CVE-2026-99032

In the Linux kernel, the following vulnerability has been resolved:

net/smc: protect clcsock lifetime in smc getname

smc getname() dereferences smc->clcsock without holding clcsock release lock. Link-group termination can release the CLC socket through smc close active abort() while the SMC socket is still open, for example after shutdown(SHUT WR).

A getsockname() caller can load smc->clcsock, then the termination worker can clear the pointer and call sock release() before the caller accesses clcsock->ops or invokes getname(). This causes a use-after-free; if the worker clears the pointer before the load, it causes a NULL dereference. The syscall's file reference keeps the SMC socket alive but does not prevent asynchronous release of its CLC socket.

KASAN reported the following with test-only timing instrumentation:

BUG: KASAN: slab-use-after-free in smc getname+0x19e/0x1b0 Read of size 8 at addr ffff888109abb4e0 by task poc/103 Call Trace: smc getname+0x19e/0x1b0 do getsockname+0xe5/0x170 sys getsockname+0x8c/0x100

Allocated by task 95: sock alloc inode+0x1e/0x280 sock alloc+0x3d/0x240 sock create+0x7e/0x430 smc create+0x121/0x240

Freed by task 0: kmem cache free+0xcc/0x340 rcu core+0x50a/0x1850

Last potentially related work creation: evict+0x446/0x6c0 smc clcsock release+0xa8/0xd0 smc close active abort+0x26a/0x3a0 smc lgr terminate.part.0+0x137/0x2e0

Hold clcsock release lock across the pointer check and the getname() callback to serialize with smc clcsock release(). Return -EBADF if the CLC socket has already been released, preserving the existing peer state check and the callback's return value otherwise.