linux/net/smc
Qi Tang ae5575e660 net/smc: fix double-free of smc_spd_priv when tee() duplicates splice pipe buffer
[ Upstream commit 24dd586bb4 ]

smc_rx_splice() allocates one smc_spd_priv per pipe_buffer and stores
the pointer in pipe_buffer.private.  The pipe_buf_operations for these
buffers used .get = generic_pipe_buf_get, which only increments the page
reference count when tee(2) duplicates a pipe buffer.  The smc_spd_priv
pointer itself was not handled, so after tee() both the original and the
cloned pipe_buffer share the same smc_spd_priv *.

When both pipes are subsequently released, smc_rx_pipe_buf_release() is
called twice against the same object:

  1st call: kfree(priv)  sock_put(sk)  smc_rx_update_cons()  [correct]
  2nd call: kfree(priv)  sock_put(sk)  smc_rx_update_cons()  [UAF]

KASAN reports a slab-use-after-free in smc_rx_pipe_buf_release(), which
then escalates to a NULL-pointer dereference and kernel panic via
smc_rx_update_consumer() when it chases the freed priv->smc pointer:

  BUG: KASAN: slab-use-after-free in smc_rx_pipe_buf_release+0x78/0x2a0
  Read of size 8 at addr ffff888004a45740 by task smc_splice_tee_/74
  Call Trace:
   <TASK>
   dump_stack_lvl+0x53/0x70
   print_report+0xce/0x650
   kasan_report+0xc6/0x100
   smc_rx_pipe_buf_release+0x78/0x2a0
   free_pipe_info+0xd4/0x130
   pipe_release+0x142/0x160
   __fput+0x1c6/0x490
   __x64_sys_close+0x4f/0x90
   do_syscall_64+0xa6/0x1a0
   entry_SYSCALL_64_after_hwframe+0x77/0x7f
   </TASK>

  BUG: kernel NULL pointer dereference, address: 0000000000000020
  RIP: 0010:smc_rx_update_consumer+0x8d/0x350
  Call Trace:
   <TASK>
   smc_rx_pipe_buf_release+0x121/0x2a0
   free_pipe_info+0xd4/0x130
   pipe_release+0x142/0x160
   __fput+0x1c6/0x490
   __x64_sys_close+0x4f/0x90
   do_syscall_64+0xa6/0x1a0
   entry_SYSCALL_64_after_hwframe+0x77/0x7f
   </TASK>
  Kernel panic - not syncing: Fatal exception

Beyond the memory-safety problem, duplicating an SMC splice buffer is
semantically questionable: smc_rx_update_cons() would advance the
consumer cursor twice for the same data, corrupting receive-window
accounting.  A refcount on smc_spd_priv could fix the double-free, but
the cursor-accounting issue would still need to be addressed separately.

The .get callback is invoked by both tee(2) and splice_pipe_to_pipe()
for partial transfers; both will now return -EFAULT.  Users who need
to duplicate SMC socket data must use a copy-based read path.

Fixes: 9014db202c ("smc: add support for splice()")
Signed-off-by: Qi Tang <tpluszz77@gmail.com>
Link: https://patch.msgid.link/20260318064847.23341-1-tpluszz77@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
History 2026-04-18 10:33:24 +02:00
..
af_smc.c net/smc: fix NULL dereference and UAF in smc_tcp_syn_recv_sock() 2026-04-18 10:33:21 +02:00
Kconfig treewide: replace '---help---' in Kconfig files with 'help' 2020-06-14 01:57:21 +09:00
Makefile net/smc: Add SMC statistics support 2021-06-16 12:54:02 -07:00
smc.h net/smc: fix NULL dereference and UAF in smc_tcp_syn_recv_sock() 2026-04-18 10:33:21 +02:00
smc_cdc.c net/smc: allow cdc msg send rather than drop it with NULL sndbuf_desc 2023-11-20 11:08:29 +01:00
smc_cdc.h net/smc: fix kernel panic caused by race of smc_sock 2022-01-05 12:42:36 +01:00
smc_clc.c net/smc: fix mismatch between CLC header and proposal 2025-12-07 06:09:27 +09:00
smc_clc.h net/smc: check smcd_v2_ext_offset when receiving proposal msg 2025-01-09 13:28:31 +01:00
smc_close.c net/smc: fix NULL dereference and UAF in smc_tcp_syn_recv_sock() 2026-04-18 10:33:21 +02:00
smc_close.h net/smc: remove close abort worker 2019-10-22 11:23:44 -07:00
smc_core.c net/smc: set rmb's SG_MAX_SINGLE_ALLOC limitation only when CONFIG_ARCH_NO_SG_CHAIN is defined 2024-08-19 05:44:56 +02:00
smc_core.h net/smc: Reset conn->lgr when link group registration fails 2022-01-27 11:03:53 +01:00
smc_diag.c net/smc: fix illegal rmb_desc access in SMC-D connection dump 2024-02-23 08:54:27 +01:00
smc_ib.c net/smc: fix kernel panic caused by race of smc_sock 2022-01-05 12:42:36 +01:00
smc_ib.h net/smc: fix kernel panic caused by race of smc_sock 2022-01-05 12:42:36 +01:00
smc_ism.c net/smc: no need to flush smcd_dev's event_wq before destroying it 2021-06-03 13:54:49 -07:00
smc_ism.h net/smc: Add support for obtaining SMCD device list 2020-12-01 17:56:13 -08:00
smc_llc.c net/smc: Avoid to access invalid RMBs' MRs in SMCRv1 ADD LINK CONT 2023-06-14 11:13:01 +02:00
smc_llc.h net/smc: move add link processing for new device into llc layer 2020-07-19 15:30:22 -07:00
smc_netlink.c net/smc: Add netlink support for SMC fallback statistics 2021-06-16 12:54:02 -07:00
smc_netlink.h net/smc: Add netlink support for SMC fallback statistics 2021-06-16 12:54:02 -07:00
smc_netns.h net/smc: introduce list of pnetids for Ethernet devices 2020-09-28 15:19:03 -07:00
smc_pnet.c net/smc: Fix searching in list of known pnetids in smc_pnet_add_pnetid 2024-11-01 01:52:34 +01:00
smc_pnet.h net/smc: Use a mutex for locking "struct smc_pnettable" 2022-03-02 11:47:59 +01:00
smc_rx.c net/smc: fix double-free of smc_spd_priv when tee() duplicates splice pipe buffer 2026-04-18 10:33:24 +02:00
smc_rx.h net/smc: fix data error when recvmsg with MSG_PEEK flag 2025-03-13 12:49:57 +01:00
smc_stats.c net/smc: Fix ENODATA tests in smc_nl_get_fback_stats() 2021-06-21 12:16:58 -07:00
smc_stats.h net/smc: Fix pos miscalculation in statistics 2023-10-19 23:05:34 +02:00
smc_tx.c net: deal with most data-races in sk_wait_event() 2023-05-24 17:36:42 +01:00
smc_tx.h net/smc: Send directly when TCP_CORK is cleared 2022-04-13 20:59:02 +02:00
smc_wr.c net/smc: fix kernel panic caused by race of smc_sock 2022-01-05 12:42:36 +01:00
smc_wr.h net/smc: fix kernel panic caused by race of smc_sock 2022-01-05 12:42:36 +01:00