Re: [PATCH 0/7] objtool: retpoline validation

[Peter Zijlstra]

On Thu, Feb 01, 2018 at 03:32:11PM +0000, David Woodhouse wrote:
> On Thu, 2018-02-01 at 09:28 -0600, Josh Poimboeuf wrote:
> > On Thu, Feb 01, 2018 at 03:34:21PM +0100, Peter Zijlstra wrote:
> > > 
> > > There are the retpoline validation patches; they work with the
> > > __noretpoline
> > > thing from David.
> > Have you run this through 0-day bot yet?  A manual awk/sed found
> > another
> > one, which objtool confirms:
> > 
> >   drivers/watchdog/.tmp_hpwdt.o: warning: objtool: .text+0x24:
> > indirect call found in RETPOLINE build
> > 
> > And my search wasn't exhaustive so it would be good to sic 0-day bot on
> > it.
> 
> We discussed that one. It's correct; we're calling into firmware so
> there's *no* point in retpolining that one. We need to set IBRS before
> any runtime calls into firmware, if we want to be safe.

Ideally we'd have a way to mark the module 'unsafe' or something.