From a5f7a5bb3b7f28ba7e4fa246775b29a0e5537255 Mon Sep 17 00:00:00 2001 From: Jinke Han Date: Tue, 8 Sep 2026 15:37:42 +0800 Subject: x86/kprobes: Fix crash when probing CS CALL instructions When using eBPF to probe CS CALL instructions within a function, a crash can be triggered. The eBPF tool probes offset 257 of the __hrtimer_run_queues() function: <__hrtimer_run_queues+249>: nopl 0x0(%rax,%rax,1) <__hrtimer_run_queues+254>: mov %r14,%rdi <__hrtimer_run_queues+257>: cs call <__x86_indirect_thunk_r12> <__hrtimer_run_queues+263>: mov %eax,%r12d <__hrtimer_run_queues+266>: xchg %ax,%ax <__hrtimer_run_queues+268>: mov %r13,%rdi Which triggers this crash: BUG: unable to handle page fault for address: 00000000000f41c9 #PF: supervisor write access in kernel mode #PF: error_code(0x0002) - not-present page PGD 0 P4D 0 Oops: 0002 [#1] SMP NOPTI CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P RIP: 0010:__hrtimer_run_queues+0x106/0x230 Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is at the 6th byte of the above CS CALL instruction. Since the CS CALL instruction occupies 6 bytes, the exception occurred in the middle of that call instruction. The root cause is that when using eBPF tools to probe in the middle of a function, a kprobe with INT3 is used as the underlying implementation. During single-step emulation of the original CALL instruction, int3_emulate_call() assumes that the probed CALL instruction is 5 bytes long. However, the actual CS-prefixed CALL instruction occupies 6 bytes, so it constructs an incorrect exception return address. When the CPU returns from the kprobe handler, the next instruction to be executed is at the address of the last byte of that CS CALL instruction. Coincidentally, starting from that address, the CPU fetches and decodes a completely different instruction, which ultimately triggers a kernel crash. Fix the issue by using the actual instruction length obtained from the instruction decoder when constructing the exception return address, rather than relying on the hardcoded CALL_INSN_SIZE macro. [ mingo: Refined the changelog ] Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step") Suggested-by: Masami Hiramatsu (Google) Signed-off-by: Jinke Han Signed-off-by: Ingo Molnar Reviewed-by: Masami Hiramatsu (Google) Acked-by: Yafang Shao Acked-by: Borislav Petkov Cc: Peter Zijlstra Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000 --- net/sctp/objcnt.c | 105 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 105 insertions(+) create mode 100644 net/sctp/objcnt.c (limited to 'net/sctp/objcnt.c') diff --git a/net/sctp/objcnt.c b/net/sctp/objcnt.c new file mode 100644 index 000000000..0400c964e --- /dev/null +++ b/net/sctp/objcnt.c @@ -0,0 +1,105 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* SCTP kernel implementation + * (C) Copyright IBM Corp. 2001, 2004 + * + * This file is part of the SCTP kernel implementation + * + * Support for memory object debugging. This allows one to monitor the + * object allocations/deallocations for types instrumented for this + * via the proc fs. + * + * Please send any bug reports or fixes you make to the + * email address(es): + * lksctp developers + * + * Written or modified by: + * Jon Grimm + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include +#include + +/* + * Global counters to count raw object allocation counts. + * To add new counters, choose a unique suffix for the variable + * name as the helper macros key off this suffix to make + * life easier for the programmer. + */ + +SCTP_DBG_OBJCNT(sock); +SCTP_DBG_OBJCNT(ep); +SCTP_DBG_OBJCNT(transport); +SCTP_DBG_OBJCNT(assoc); +SCTP_DBG_OBJCNT(bind_addr); +SCTP_DBG_OBJCNT(bind_bucket); +SCTP_DBG_OBJCNT(chunk); +SCTP_DBG_OBJCNT(addr); +SCTP_DBG_OBJCNT(datamsg); +SCTP_DBG_OBJCNT(keys); + +/* An array to make it easy to pretty print the debug information + * to the proc fs. + */ +static struct sctp_dbg_objcnt_entry sctp_dbg_objcnt[] = { + SCTP_DBG_OBJCNT_ENTRY(sock), + SCTP_DBG_OBJCNT_ENTRY(ep), + SCTP_DBG_OBJCNT_ENTRY(assoc), + SCTP_DBG_OBJCNT_ENTRY(transport), + SCTP_DBG_OBJCNT_ENTRY(chunk), + SCTP_DBG_OBJCNT_ENTRY(bind_addr), + SCTP_DBG_OBJCNT_ENTRY(bind_bucket), + SCTP_DBG_OBJCNT_ENTRY(addr), + SCTP_DBG_OBJCNT_ENTRY(datamsg), + SCTP_DBG_OBJCNT_ENTRY(keys), +}; + +/* Callback from procfs to read out objcount information. + * Walk through the entries in the sctp_dbg_objcnt array, dumping + * the raw object counts for each monitored type. + */ +static int sctp_objcnt_seq_show(struct seq_file *seq, void *v) +{ + int i; + + i = (int)*(loff_t *)v; + seq_setwidth(seq, 127); + seq_printf(seq, "%s: %d", sctp_dbg_objcnt[i].label, + atomic_read(sctp_dbg_objcnt[i].counter)); + seq_pad(seq, '\n'); + return 0; +} + +static void *sctp_objcnt_seq_start(struct seq_file *seq, loff_t *pos) +{ + return (*pos >= ARRAY_SIZE(sctp_dbg_objcnt)) ? NULL : (void *)pos; +} + +static void sctp_objcnt_seq_stop(struct seq_file *seq, void *v) +{ +} + +static void *sctp_objcnt_seq_next(struct seq_file *seq, void *v, loff_t *pos) +{ + ++*pos; + return (*pos >= ARRAY_SIZE(sctp_dbg_objcnt)) ? NULL : (void *)pos; +} + +static const struct seq_operations sctp_objcnt_seq_ops = { + .start = sctp_objcnt_seq_start, + .next = sctp_objcnt_seq_next, + .stop = sctp_objcnt_seq_stop, + .show = sctp_objcnt_seq_show, +}; + +/* Initialize the objcount in the proc filesystem. */ +void sctp_dbg_objcnt_init(struct net *net) +{ + struct proc_dir_entry *ent; + + ent = proc_create_seq("sctp_dbg_objcnt", 0, + net->sctp.proc_net_sctp, &sctp_objcnt_seq_ops); + if (!ent) + pr_warn("sctp_dbg_objcnt: Unable to create /proc entry.\n"); +} -- cgit v1.3.1