing to all the calls currently in progress on a connection. To this end, make the following changes: (1) Handle call number allocation independently per channel. (2) Get rid of the conn->calls rb-tree. This is overkill as a connection may have a maximum of four calls in progress at any one time. Use the pointers in the channels[] array instead, indexed by the channel number from the packet. (3) For each channel, save the result of the last call that was in progress on that channel in conn->channels[] so that the final ACK or ABORT packet can be replayed if necessary. Any call earlier than that is just ignored. If we've seen the next call number in a packet, the last one is most definitely defunct. (4) When generating a RESPONSE packet for a connection, the call number counter for each channel must be included in it. (5) When parsing a RESPONSE packet for a connection, the call number counters contained therein should be used to set the minimum expected call numbers on each channel. To do in future commits: (1) Replay terminal packets based on the last call stored in conn->channels[]. (2) Connections should be retired before the callNumber space on any channel runs out. (3) A server is expected to disregard or reject any new incoming call that has a call number less than the current call number counter. The call number counter for that channel must be advanced to the new call number. Note that the server cannot just require that the next call that it sees on a channel be exactly the call number counter + 1 because then there's a scenario that could cause a problem: The client transmits a packet to initiate a connection, the network goes out, the server sends an ACK (which gets lost), the client sends an ABORT (which also gets lost); the network then reconnects, the client then reuses the call number for the next call (it doesn't know the server already saw the call number), but the server thinks it already has the first packet of this call (it doesn't know that the client doesn't know that it saw the call number the first time). Signed-off-by: David Howells <dhowells@redhat.com>
2016-06-27 21:39:44 +08:00
eproto = tracepoint_string("rxkad_rsp_callid");
rxrpc: Call channels should have separate call number spaces Each channel on a connection has a separate, independent number space from which to allocate callNumber values. It is entirely possible, for example, to have a connection with four active calls, each with call number 1. Note that the callNumber values for any particular channel don't have to start at 1, but they are supposed to increment monotonically for that channel from a client's perspective and may not be reused once the call number is transmitted (until the epoch cycles all the way back round). Currently, however, call numbers are allocated on a per-connection basis and, further, are held in an rb-tree. The rb-tree is redundant as the four channel pointers in the rxrpc_connection struct are entirely capable of pointing to all the calls currently in progress on a connection. To this end, make the following changes: (1) Handle call number allocation independently per channel. (2) Get rid of the conn->calls rb-tree. This is overkill as a connection may have a maximum of four calls in progress at any one time. Use the pointers in the channels[] array instead, indexed by the channel number from the packet. (3) For each channel, save the result of the last call that was in progress on that channel in conn->channels[] so that the final ACK or ABORT packet can be replayed if necessary. Any call earlier than that is just ignored. If we've seen the next call number in a packet, the last one is most definitely defunct. (4) When generating a RESPONSE packet for a connection, the call number counter for each channel must be included in it. (5) When parsing a RESPONSE packet for a connection, the call number counters contained therein should be used to set the minimum expected call numbers on each channel. To do in future commits: (1) Replay terminal packets based on the last call stored in conn->channels[]. (2) Connections should be retired before the callNumber space on any channel runs out. (3) A server is expected to disregard or reject any new incoming call that has a call number less than the current call number counter. The call number counter for that channel must be advanced to the new call number. Note that the server cannot just require that the next call that it sees on a channel be exactly the call number counter + 1 because then there's a scenario that could cause a problem: The client transmits a packet to initiate a connection, the network goes out, the server sends an ACK (which gets lost), the client sends an ABORT (which also gets lost); the network then reconnects, the client then reuses the call number for the next call (it doesn't know the server already saw the call number), but the server thinks it already has the first packet of this call (it doesn't know that the client doesn't know that it saw the call number the first time). Signed-off-by: David Howells <dhowells@redhat.com>
