Re: v5.15.57 regression - boot panic after retbleed backports with CONFIG_KPROBES_SANITY_TEST=y

[Paul Gortmaker <paul.gortmaker@windriver.com>]

[Re: v5.15.57 regression - boot panic after retbleed backports with CONFIG_KPROBES_SANITY_TEST=y] On 16/08/2022 (Tue 04:29) Thadeu Lima de Souza Cascardo wrote:

> On Tue, Aug 16, 2022 at 12:12:24AM -0400, Paul Gortmaker wrote:

> > Cascardo,
> > 
> > Was this a final patch or just a "can you test this so we can better
> > understand the regression"  intermediate step?  I don't think anyone
> > wants to guess at that or at what you wanted to put in the commit log.
> > 
> > It would be nice to close this out vs. leaving it hanging out there.
> > 
> > Thanks,
> > Paul.
> > --
> 
> Not a final thing, I was catching up on other stuff before I could go back to
> it, was planning on doing it this week. Just need to sort out the order of some
> commits here to get to that end result.

Great - I didn't want to nag, but some new ftrace/kprobe regressions
were reported on a Yocto variant of v5.15 and I immediately thought of
this being the underlying cause and the breakage being more than just
the CONFIG_KPROBES_SANITY_TEST corner case I originally reported.

The good news is that I've tested with your new three pack and that
fixes both new issues.  I'll loop back and re-check the original issue
as well for completeness, but if you don't hear from me again, you can
assume that went well too.  So, if Greg wants to, feel free to add: 

Tested-by: Paul Gortmaker <paul.gortmaker@windriver.com>

Thanks again for fixing this.  Much appreciated.
Paul.
--

> 
> Cascardo.
> 
> > 
> > > 
> > > thanks,
> > > 
> > > greg k-h