Re: [RFC PATCH 00/21] x86: Proposed fixes for stackvalidate warnings

[Ingo Molnar <mingo@kernel.org>]

* Andy Lutomirski <luto@amacapital.net> wrote:

> On Fri, Jul 17, 2015 at 9:47 AM, Josh Poimboeuf <jpoimboe@redhat.com> wrote:
> > These patches fix many of the warnings reported by stackvalidate.
> > They're based on top of the "Compile-time stack validation" v7 patch set
> > [1].
> >
> > They've been compile-tested and boot tested in a VM, but I haven't
> > attempted any meaningful testing for most of them.  This should give an
> > idea of what kinds of changes I think are needed.
> >
> > [1] https://lkml.kernel.org/r/cover.1436893563.git.jpoimboe@redhat.com
> >
> 
> Nothing here looks all that bad, but I think the extra frame pointers
> staring us in the face (as opposed to the ones that gcc adds
> transparently) might serve as added incentive to suck it up and get
> CFI unwinding working.

So there are two aspects to frame pointers staring us in the face:

 - Syntactically there's extra debuginfo cruft added to the source code.
   Not good but very fixable: I already made a couple of suggestions of
   how to trim all that in a way that improves general readability as well over 
   what we have today.

 - Instruction wise in the generated code. I'm afraid the 'transparent' frame 
   pointers added by GCC are staring me in the face in perf top/report disassembly 
   output just as much! ;-)

But naming details aside, I like the direction of these patches a lot better than 
that of any previous debuginfo approach, because the approach is very proactive.

( Which it really has to be, given that deep down these patches are motivated by 
  enabling more complex models of live kernel patching. But that's a win-win. )

Once the cleanliness of the annotations is improved and we have a reasonable path 
towards having no warnings on a regular kernel build, I plan on starting to apply 
the patches.

So if anyone has any deep objections that were kept close to the vest so far, 
please holler now.

Thanks,

	Ingo