Re: PANIC: double fault in fixup_bad_iret

From: Peter Zijlstra
Date: Tue Jun 02 2020 - 13:59:35 EST

On Tue, Jun 02, 2020 at 07:51:40PM +0200, Marco Elver wrote:

> We have all attributes except __no_sanitize_coverage. GCC <= 7 has
> problems with __always_inline, so we may just have to bump the
> required compiler or emit a warning.

GCC <= 7 will hard fail the compile with those function attributes.
Bumping the min GCC version for KASAN/UBSAN to avoid that might be best.

> > > Not sure what the best strategy is to minimize patch conflicts. For
> > > now I could send just the patches to add missing definitions. If you'd
> > > like me to send all patches (including modifying 'noinstr'), let me
> > > know.
> >
> > If you're going to do patches anyway, might as well do that :-)
> I was stuck on trying to find ways to emulate __no_sanitize_coverage
> (with no success), and then agonizing which patches to send in which
> sequence. ;-) You made that decision by sending the KCSAN noinstr
> series first, so let me respond to that with what I think we can add
> for KASAN and UBSAN at least.

Excellent, thanks!