On Fri, 10 Jun 2016, Jakub Jelinek wrote:
> On Fri, Jun 10, 2016 at 03:13:32PM +0300, Maxim Ostapenko wrote:
> > gcc/ChangeLog:
> >
> > 2016-06-10 Maxim Ostapenko <m.ostape...@samsung.com>
> >
> > PR sanitizer/71480
> > * varasm.c (place_block_symbol): Adjust alignment for asan protected
> > STRING_CSTs even if TREE_CONSTANT_POOL_ADDRESS_P.
> >
> > gcc/testsuite/ChangeLog:
> >
> > 2016-06-10 Maxim Ostapenko <m.ostape...@samsung.com>
> >
> > PR sanitizer/71480
> > * c-c++-common/asan/pr71480.c: New test.
> >
> > diff --git a/gcc/testsuite/c-c++-common/asan/pr71480.c
> > b/gcc/testsuite/c-c++-common/asan/pr71480.c
> > new file mode 100644
> > index 0000000..3cf2c05
> > --- /dev/null
> > +++ b/gcc/testsuite/c-c++-common/asan/pr71480.c
> > @@ -0,0 +1,41 @@
> > +/* { dg-do run } */
> > +
> > +__attribute__ ((noinline, noclone)) int
> > +foo (char *c)
> > +{
>
> I'd add asm volatile ("" : : "r" (c) : "memory"); here
I'd say asm("") is sufficient.
> > + return 1;
> > +}
> > +
> > +__attribute__ ((noinline, noclone)) void
> > +bar (char *c)
> > +{
>
> and here (so that gcc doesn't try to figure out the functions are const or
> pure.
This is exactly what the special-case-empty-asm is for, unless
I miss something extra-special about this case. Maybe to avoid
a warning for the otherwise unused c?
> No need for return; in function returning void.
> Ok with that change, thanks.
>
> Jakub
>