2016-06-27 21:39:44 +08:00
if (call_id > INT_MAX)
goto protocol_error_unlock;
eproto = tracepoint_string("rxkad_rsp_callctr");
rxrpc: Call channels should have separate call number spaces Each channel on a connection has a separate, independent number space from which to allocate callNumber values. It is entirely possible, for example, to have a connection with four active calls, each with call number 1. Note that the callNumber values for any particular channel don't have to start at 1, but they are supposed to increment monotonically for that channel from a client's perspective and may not be reused once the call number is transmitted (until the epoch cycles all the way back round). Currently, however, call numbers are allocated on a per-connection basis and, further, are held in an rb-tree. The rb-tree is redundant as the four channel pointers in the rxrpc_connection struct are entirely capable of pointing to all the calls currently in progress on a connection. To this end, make the following changes: (1) Handle call number allocation independently per channel. (2) Get rid of the conn->calls rb-tree. This is overkill as a connection may have a maximum of four calls in progress at any one time. Use the pointers in the channels[] array instead, indexed by the channel number from the packet. (3) For each channel, save the result of the last call that was in progress on that channel in conn->channels[] so that the final ACK or ABORT packet can be replayed if necessary. Any call earlier than that is just ignored. If we've seen the next call number in a packet, the last one is most definitely defunct. (4) When generating a RESPONSE packet for a connection, the call number counter for each channel must be included in it. (5) When parsing a RESPONSE packet for a connection, the call number counters contained therein should be used to set the minimum expected call numbers on each channel. To do in future commits: (1) Replay terminal packets based on the last call stored in conn->channels[]. (2) Connections should be retired before the callNumber space on any channel runs out. (3) A server is expected to disregard or reject any new incoming call that has a call number less than the current call number counter. The call number counter for that channel must be advanced to the new call number. Note that the server cannot just require that the next call that it sees on a channel be exactly the call number counter + 1 because then there's a scenario that could cause a problem: The client transmits a packet to initiate a connection, the network goes out, the server sends an ACK (which gets lost), the client sends an ABORT (which also gets lost); the network then reconnects, the client then reuses the call number for the next call (it doesn't know the server already saw the call number), but the server thinks it already has the first packet of this call (it doesn't know that the client doesn't know that it saw the call number the first time). Signed-off-by: David Howells <dhowells@redhat.com>
2016-06-27 21:39:44 +08:00
if (call_id < conn->channels[i].call_counter)
goto protocol_error_unlock;
eproto = tracepoint_string("rxkad_rsp_callst");
rxrpc: Call channels should have separate call number spaces Each channel on a connection has a separate, independent number space from which to allocate callNumber values. It is entirely possible, for example, to have a connection with four active calls, each with call number 1. Note that the callNumber values for any particular channel don't have to start at 1, but they are supposed to increment monotonically for that channel from a client's perspective and may not be reused once the call number is transmitted (until the epoch cycles all the way back round). Currently, however, call numbers are allocated on a per-connection basis and, further, are held in an rb-tree. The rb-tree is redundant as the four channel pointers in the rxrpc_connection struct are entirely capable of pointing to all the calls currently in progress on a connection. To this end, make the following changes: (1) Handle call number allocation independently per channel. (2) Get rid of the conn->calls rb-tree. This is overkill as a connection may have a maximum of four calls in progress at any one time. Use the pointers in the channels[] array instead, indexed by the