Messages in this thread | | | Date | Fri, 1 May 2015 16:15:57 +0200 | From | Peter Zijlstra <> | Subject | Re: perf: recursive locking of ctx->mutex |
| |
On Fri, May 01, 2015 at 04:13:37PM +0200, Peter Zijlstra wrote: > > A little something like so should cure that me thinks. > > I would much appreciate other people reviewing this with care though; my > snot addled brain isn't too bright. > > > @@ -8728,7 +8753,7 @@ static int perf_event_init_context(struc > * Lock the parent list. No need to lock the child - not PID > * hashed yet and not running, so nobody can access it. > */ > - mutex_lock(&parent_ctx->mutex); > + mutex_lock_nested(&parent_ctx->mutex); > > /* > * We dont have to disable NMIs - we are only looking at
It will help with building to not apply this last hunk..
| |