KCSAN: data-race in perf_pending_task / perf_swevent_get_recursion_context

Status: moderation: reported on 2026/11/18 05:04
Subsystems: perf
Labels: prio:low race:benign
[Documentation on labels]
Reported-by: syzbot+286ba4942f371412d8e3@syzkaller.appspotmail.com
First crash: 1d02h, last: 10h41m
✨ AI cause bisection: introduced by (AI job) :
commit 16b9569df9d2ab07eeee075cb7895e9d3e08e8f0
Author: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Date: Thu Jul 4 17:03:40 2024 +0000

  perf: Don't disable preemption in perf_pending_task().

  
✨ AI-generated root cause summary:

perf_swevent_get_recursion_context() indexes current->perf_recursion[] using interrupt_context_level() (level 0 for task context), relying on non-preemptibility to prevent concurrent re-entrant access on the same task. However, perf_pending_task() executes as a task_work callback with preemption and interrupts enabled. When an interrupt interrupts perf_pending_task() and triggers preemption on exit (via preempt_schedule_irq()), __schedule() invokes the sched_switch tracepoint on the same preempted current task, leading to concurrent unprotected access to current->perf_recursion[0].

▶ ▼ ✨ Full AI root cause analysis
Based on kernel commit 666539eecdf4:

Root Cause Analysis

1. Mechanism and Invariant Violation

The functions perf_swevent_get_recursion_context() and perf_swevent_put_recursion_context() (kernel/events/core.c:11233-11242) track software event recursion by indexing into a per-task context array, current->perf_recursion[PERF_NR_CONTEXTS].

The helper get_recursion_context(u8 *recursion) in kernel/events/internal.h:214-225:

static inline int get_recursion_context(u8 *recursion)
{
	unsigned char rctx = interrupt_context_level();

	if (recursion[rctx])
		return -1;

	recursion[rctx]++;
	barrier();

	return rctx;
}

interrupt_context_level() (include/linux/preempt.h:118-128) computes the execution level based on preempt_count():

  • 0 - task (normal context)
  • 1 - softirq context
  • 2 - hardirq context
  • 3 - NMI context

get_recursion_context() and put_recursion_context() use plain, non-atomic reads and increments/decrements with only compiler barriers. This mechanism relies on an architectural invariant:

  • Interrupts (softirq, hardirq, NMI) operate at rctx >= 1, accessing disjoint entries of the array without colliding with recursion[0].
  • For rctx == 0 (task context), execution must be non-preemptible to guarantee that multiple task-context callers on the same task cannot execute re-entrantly or concurrently.

Standard callers adhere to this requirement:

  • __perf_sw_event() runs enclosed within preempt_disable_notrace() and preempt_enable_notrace().
  • perf_trace_##call() macros run under guard(preempt_notrace)().
  • perf_pending_irq() and perf_pending_disable() execute from irq_work_run() in interrupt or softirq contexts (rctx >= 1).

2. Triggering Condition in perf_pending_task()

In contrast to the other callers, perf_pending_task() (kernel/events/core.c:7767-7787) is queued via task_work_add() and executed from task_work_run() in normal, sleepable, preemptible task context (rctx == 0). Both IRQs and preemption are enabled throughout the function.

  1. Task T enters perf_pending_task() and invokes perf_swevent_get_recursion_context().
  2. Inside get_recursion_context(), interrupt_context_level() returns 0, and T reads current->perf_recursion[0].
  3. An asynchronous interrupt (e.g., sysvec_apic_timer_interrupt) interrupts task T.
  4. The interrupt executes. On return to kernel mode (irqentry_exit() -> irqentry_exit_to_kernel_mode_preempt() -> irqentry_exit_cond_resched()), the hardirq state has already ended (!preempt_count()).
  5. Because rescheduling is needed (need_resched()), preempt_schedule_irq() calls __schedule(SM_PREEMPT).
  6. Inside preempt_schedule_irq(), preempt_disable() sets only PREEMPT_OFFSET. Therefore, interrupt_context_level() evaluates to 0. Prior to switching out the running process, current is still task T.
  7. __schedule() fires the trace_sched_switch tracepoint. The attached perf handler (perf_trace_sched_switch() -> perf_trace_buf_alloc()) invokes perf_swevent_get_recursion_context().
  8. get_recursion_context() reads and increments current->perf_recursion[0], concurrently racing with the preempted access in perf_pending_task() on task T.

Key Functions, Structs, and Files Involved

▶ ▼ ✨ AI Jobs (3)
ID Workflow Result Correct Bug Created Started Finished Revision
88c1f717 bisect-cause ❓ KCSAN: data-race in perf_pending_task / perf_swevent_get_recursion_context 2026/11/18 05:13 2026/11/18 05:13 2026/11/18 05:44 ca622473
b0a4bba3 assessment-security DenialOfService: ❌ Exploitable: ❌ FilesystemTrigger: ❌ FuzzingArtifact: ❌ MemoryCorruption: ❌ NetworkNeedsSetup: ❌ NetworkTrigger: ❌ PeripheralTrigger: ❌ RemoteTrigger: ❌ Unprivileged: ✅ UserNamespace: ✅ VMGuestTrigger: ❌ VMHostTrigger: ❌ ❓ KCSAN: data-race in perf_pending_task / perf_swevent_get_recursion_context 2026/11/18 04:48 2026/11/18 04:48 2026/11/18 05:03 ca622473
951acab7 assessment-kcsan Benign: ✅ ❓ KCSAN: data-race in perf_pending_task / perf_swevent_get_recursion_context 2026/11/18 04:37 2026/11/18 04:37 2026/11/18 05:55 ca622473