channel number from the packet. (3) For each channel, save the result of the last call that was in progress on that channel in conn->channels[] so that the final ACK or ABORT packet can be replayed if necessary. Any call earlier than that is just ignored. If we've seen the next call number in a packet, the last one is most definitely defunct. (4) When generating a RESPONSE packet for a connection, the call number counter for each channel must be included in it. (5) When parsing a RESPONSE packet for a connection, the call number counters contained therein should be used to set the minimum expected call numbers on each channel. To do in future commits: (1) Replay terminal packets based on the last call stored in conn->channels[]. (2) Connections should be retired before the callNumber space on any channel runs out. (3) A server is expected to disregard or reject any new incoming call that has a call number less than the current call number counter. The call number counter for that channel must be advanced to the new call number. Note that the server cannot just require that the next call that it sees on a channel be exactly the call number counter + 1 because then there's a scenario that could cause a problem: The client transmits a packet to initiate a connection, the network goes out, the server sends an ACK (which gets lost), the client sends an ABORT (which also gets lost); the network then reconnects, the client then reuses the call number for the next call (it doesn't know the server already saw the call number), but the server thinks it already has the first packet of this call (it doesn't know that the client doesn't know that it saw the call number the first time). Signed-off-by: David Howells <dhowells@redhat.com>
2016-06-27 21:39:44 +08:00
if (call_id > conn->channels[i].call_counter) {
call = rcu_dereference_protected(
conn->channels[i].call,
rxrpc: Rewrite the client connection manager Rewrite the rxrpc client connection manager so that it can support multiple connections for a given security key to a peer. The following changes are made: (1) For each open socket, the code currently maintains an rbtree with the connections placed into it, keyed by communications parameters. This is tricky to maintain as connections can be culled from the tree or replaced within it. Connections can require replacement for a number of reasons, e.g. their IDs span too great a range for the IDR data type to represent efficiently, the call ID numbers on that conn would overflow or the conn got aborted. This is changed so that there's now a connection bundle object placed in the tree, keyed on the same parameters. The bundle, however, does not need to be replaced. (2) An rxrpc_bundle object can now manage the available channels for a set of parallel connections. The lock that manages this is moved there from the rxrpc_connection struct (channel_lock). (3) There'a a dummy bundle for all incoming connections to share so that they have a channel_lock too. It might be better to give each incoming connection its own bundle. This bundle is not needed to manage which channels incoming calls are made on because that's the solely at whim of the client. (4) The restrictions on how many client connections are around are removed. Instead, a previous patch limits the number of client calls that can be allocated. Ordinarily, client connections are reaped after 2 minutes on the idle queue, but when more than a certain number of connections are in existence, the reaper starts reaping them after 2s of idleness instead to get the numbers back down. It could also be made such that new call allocations are forced to wait until the number of outstanding connections subsides. Signed-off-by: David Howells <dhowells@redhat.com>
2020-07-01 18:15:32 +08:00
lockdep_is_held(&conn->bundle->channel_lock));
rxrpc: Call channels should have separate call number spaces Each channel on a connection has a separate, independent number space from which to allocate callNumber values. It is entirely possible, for example, to have a connection with four active calls, each with call number 1. Note that the callNumber values for any particular channel don't have to start at 1, but they are supposed to increment monotonically for that channel from a client's perspective and may not be reused once the call number is transmitted (until the epoch cycles all the way back round). Currently, however, call numbers are allocated on a per-connection basis and, further, are held in an rb-tree. The rb-tree is redundant as the four channel pointers in the rxrpc_connection struct are entirely capable of pointing to all the calls currently in progress on a connection. To this end, make the following