Hi!
On Wed, Jul 08, 2020 at 11:02:45PM -0500, Peter Bergner wrote:
> PR96125 shows a bug when we try to use an MMA built-in within a function
> that uses #pragma target/attribute target to enable power10 code generation
> and the -mcpu=<CPU> command line option is pre-power10.
>
> The problem is that we only initialize built-ins once, fairly early, when
> the command line options are in force. If the -mcpu=<CPU> is pre-power10,
> then we fail to initialize the MMA built-ins at all and so they are not
> available to call in a #pragma target/attribute target function.
>
> The patch below basically always (on server type cpus) initializes the MMA
> built-ins so we can use them in #pragma target/attribute target functions.
> The patch below fixes the bug and is currently in the middle of testing.
>
> Is this ok for trunk assuming the bootstrap and regression testing
> show no regressions?
>
> This also affects GCC10, so I'd like to backport this before the release.
> Ok there too after it sits on trunk a day or two?
> gcc/
> PR target/96125
> * config/rs6000/rs6000-call.c (rs6000_init_builtins): Define the MMA
> specific types __vector_quad and __vector_pair, and initialize the
> MMA built-ins if TARGET_EXTRA_BUILTINS is set.
> (mma_init_builtins): Don't test for mask set in rs6000_builtin_mask.
> Remove now unneeded mask variable.
> * config/rs6000/rs6000.c (rs6000_option_override_internal): Add the
> OPTION_MASK_MMA flag for power10 if not already set.
>
> gcc/testsuite/
> PR target/96125
> * gcc.target/powerpc/pr96125.c: New test.
Okay for trunk and 10 (but see test nit below). Thanks!
> - /* Create Altivec and VSX builtins on machines with at least the
> + /* Create Altivec, VSX and MMA builtins on machines with at least the
> general purpose extensions (970 and newer) to allow the use of
> the target attribute. */
> if (TARGET_EXTRA_BUILTINS)
> - altivec_init_builtins ();
> - if (TARGET_MMA)
> - mma_init_builtins ();
> + {
> + altivec_init_builtins ();
> + mma_init_builtins ();
> + }
> if (TARGET_HTM)
> htm_init_builtins ();
So maybe we should just do all builtins always?
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/powerpc/pr96125.c
> @@ -0,0 +1,47 @@
> +/* { dg-do compile } */
> +/* { dg-require-effective-target powerpc_vsx_ok } */
> +/* { dg-options "-mdejagnu-cpu=power8 -O2" } */
powerpc_vsx_ok is not the right test for -mcpu=power8 (it means p7).
Usually powerpc_p8vector_ok is used... not a great situation, but :-)
Segher
