This testcase has never tested the problem in the PR it is named for
(except perhaps very indirectly), and over the years this argumentation
has become thinner all the time. Now we are faced with on the one hand
having to accept various forms of code, but on the other hand very
similar code is generated elsewhere in the testcase, and the crux of the
check is that we have to make sure no duplicate happens.
I see no better way forward than to delete this testcase.
Comments? I'll commit this tomorrow.
Segher
2022-08-18 Segher Boessenkool <seg...@kernel.crashing.org>
gcc/testsuite/
PR target/102146
* gcc.target/powerpc/pr56605.c: Delete.
---
gcc/testsuite/gcc.target/powerpc/pr56605.c | 15 ---------------
1 file changed, 15 deletions(-)
delete mode 100644 gcc/testsuite/gcc.target/powerpc/pr56605.c
diff --git a/gcc/testsuite/gcc.target/powerpc/pr56605.c
b/gcc/testsuite/gcc.target/powerpc/pr56605.c
deleted file mode 100644
index 7695f87db6f6..000000000000
--- a/gcc/testsuite/gcc.target/powerpc/pr56605.c
+++ /dev/null
@@ -1,15 +0,0 @@
-/* PR rtl-optimization/56605 */
-/* { dg-do compile { target { powerpc*-*-* && lp64 } } } */
-/* { dg-skip-if "" { powerpc*-*-darwin* } } */
-/* { dg-require-effective-target powerpc_vsx_ok } */
-/* { dg-options "-O3 -mvsx -mdejagnu-cpu=power7 -fno-unroll-loops
-fdump-rtl-combine" } */
-
-void foo (short* __restrict sb, int* __restrict ia)
-{
- int i;
- for (i = 0; i < 4000; i++)
- ia[i] = (int) sb[i];
-}
-
-/* { dg-final { scan-rtl-dump-times {\(compare:CC
\((?:and|zero_extend):(?:[SD]I) \((?:sub)?reg:[SD]I} 1 "combine" } } */
-
--
1.8.3.1