changes: (1) Handle call number allocation independently per channel. (2) Get rid of the conn->calls rb-tree. This is overkill as a connection may have a maximum of four calls in progress at any one time. Use the pointers in the channels[] array instead, indexed by the channel number from the packet. (3) For each channel, save the result of the last call that was in progress on that channel in conn->channels[] so that the final ACK or ABORT packet can be replayed if necessary. Any call earlier than that is just ignored. If we've seen the next call number in a packet, the last one is most definitely defunct. (4) When generating a RESPONSE packet for a connection, the call number counter for each channel must be included in it. (5) When parsing a RESPONSE packet for a connection, the call number counters contained therein should be used to set the minimum expected call numbers on each channel. To do in future commits: (1) Replay terminal packets based on the last call stored in conn->channels[]. (2) Connections should be retired before the callNumber space on any channel runs out. (3) A server is expected to disregard or reject any new incoming call that has a call number less than the current call number counter. The call number counter for that channel must be advanced to the new call number. Note that the server cannot just require that the next call that it sees on a channel be exactly the call number counter + 1 because then there's a scenario that could cause a problem: The client transmits a packet to initiate a connection, the network goes out, the server sends an ACK (which gets lost), the client sends an ABORT (which also gets lost); the network then reconnects, the client then reuses the call number for the next call (it doesn't know the server already saw the call number), but the server thinks it already has the first packet of this call (it doesn't know that the client doesn't know that it saw the call number the first time). Signed-off-by: David Howells <dhowells@redhat.com>
2016-06-27 21:39:44 +08:00
if (call && call->state < RXRPC_CALL_COMPLETE)
goto protocol_error_unlock;
conn->channels[i].call_counter = call_id;
}
}
rxrpc: Rewrite the client connection manager Rewrite the rxrpc client connection manager so that it can support multiple connections for a given security key to a peer. The following changes are made: (1) For each open socket, the code currently maintains an rbtree with the connections placed into it, keyed by communications parameters. This is tricky to maintain as connections can be culled from the tree or replaced within it. Connections can require replacement for a number of reasons, e.g. their IDs span too great a range for the IDR data type to represent efficiently, the call ID numbers on that conn would overflow or the conn got aborted. This is changed so that there's now a connection bundle object placed in the tree, keyed on the same parameters. The bundle, however, does not need to be replaced. (2) An rxrpc_bundle object can now manage the available channels for a set of parallel connections. The lock that manages this is moved there from the rxrpc_connection struct (channel_lock). (3) There'a a dummy bundle for all incoming connections to share so that they have a channel_lock too. It might be better to give each incoming connection its own bundle. This bundle is not needed to manage which channels incoming calls are made on because that's the solely at whim of the client. (4) The restrictions on how many client connections are around are removed. Instead, a previous patch limits the number of client calls that can be allocated. Ordinarily, client connections are reaped after 2 minutes on the idle queue, but when more than a certain number of connections are in existence, the reaper starts reaping them after 2s of idleness instead to get the numbers back down. It could also be made such that new call allocations are forced to wait until the number of outstanding connections subsides. Signed-off-by: David Howells <dhowells@redhat.com>
2020-07-01 18:15:32 +08:00
spin_unlock(&conn->bundle->channel_lock);
eproto = tracepoint_string("rxkad_rsp_seq");
abort_code = RXKADOUTOFSEQUENCE;
if (ntohl(response->encrypted.inc_nonce) != conn->rxkad.nonce + 1)
goto protocol_error_free;
eproto = tracepoint_string("rxkad_rsp_level");
abort_code = RXKADLEVELFAIL;
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
level = ntohl(response->encrypted.level);
if (level > RXRPC_SECURITY_ENCRYPT)
goto protocol_error_free;
conn->params.security_level = level;
/* create a key to hold the security data and expiration time - after
* this the connection security can be handled in exactly the same way
* as for a client connection */
