[PATCH v2] tracing: perf: Fix stale head for perf syscall tracing

Steven Rostedt posted 1 patch 5 hours ago
kernel/trace/trace_syscalls.c | 5 +++++
1 file changed, 5 insertions(+)
[PATCH v2] tracing: perf: Fix stale head for perf syscall tracing
Posted by Steven Rostedt 5 hours ago
From: Steven Rostedt <rostedt@goodmis.org>

The code that can read the user space parameters of a system call may
enable preemption and migrate. The head of the per CPU perf events list
may be pointing to the wrong CPU event if the code migrates the task.

Reassign the head pointer if the system call event called the code that
may have caused a migration.

Cc: stable@vger.kernel.org
Reported-by: Sashiko <>
Link: https://sashiko.dev/#/patchset/20260717173252.3431565-1-usama.arif%40linux.dev
Fixes: edca33a56297d ("tracing: Fix failure to read user space from system call trace events")
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
Changes since v1: https://lore.kernel.org/all/20260724184250.357e91f1@gandalf.local.home/

- Keep the shortcut out if the current CPU is not tracing the system call.
  Instead, reassign the head pointer only if the syscall_get_data() is called.

 kernel/trace/trace_syscalls.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c
index e98ee7e1e66f..8a4f3c75e39f 100644
--- a/kernel/trace/trace_syscalls.c
+++ b/kernel/trace/trace_syscalls.c
@@ -1451,6 +1451,11 @@ static void perf_syscall_enter(void *ignore, struct pt_regs *regs, long id)
 		if (syscall_get_data(sys_data, args, &user_ptr,
 				     &size, user_sizes, &uargs, buf_size) < 0)
 			return;
+
+		/* The above may have caused a migration */
+		head = this_cpu_ptr(sys_data->enter_event->perf_events);
+		if (hlist_empty(head))
+			return;
 	}
 
 	/* get the size after alignment with the u32 buffer size field */
-- 
2.53.0