On Fri, Feb 04, 2022 at 11:12:49AM +0100, Pavel Machek wrote:
> On Fri 2022-02-04 10:40:45, Greg Kroah-Hartman wrote:
> > On Fri, Feb 04, 2022 at 10:37:35AM +0100, Pavel Machek wrote:
> > > Hi!
> > >
> > > > From: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> > > >
> > > > commit ef54c1a476aef7eef26fe13ea10dc090952c00f8 upstream.
> > > >
> > > > Make perf_event_exit_event() more robust, such that we can use it from
> > > > other contexts. Specifically the up and coming remove_on_exec.
> > >
> > > Do we need this in 5.10? AFAICT the remove_on_exec work is not queued
> > > for 5.10, and this patch is buggy and needs following one to fix it
> > > up.
> >
> > It's needed by the following patch, which says 5.10 is affected.
>
> 9/25 says this patch broke 5.10: Fixes: ef54c1a476ae ("perf: Rework
> perf_event_exit_event()"). 8/25 is not claiming to fix anything.
>
> Simply drop 8/25 and 9/25, and 5.10 is okay...
Ah, yeah, the patch this was to fix was added and then reverted which
confused things. I'll go drop both of these now, thanks.
greg k-h