ret = rxrpc_get_server_data_key(conn, &session_key, expiry, kvno);
if (ret < 0)
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
goto temporary_error_free_ticket;
kfree(ticket);
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
kfree(response);
_leave(" = 0");
return 0;
rxrpc: Call channels should have separate call number spaces Each channel on a connection has a separate, independent number space from which to allocate callNumber values. It is entirely possible, for example, to have a connection with four active calls, each with call number 1. Note that the callNumber values for any particular channel don't have to start at 1, but they are supposed to increment monotonically for that channel from a client's perspective and may not be reused once the call number is transmitted (until the epoch cycles all the way back round). Currently, however, call numbers are allocated on a per-connection basis and, further, are held in an rb-tree. The rb-tree is redundant as the four channel pointers in the rxrpc_connection struct are entirely capable of pointing to all the calls currently in progress on a connection. To this end, make the following changes: (1) Handle call number allocation independently per channel. (2) Get rid of the conn->calls rb-tree. This is overkill as a connection may have a maximum of four calls in progress at any one time. Use the pointers in the channels[] array instead, indexed by the channel number from the packet. (3) For each channel, save the result of the last call that was in progress on that channel in conn->channels[] so that the final ACK or ABORT packet can be replayed if necessary. Any call earlier than that is just ignored. If we've seen the next call number in a packet, the last one is most definitely defunct. (4) When generating a RESPONSE packet for a connection, the call number counter for each channel must be included in it. (5) When parsing a RESPONSE packet for a connection, the call number counters contained therein should be used to set the minimum expected call numbers on each channel. To do in future commits: (1) Replay terminal packets based on the last call stored in conn->channels[]. (2) Connections should be retired before the callNumber space on any channel runs out. (3) A server is expected to disregard or reject any new incoming call that has a call number less than the current call number counter. The call number counter for that channel must be advanced to the new call number. Note that the server cannot just require that the next call that it sees on a channel be exactly the call number counter + 1 because then there's a scenario that could cause a problem: The client transmits a packet to initiate a connection, the network goes out, the server sends an ACK (which gets lost), the client sends an ABORT (which also gets lost); the network then reconnects, the client then reuses the call number for the next call (it doesn't know the server already saw the call number), but the server thinks it already has the first packet of this call (it doesn't know that the client doesn't know that it saw the call number the first time). Signed-off-by: David Howells <dhowells@redhat.com>
2016-06-27 21:39:44 +08:00
protocol_error_unlock:
rxrpc: Rewrite the client connection manager Rewrite the rxrpc client connection manager so that it can support multiple connections for a given security key to a peer. The following changes are made: (1) For each open socket, the code currently maintains an rbtree with the connections placed into it, keyed by communications parameters. This is tricky to maintain as connections can be culled from the tree or replaced within it. Connections can require replacement for a number of reasons, e.g. their IDs span too great a range for the IDR data type to represent efficiently, the call ID numbers on that conn would overflow or the conn got aborted. This is changed so that there's now a connection bundle object placed in the tree, keyed on the same parameters. The bundle, however, does not need to be replaced. (2) An rxrpc_bundle object can now manage the available channels for a set of parallel connections. The lock that manages this is moved there from the rxrpc_connection struct (channel_lock). (3) There'a a dummy bundle for all incoming connections to share so that they have a channel_lock too. It might be better to give each incoming connection its own bundle. This bundle is not needed to manage which channels incoming calls are made on because that's the solely at whim of the client. (4) The restrictions on how many client connections are around are removed. Instead, a previous patch limits the number of client calls that can be allocated. Ordinarily, client connections are reaped after 2 minutes on the idle queue, but when more than a certain number of connections are in existence, the reaper starts reaping them after 2s of idleness instead to get the numbers back down. It could also be made such that new call allocations are forced to wait until the number of outstanding connections subsides. Signed-off-by: David Howells <dhowells@redhat.com>
