On Thu, Sep 03, 2026 at 10:31:16AM -0400, Steven Rostedt wrote:
> On Tue,  1 Sep 2026 16:54:41 +0100
> Vincent Donnefort <[email protected]> wrote:
> 
> > This series addresses multiple issues discovered with the dynamic ring
> > buffer resizing.
> > 
> > I have managed to reproduce a ring_buffer_read_page() race with
> > 
> >  $ while true; do for i in 8 16 32; do echo $i > 
> > /sys/kernel/tracing/buffer_subbuf_size_kb; sleep 0.1; done; done
> 
> Did you do anything else while running the above?
> 
> I ask because I ran this for several hours and it didn't produce anything.
> What bug does it supposed to trigger?
> 
> -- Steve

I clearly have been imprecise here.

I had "trace-record -e sched" and I have just verified with a print that it has
run into the reader->order != read_page->order case.

-- 
Vincent

Reply via email to