On Sun, Apr 25, 2021 at 08:50:18PM -0500, Bill Schmidt wrote:
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/powerpc/rop-1.c
> @@ -0,0 +1,16 @@
> +/* { dg-do compile } */
> +/* { dg-options "-O2 -mdejagnu-cpu=power10 -mrop-protect" } */
This should only run on ELFv2 currently, no?
> +/* { dg-final { scan-assembler {\mhashst\M} } } */
> +/* { dg-final { scan-assembler {\mhashchk\M} } } */
Maybe check these are emitted only once each?
> +/* { dg-final { scan-assembler {\mhashchkp\M} } } */
> diff --git a/gcc/testsuite/gcc.target/powerpc/rop-3.c
> b/gcc/testsuite/gcc.target/powerpc/rop-3.c
> new file mode 100644
> index 00000000000..054f94fda99
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/powerpc/rop-3.c
> @@ -0,0 +1,19 @@
> +/* { dg-do run { target { power10_hw } } } */
> +/* { dg-require-effective-target power10_ok } */
> +/* { dg-require-effective-target powerpc_elfv2 } */
> +/* { dg-options "-O2 -mdejagnu-cpu=power10 -mrop-protect" } */
Why does this need power10_ok? We always are 64-bit if ELFv2 (and even
if you don't want to depend on that, check for lp64 instead?)
> +/* Verify that ROP-protect instructions execute correctly when a
> + call is present. */
> +
> +void __attribute__((noinline)) foo ()
> +{
> + asm ("");
> +}
Use noipa instead? noinline should be fine in this super trivial case,
but it is less typing at least ;-)
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/powerpc/rop-5.c
> @@ -0,0 +1,17 @@
> +/* { dg-do run { target { power10_hw } } } */
> +/* { dg-require-effective-target power10_ok } */
> +/* { dg-require-effective-target powerpc_elfv2 } */
> +/* { dg-options "-O2 -mdejagnu-cpu=power10 -mrop-protect" } */
> +
> +/* Verify that __ROP_PROTECT__ is predefined for -mrop-protect. */
> +
> +extern void abort (void);
> +
> +int main ()
> +{
> +#ifndef __ROP_PROTECT__
> + abort ();
> +#endif
> + return 0;
> +}
Please do this in a compile test instead?
#ifndef __ROP_PROTECT__
har har har
#endif
Okay for trunk and 11 with such changes. Thank you!
Segher