2020-07-01 18:15:32 +08:00
spin_unlock(&conn->bundle->channel_lock);
protocol_error_free:
kfree(ticket);
protocol_error:
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
kfree(response);
trace_rxrpc_rx_eproto(NULL, sp->hdr.serial, eproto);
key_put(server_key);
*_abort_code = abort_code;
return -EPROTO;
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
temporary_error_free_ticket:
kfree(ticket);
temporary_error_free_resp:
rxrpc: Don't put crypto buffers on the stack Don't put buffers of data to be handed to crypto on the stack as this may cause an assertion failure in the kernel (see below). Fix this by using an kmalloc'd buffer instead. kernel BUG at ./include/linux/scatterlist.h:147! ... RIP: 0010:rxkad_encrypt_response.isra.6+0x191/0x1b0 [rxrpc] RSP: 0018:ffffbe2fc06cfca8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff989277d59900 RCX: 0000000000000028 RDX: 0000259dc06cfd88 RSI: 0000000000000025 RDI: ffffbe30406cfd88 RBP: ffffbe2fc06cfd60 R08: ffffbe2fc06cfd08 R09: ffffbe2fc06cfd08 R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff7c5f80d9f95 R13: ffffbe2fc06cfd88 R14: ffff98927a3f7aa0 R15: ffffbe2fc06cfd08 FS: 0000000000000000(0000) GS:ffff98927fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b1ff28f0f8 CR3: 000000001b412003 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: rxkad_respond_to_challenge+0x297/0x330 [rxrpc] rxrpc_process_connection+0xd1/0x690 [rxrpc] ? process_one_work+0x1c3/0x680 ? __lock_is_held+0x59/0xa0 process_one_work+0x249/0x680 worker_thread+0x3a/0x390 ? process_one_work+0x680/0x680 kthread+0x121/0x140 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x3a/0x50 Reported-by: Jonathan Billings <jsbillings@jsbillings.org> Reported-by: Marc Dionne <marc.dionne@auristor.com> Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Jonathan Billings <jsbillings@jsbillings.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-02-08 23:59:07 +08:00
kfree(response);
temporary_error:
/* Ignore the response packet if we got a temporary error such as
* ENOMEM. We just want to send the challenge again. Note that we
* also come out this way if the ticket decryption fails.
*/
key_put(server_key);
return ret;
}
/*
* clear the connection security
*/
static void rxkad_clear(struct rxrpc_connection *conn)
{
_enter("");
if (conn->rxkad.cipher)
crypto_free_sync_skcipher(conn->rxkad.cipher);
}
/*
* Initialise the rxkad security service.
*/
static int rxkad_init(void)
{
struct crypto_sync_skcipher *tfm;
struct skcipher_request *req;
/* pin the cipher we need so that the crypto layer doesn't invoke
* keventd to go get it */
tfm = crypto_alloc_sync_skcipher("pcbc(fcrypt)", 0, 0);
if (IS_ERR(tfm))
return PTR_ERR(tfm);
req = skcipher_request_alloc(&tfm->base, GFP_KERNEL);
if (!req)
goto nomem_tfm;
rxkad_ci_req = req;
rxkad_ci = tfm;
return 0;
nomem_tfm:
crypto_free_sync_skcipher(tfm);
return -ENOMEM;
}
/*
* Clean up the rxkad security service.
*/
static void rxkad_exit(void)
{
crypto_free_sync_skcipher(rxkad_ci);
skcipher_request_free(rxkad_ci_req);
}
/*
* RxRPC Kerberos-based security
*/
const struct rxrpc_security rxkad = {
.name = "rxkad",
.security_index = RXRPC_SECURITY_RXKAD,
.no_key_abort = RXKADUNKNOWNKEY,
.init = rxkad_init,
.exit = rxkad_exit,
.preparse_server_key = rxkad_preparse_server_key,
.free_preparse_server_key = rxkad_free_preparse_server_key,
.destroy_server_key = rxkad_destroy_server_key,
.init_connection_security = rxkad_init_connection_security,
.how_much_data = rxkad_how_much_data,
.secure_packet = rxkad_secure_packet,
.verify_packet = rxkad_verify_packet,
.free_call_crypto = rxkad_free_call_crypto,
