On Wed, 11 Jan 2017, Palmer Dabbelt wrote:
> diff --git a/gcc/testsuite/gcc.dg/tree-ssa/ssa-dom-cse-2.c
> b/gcc/testsuite/gcc.dg/tree-ssa/ssa-dom-cse-2.c
> index 1a4bfe6..665ac23 100644
> --- a/gcc/testsuite/gcc.dg/tree-ssa/ssa-dom-cse-2.c
> +++ b/gcc/testsuite/gcc.dg/tree-ssa/ssa-dom-cse-2.c
> @@ -25,4 +25,4 @@ foo ()
> but the loop reads only one element at a time, and DOM cannot resolve
> these.
> The same happens on powerpc depending on the SIMD support available. */
>
> -/* { dg-final { scan-tree-dump "return 28;" "optimized" { xfail { {
> alpha*-*-* hppa*64*-*-* powerpc64*-*-* } || { sparc*-*-* && lp64 } } } } } */
> +/* { dg-final { scan-tree-dump "return 28;" "optimized" { xfail { {
> alpha*-*-* hppa*64*-*-* powerpc64*-*-* riscv*64*-*-* } || { sparc*-*-* &&
> lp64 } } } } } */
> diff --git a/gcc/testsuite/gcc.dg/tree-ssa/ssa-fre-3.c
> b/gcc/testsuite/gcc.dg/tree-ssa/ssa-fre-3.c
> index a287dad..0aecfed 100644
> --- a/gcc/testsuite/gcc.dg/tree-ssa/ssa-fre-3.c
> +++ b/gcc/testsuite/gcc.dg/tree-ssa/ssa-fre-3.c
> @@ -5,7 +5,7 @@
>
> When the condition is true, we distribute "(int) (a + b)" as
> "(int) a + (int) b", otherwise we keep the original. */
> -/* { dg-do compile { target { { ! mips64 } && { ! spu-*-* } } } } */
> +/* { dg-do compile { target { { ! mips64 } && { { ! spu-*-* } && { !
> riscv*64*-*-* } } } } } */
> /* { dg-options "-O -fno-tree-forwprop -fno-tree-ccp -fwrapv
> -fdump-tree-fre1-details" } */
>
> /* From PR14844. */
As I understand it, you have -m options that can change the ABI between
32-bit and 64-bit. In such a case you mustn't check for riscv*64*-*-*
target triplets in the testsuite, since that behave incorrectly when
testing with an option that changes to the other one of 32-bit and 64-bit
from the configured default. Instead, you need to allow for all
riscv*-*-* triplets and combine as needed with a test such as lp64 or
ilp32.
--
Joseph S. Myers
jos...@codesourcery.com
