[Re: ftrace regression with CONFIG_UNWINDER_FRAME_POINTER] On 15/08/2022 (Mon 13:29) Johannes Erdfelt wrote: > On Mon, Aug 15, 2022, Borislav Petkov <bp@xxxxxxxxx> wrote: > > On Mon, Aug 15, 2022 at 04:06:46PM -0400, Steven Rostedt wrote: > > > On Mon, 15 Aug 2022 12:58:28 -0700 > > > Johannes Erdfelt <johannes@xxxxxxxxxxx> wrote: > > > > > > > I recently ran across an oops with ftrace when using the frame pointer > > > > unwinder. It does not affect Linus' tree (tested 6.0-rc1), but it does > > > > affect 5.15.60 and 5.10.136 (at least, have not tested other kernels). > > > > > > > > It appears to be related to the recently merged retbleed mitigations, > > > > > > If you believe this, then Cc those that are involved with those > > > mitigations, which I did. > > > > Sounds like this one: > > > > https://lore.kernel.org/all/Yu2H%2FRdg%2FU4bHWaY@quatroqueijos/ > > Ahh, I missed this in my search to see if this was posted already. > > The patch in that post appears to solve the oops I'm seeing (at least > on top of 5.15.60). Note that if you reload that thread, you'll see what you tested was not a full fix - a series of three commits is pending for 5.15-stable now. Paul. -- > > JE >