Sample crash report:
==================================================================
BUG: KCSAN: data-race in perf_pending_task / perf_swevent_get_recursion_context

write to 0xffff8881047cacb0 of 1 bytes by task 7070 on cpu 0:
 get_recursion_context kernel/events/internal.h:221 [inline]
 perf_swevent_get_recursion_context+0xa7/0xc0 kernel/events/core.c:11235
 perf_trace_buf_alloc+0x36/0x100 kernel/trace/trace_event_perf.c:409
 do_perf_trace_sched_switch include/trace/events/sched.h:220 [inline]
 perf_trace_sched_switch+0x77/0x1d0 include/trace/events/sched.h:220
 __traceiter_sched_switch+0x3f/0x60 include/trace/events/sched.h:220
 __do_trace_sched_switch include/trace/events/sched.h:220 [inline]
 trace_sched_switch include/trace/events/sched.h:220 [inline]
 __schedule+0xc3d/0xd30 kernel/sched/core.c:7383
 preempt_schedule_irq+0x2c/0x50 kernel/sched/core.c:7666
 irqentry_exit_cond_resched kernel/entry/common.c:145 [inline]
 irqentry_exit_to_kernel_mode_preempt+0x63/0x80 include/linux/irq-entry-common.h:453
 irqentry_exit_to_kernel_mode include/linux/irq-entry-common.h:543 [inline]
 irqentry_exit+0x99/0x640 kernel/entry/common.c:154
 sysvec_apic_timer_interrupt+0x44/0x80 arch/x86/kernel/apic/apic.c:1062
 asm_sysvec_apic_timer_interrupt+0x1a/0x20 arch/x86/include/asm/idtentry.h:674
 native_pause arch/x86/include/asm/vdso/processor.h:13 [inline]
 delay_tsc+0x43/0xa0 arch/x86/lib/delay.c:76
 udelay include/asm-generic/delay.h:64 [inline]
 delay_access kernel/kcsan/core.c:332 [inline]
 kcsan_setup_watchpoint+0x255/0x3f0 kernel/kcsan/core.c:622
 get_recursion_context kernel/events/internal.h:218 [inline]
 perf_swevent_get_recursion_context kernel/events/core.c:11235 [inline]
 perf_pending_task+0x80/0x220 kernel/events/core.c:7776
 task_work_run+0x145/0x1a0 kernel/task_work.c:233
 resume_user_mode_work+0x85/0xa0 include/linux/resume_user_mode.h:50
 __exit_to_user_mode_loop kernel/entry/common.c:70 [inline]
 exit_to_user_mode_loop kernel/entry/common.c:101 [inline]
 __exit_to_user_mode_prepare include/linux/irq-entry-common.h:207 [inline]
 irqentry_exit_to_user_mode_prepare include/linux/irq-entry-common.h:244 [inline]
 irqentry_exit_to_user_mode include/linux/irq-entry-common.h:315 [inline]
 irqentry_exit+0x170/0x640 kernel/entry/common.c:152
 sysvec_apic_timer_interrupt+0x44/0x80 arch/x86/kernel/apic/apic.c:1062
 asm_sysvec_apic_timer_interrupt+0x1a/0x20 arch/x86/include/asm/idtentry.h:674

read to 0xffff8881047cacb0 of 1 bytes by task 7070 on cpu 0:
 get_recursion_context kernel/events/internal.h:218 [inline]
 perf_swevent_get_recursion_context kernel/events/core.c:11235 [inline]
 perf_pending_task+0x80/0x220 kernel/events/core.c:7776
 task_work_run+0x145/0x1a0 kernel/task_work.c:233
 resume_user_mode_work+0x85/0xa0 include/linux/resume_user_mode.h:50
 __exit_to_user_mode_loop kernel/entry/common.c:70 [inline]
 exit_to_user_mode_loop kernel/entry/common.c:101 [inline]
 __exit_to_user_mode_prepare include/linux/irq-entry-common.h:207 [inline]
 irqentry_exit_to_user_mode_prepare include/linux/irq-entry-common.h:244 [inline]
 irqentry_exit_to_user_mode include/linux/irq-entry-common.h:315 [inline]
 irqentry_exit+0x170/0x640 kernel/entry/common.c:152
 sysvec_apic_timer_interrupt+0x44/0x80 arch/x86/kernel/apic/apic.c:1062
 asm_sysvec_apic_timer_interrupt+0x1a/0x20 arch/x86/include/asm/idtentry.h:674

Reported by Kernel Concurrency Sanitizer on:
CPU: 0 UID: 0 PID: 7070 Comm: syz.2.361 Not tainted syzkaller #0 PREEMPT(lazy) 
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 11/06/2026
==================================================================

Crashes (3):
Time Kernel Commit Syzkaller Config Log Report Syz repro C repro VM info Assets (help?) Manager Title
2026/11/17 19:30 upstream 666539eecdf4 122f94d5 .config console log report info [disk image] [vmlinux] [kernel image] ci2-upstream-kcsan-gce KCSAN: data-race in perf_pending_task / perf_swevent_get_recursion_context
2026/11/17 11:53 upstream 666539eecdf4 43af6cac .config console log report info [disk image] [vmlinux] [kernel image] ci2-upstream-kcsan-gce KCSAN: data-race in perf_pending_task / perf_swevent_get_recursion_context
2026/11/17 03:18 upstream 666539eecdf4 a3e64cfc .config console log report info [disk image] [vmlinux] [kernel image] ci2-upstream-kcsan-gce KCSAN: data-race in perf_pending_task / perf_swevent_get_recursion_context
* Struck through repros no longer work on HEAD.