rxrpc: Rewrite the data and ack handling code Rewrite the data and ack handling code such that: (1) Parsing of received ACK and ABORT packets and the distribution and the filing of DATA packets happens entirely within the data_ready context called from the UDP socket. This allows us to process and discard ACK and ABORT packets much more quickly (they're no longer stashed on a queue for a background thread to process). (2) We avoid calling skb_clone(), pskb_pull() and pskb_trim(). We instead keep track of the offset and length of the content of each packet in the sk_buff metadata. This means we don't do any allocation in the receive path. (3) Jumbo DATA packet parsing is now done in data_ready context. Rather than cloning the packet once for each subpacket and pulling/trimming it, we file the packet multiple times with an annotation for each indicating which subpacket is there. From that we can directly calculate the offset and length. (4) A call's receive queue can be accessed without taking locks (memory barriers do have to be used, though). (5) Incoming calls are set up from preallocated resources and immediately made live. They can than have packets queued upon them and ACKs generated. If insufficient resources exist, DATA packet #1 is given a BUSY reply and other DATA packets are discarded). (6) sk_buffs no longer take a ref on their parent call. To make this work, the following changes are made: (1) Each call's receive buffer is now a circular buffer of sk_buff pointers (rxtx_buffer) rather than a number of sk_buff_heads spread between the call and the socket. This permits each sk_buff to be in the buffer multiple times. The receive buffer is reused for the transmit buffer. (2) A circular buffer of annotations (rxtx_annotations) is kept parallel to the data buffer. Transmission phase annotations indicate whether a buffered packet has been ACK'd or not and whether it needs retransmission. Receive phase annotations indicate whether a slot holds a whole packet or a jumbo subpacket and, if the latter, which subpacket. They also note whether the packet has been decrypted in place. (3) DATA packet window tracking is much simplified. Each phase has just two numbers representing the window (rx_hard_ack/rx_top and tx_hard_ack/tx_top). The hard_ack number is the sequence number before base of the window, representing the last packet the other side says it has consumed. hard_ack starts from 0 and the first packet is sequence number 1. The top number is the sequence number of the highest-numbered packet residing in the buffer. Packets between hard_ack+1 and top are soft-ACK'd to indicate they've been received, but not yet consumed. Four macros, before(), before_eq(), after() and after_eq() are added to compare sequence numbers within the window. This allows for the top of the window to wrap when the hard-ack sequence number gets close to the limit. Two flags, RXRPC_CALL_RX_LAST and RXRPC_CALL_TX_LAST, are added also to indicate when rx_top and tx_top point at the packets with the LAST_PACKET bit set, indicating the end of the phase. (4) Calls are queued on the socket 'receive queue' rather than packets. This means that we don't need have to invent dummy packets to queue to indicate abnormal/terminal states and we don't have to keep metadata packets (such as ABORTs) around (5) The offset and length of a (sub)packet's content are now passed to the verify_packet security op. This is currently expected to decrypt the packet in place and validate it. However, there's now nowhere to store the revised offset and length of the actual data within the decrypted blob (there may be a header and padding to skip) because an sk_buff may represent multiple packets, so a locate_data security op is added to retrieve these details from the sk_buff content when needed. (6) recvmsg() now has to handle jumbo subpackets, where each subpacket is individually secured and needs to be individually decrypted. The code to do this is broken out into rxrpc_recvmsg_data() and shared with the kernel API. It now iterates over the call's receive buffer rather than walking the socket receive queue. Additional changes: (1) The timers are condensed to a single timer that is set for the soonest of three timeouts (delayed ACK generation, DATA retransmission and call lifespan). (2) Transmission of ACK and ABORT packets is effected immediately from process-context socket ops/kernel API calls that cause them instead of them being punted off to a background work item. The data_ready handler still has to defer to the background, though. (3) A shutdown op is added to the AF_RXRPC socket so that the AFS filesystem can shut down the socket and flush its own work items before closing the socket to deal with any in-progress service calls. Future additional changes that will need to be considered: (1) Make sure that a call doesn't hog the front of the queue by receiving data from the network as fast as userspace is consuming it to the exclusion of other calls. (2) Transmit delayed ACKs from within recvmsg() when we've consumed sufficiently more packets to avoid the background work item needing to run. Signed-off-by: David Howells <dhowells@redhat.com>
2016-09-08 18:10:12 +08:00
.locate_data = rxkad_locate_data,
.issue_challenge = rxkad_issue_challenge,
.respond_to_challenge = rxkad_respond_to_challenge,
.verify_response = rxkad_verify_response,
.clear = rxkad_clear,
};