Skip to contentSkip to conversation

trace: Fix race in trace_open and buffer resize call

PATCH In review Gaurav Kohli 3 participants 3 messages Jan 24, 2021

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 trace: Fix race in trace_open and buffer resize call

Below race can come, if trace_open and resize of
cpu buffer is running parallely on different cpus
CPUX CPUY

				    ring_buffer_resize
				    atomic_read(&buffer->resize_disabled)

tracing_open
tracing_reset_online_cpus ring_buffer_reset_cpu
rb_reset_cpu
rb_update_pages remove/insert pages
resetting pointer

This race can cause data abort or some times infinite loop in rb_remove_pages and rb_insert_pages while checking pages for sanity.

Take buffer lock to fix this.

kernel/trace/ring_buffer.c +4 -0
index 077877e..7283741 100644
@@ -4448,6 +4448,8 @@ void ring_buffer_reset_cpu(struct ring_buffer *buffer, int cpu)
 
 	if (!cpumask_test_cpu(cpu, buffer->cpumask))
 		return;
+	/* prevent another thread from changing buffer sizes */
+	mutex_lock(&buffer->mutex);
 
 	atomic_inc(&buffer->resize_disabled);
 	atomic_inc(&cpu_buffer->record_disabled);
@@ -4471,6 +4473,8 @@ void ring_buffer_reset_cpu(struct ring_buffer *buffer, int cpu)
 
 	atomic_dec(&cpu_buffer->record_disabled);
 	atomic_dec(&buffer->resize_disabled);
+
+	mutex_unlock(&buffer->mutex);
 }
 EXPORT_SYMBOL_GPL(ring_buffer_reset_cpu);
 
Signature · 2 lines
Qualcomm India Private Limited, on behalf of Qualcomm Innovation Center,
Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project
#1 lore raw
Quoted 1 line · Gaurav Kohli

Below race can come, if trace_open and resize of

<snip>

You forgot to cc: stable, and put the git id that this patch is upstream...

maintainer
#2 lore raw
Quoted 7 lines · On Sun, 24 Jan 2021 11:05:28 +0100 Greg KH <[email protected]> wrote:
Quoted 1 line · Gaurav Kohli

Below race can come, if trace_open and resize of

<snip>

You forgot to cc: stable, and put the git id that this patch is upstream...

Hi Greg,

Sorry for the confusion. I wanted to give it a test before pushing it officially to stable. Gaurav was just following my instructions. After I get a chance to test it (tomorrow), I'll send it properly to stable.

Thanks,

-- Steve