Re: [clang] stack protector and f1f029c7bf

[Alistair Strachan]

(Resent plain text)

On Thu, May 24, 2018 at 11:24 AM Nick Desaulniers <ndesaulniers@xxxxxxxxxx>
wrote:

> On Thu, May 24, 2018 at 11:20 AM <hpa@xxxxxxxxx> wrote:
> > A stack canary on an *inlined* function? That's bound to break things
> elsewhere too sooner or later.

> But it's *not* inlined by GCC or Clang.

FWIW, GCC can also insert a stack guard in an out-of-lined inline function,
it just doesn't for this one. The -fstack-protector and
-fstack-protector-strong flags are heuristic and the heuristic does not
match between gcc and clang for this function. It is working on GCC purely
by chance.

> While the function is marked `static inline`, it's not in
> arch/x86/kernel/paravirt.o due to:
> arch/x86/kernel/paravirt.c:326

> 325 __visible struct pv_irq_ops pv_irq_ops = {

> 326         .save_fl = __PV_IS_CALLEE_SAVE(native_save_fl),

> see comparison of disassembly  attached in:
> https://bugs.llvm.org/attachment.cgi?id=20338
> --
> Thanks,
> ~Nick Desaulniers