Re: Re: [PATCH][AArch64] Testcase fix for __ATOMIC_CONSUME

[Alex Velenko <Alex dot Velenko at arm dot com>]

On 28/01/15 18:50, Mike Stump wrote:
> On Jan 28, 2015, at 9:51 AM, Marcus Shawcroft <marcus.shawcroft@gmail.com> wrote:
> > Going forward we can [ â ] xfail the test case pending a proper solution to
> > 59448 ?
>
> > Mike do you prefer one of the other two approaches ?
>
> Iâd xfail the test case and mark with the fix consume PR.  If we donât have an unambiguous, fix consume PR, Iâd file that.  It should be listed as failing on aarch, and the fix for that PR should then make the aarch test case pass.  This way no one can run off with the PR and do anything else with it.

Hi, Mike!

Sorry for the delay.

The following patch makes atomic-op-consume.c XFAIL for both 
gcc.target/arm and gcc.target/aarch64 tests. XFAIL was chosen as 
prefered approach for handling __ATOMIC_CONSUME to __ATOMIC_ACQUIRE 
promotion workaround.

This patch was tested by running the modified tests on aarch64-none-elf
and arm-none-eabi compilers.

Is this patch ok?

Alex

2015-02-09  Alex Velenko  <Alex.Velenko@arm.com>

gcc/testsuite/

	* gcc.target/aarch64/atomic-op-consume.c (scan-assember-times):
	Directive adjusted to XFAIL.
	* gcc.target/arm/atomic-op-consume.c (scan-assember-times): Directive
	adjusted to XFAIL.

diff --git a/gcc/testsuite/gcc.target/aarch64/atomic-op-consume.c 
b/gcc/testsuite/gcc.target/aarch64/atomic-op-consume.c
index 38d6c2c..8150af6 100644
--- a/gcc/testsuite/gcc.target/aarch64/atomic-op-consume.c
+++ b/gcc/testsuite/gcc.target/aarch64/atomic-op-consume.c
@@ -3,5 +3,9 @@

 #include "atomic-op-consume.x"

-/* { dg-final { scan-assembler-times "ldxr\tw\[0-9\]+, 
\\\[x\[0-9\]+\\\]" 6 } } */
+/* To workaround Bugzilla 59448 issue, a request for __ATOMIC_CONSUME 
is always
+   promoted to __ATOMIC_ACQUIRE, implemented as MEMMODEL_ACQUIRE.  This 
causes
+   "LDAXR" to be generated instead of "LDXR".  Therefore, "LDXR" test is
+   expected to fail.  */
+/* { dg-final { scan-assembler-times "ldxr\tw\[0-9\]+, 
\\\[x\[0-9\]+\\\]" 6 { xfail *-*-* } } } */
 /* { dg-final { scan-assembler-times "stxr\tw\[0-9\]+, w\[0-9\]+, 
\\\[x\[0-9\]+\\\]" 6 } } */
diff --git a/gcc/testsuite/gcc.target/arm/atomic-op-consume.c 
b/gcc/testsuite/gcc.target/arm/atomic-op-consume.c
index cc6c028..060655c 100644
--- a/gcc/testsuite/gcc.target/arm/atomic-op-consume.c
+++ b/gcc/testsuite/gcc.target/arm/atomic-op-consume.c
@@ -7,7 +7,8 @@

 /* To workaround Bugzilla 59448 issue, a request for __ATOMIC_CONSUME 
is always
    promoted to __ATOMIC_ACQUIRE, implemented as MEMMODEL_ACQUIRE. 
This causes
-   "LDAEX" to be generated instead of "LDREX".  */
-/* { dg-final { scan-assembler-times "ldaex\tr\[0-9\]+, 
\\\[r\[0-9\]+\\\]" 6 } } */
+   "LDAEX" to be generated instead of "LDREX".  Therefore, "LDREX" test is
+   expected to fail.  */
+/* { dg-final { scan-assembler-times "ldrex\tr\[0-9\]+, 
\\\[r\[0-9\]+\\\]" 6 { xfail *-*-* } } } */
 /* { dg-final { scan-assembler-times "strex\t...?, r\[0-9\]+, 
\\\[r\[0-9\]+\\\]" 6 } } */
 /* { dg-final { scan-assembler-not "dmb" } } */

--------------1.8.1.2--