]> git.ipfire.org Git - thirdparty/gcc.git/commitdiff
[PR testsuite/116860] Testsuite adjustment for recently added tests
authorJeff Law <jlaw@ventanamicro.com>
Thu, 30 Jan 2025 02:42:11 +0000 (19:42 -0700)
committerJeff Law <jlaw@ventanamicro.com>
Thu, 30 Jan 2025 02:45:00 +0000 (19:45 -0700)
There's two new tests that are dependent on logical-op-non-short-circuit
settings.  The BZ is reported against ppc64 and ppc64le, but also applies to a
goodly number of the other targets.

The "regression" fix is trivial, just add the appropriate param to force the
behavior we're expecting.  I'm committing that fix momentarily.  It's been
verified on ppc64, ppc64le and x86_64 as well as the various embedded targets
in my tester where many FAILS flip to PASS.

I'm leaving the bug open without the regression marker as Jakub has noted a
couple of improvements that we can and probably should make.

PR target/116860
gcc/testsuite
* gcc.dg/tree-ssa/fold-xor-and-or.c: Set logical-op-non-short-circuit.
* gcc.dg/tree-ssa/fold-xor-or.c: Similarly.

gcc/testsuite/gcc.dg/tree-ssa/fold-xor-and-or.c
gcc/testsuite/gcc.dg/tree-ssa/fold-xor-or.c

index e5dc98e7541d57637439ae12aba3459d67d5807b..99e83d8e5aae16d584df649b1d569e4e3ecb34ed 100644 (file)
@@ -1,5 +1,5 @@
 /* { dg-do compile } */
-/* { dg-options "-O3 -fdump-tree-optimized" } */
+/* { dg-options "-O3 -fdump-tree-optimized --param logical-op-non-short-circuit=1" } */
 
 typedef unsigned long int uint64_t;
 
index c55cfbcc84c8d47fd3f535ef76dca9816203df18..51b7373af0d869f9f6a8c13e6196f9ca4ec9b3d5 100644 (file)
@@ -1,5 +1,5 @@
 /* { dg-do compile } */
-/* { dg-options "-O3 -fdump-tree-optimized" } */
+/* { dg-options "-O3 -fdump-tree-optimized --param logical-op-non-short-circuit=1" } */
 
 typedef unsigned long int uint64_t;