This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] Fixes to testcase for PR tree-optimization/86636


On 27/07/18 16:39, David Malcolm wrote:
> On Thu, 2018-07-26 at 13:22 +0100, Andre Vieira (lists) wrote:
> 
> [...snip...]
> 
>>>> diff --git a/gcc/testsuite/gcc.c-torture/compile/pr86636.c
>>>> b/gcc/testsuite/gcc.c-torture/compile/pr86636.c
>>>> new file mode 100644
>>>> index 0000000..2fe2f70
>>>> --- /dev/null
>>>> +++ b/gcc/testsuite/gcc.c-torture/compile/pr86636.c
>>>> @@ -0,0 +1,8 @@
>>>> +/* { dg-options "-fsave-optimization-record -ftree-loop-
>>>> vectorize -ftree-parallelize-loops=2" } */
>>>> +
>>>> +void
>>>> +n2 (int ih)
>>>> +{
>>>> +  while (ih < 1)
>>>> +    ++ih;
>>>> +}
>>>> --
>>>> 1.8.5.3
>>>>
>>
>> Hi David,
>>
>> I believe the test in this patch needs a "{ dg-require-effective-
>> target
>> pthread }" as -ftree-parallelize-loops seems to be turning on
>> -pthread.
>>
>> Cheers,
>> Andre
> 
> Thanks; indeed I now see that every test case using
>   -ftree-parallelize-loops
> has such a directive.
> 
> Also, the test is missing a dg-do compile.
> 
> Does the following look correct?  (borderline covered by "obvious fix"
> rule).
> 
> Dave
> 
> 
> gcc/testsuite/ChangeLog:
> 	PR tree-optimization/86636
> 	* gcc.c-torture/compile/pr86636.c: Add dg-do compile.
> 	Add dg-require-effective-target pthread.
> ---
>  gcc/testsuite/gcc.c-torture/compile/pr86636.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/gcc/testsuite/gcc.c-torture/compile/pr86636.c b/gcc/testsuite/gcc.c-torture/compile/pr86636.c
> index 2fe2f70..8527a0f 100644
> --- a/gcc/testsuite/gcc.c-torture/compile/pr86636.c
> +++ b/gcc/testsuite/gcc.c-torture/compile/pr86636.c
> @@ -1,3 +1,5 @@
> +/* { dg-do compile } */
> +/* { dg-require-effective-target pthread } */
>  /* { dg-options "-fsave-optimization-record -ftree-loop-vectorize -ftree-parallelize-loops=2" } */
>  
>  void
> 

LGTM, but I am not a maintainer. I too would think you might would be ok
with a "obvious fix" here.

Cheers,
Andre


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]