Patchwork [ARM,testsuite] Enable vect_condition tests for ARM/NEON

login
register
mail settings
Submitter Ulrich Weigand
Date Feb. 21, 2012, 6:26 p.m.
Message ID <201202211826.q1LIQqLG028085@d06av02.portsmouth.uk.ibm.com>
Download mbox | patch
Permalink /patch/142346/
State New
Headers show

Comments

Ulrich Weigand - Feb. 21, 2012, 6:26 p.m.
Hello,

I just noticed that check_effective_target_vect_condition returns false
for ARM/NEON, even though the platforms in fact supports vectorized
conditional expressions.  This causes a number of tests to be skipped
unnecessarily.

Fixed by the following patch.   Tested on arm-linux-gnueabi with no
regressions, adds about 130 PASSes.

OK for mainline?

Bye,
Ulrich


testsuite/ChangeLog:

	* lib/target-supports.exp (check_effective_target_vect_condition):
	Return true for NEON.
Mike Stump - Feb. 21, 2012, 7:37 p.m.
On Feb 21, 2012, at 10:26 AM, Ulrich Weigand wrote:
> I just noticed that check_effective_target_vect_condition returns false
> for ARM/NEON, even though the platforms in fact supports vectorized
> conditional expressions.  This causes a number of tests to be skipped
> unnecessarily.
> 
> Fixed by the following patch.   Tested on arm-linux-gnueabi with no
> regressions, adds about 130 PASSes.
> 
> OK for mainline?

Ok.  I'm sure someone will scream soon if it is a bad idea.  :-)
Ramana Radhakrishnan - Feb. 29, 2012, 12:08 a.m.
On Tue, Feb 21, 2012 at 11:37:07AM -0800, Mike Stump wrote:
> On Feb 21, 2012, at 10:26 AM, Ulrich Weigand wrote:
> > I just noticed that check_effective_target_vect_condition returns false
> > for ARM/NEON, even though the platforms in fact supports vectorized
> > conditional expressions.  This causes a number of tests to be skipped
> > unnecessarily.
> > 
> > Fixed by the following patch.   Tested on arm-linux-gnueabi with no
> > regressions, adds about 130 PASSes.
> > 
> > OK for mainline?
> 
> Ok.  I'm sure someone will scream soon if it is a bad idea.  :-)

Certainly no objections to it - in fact I'd suggest it for 4.6 as well.
vcond is working for a while now.

Ramana
Mike Stump - Feb. 29, 2012, 2:11 a.m.
On Feb 28, 2012, at 4:08 PM, Ramana Radhakrishnan wrote:
> Certainly no objections to it - in fact I'd suggest it for 4.6 as well.

The backport is fine, if someone wants to do the work.  Just let it bake a little in the top of the tree as usual first.
Ramana Radhakrishnan - Feb. 29, 2012, 1:08 p.m.
On 29 February 2012 13:06, Ulrich Weigand <uweigand@de.ibm.com> wrote:
> Mike Stump wrote:
>> On Feb 28, 2012, at 4:08 PM, Ramana Radhakrishnan wrote:
>> > Certainly no objections to it - in fact I'd suggest it for 4.6 as well.
>>
>> The backport is fine, if someone wants to do the work.  Just let it bake a =
>> little in the top of the tree as usual first.
>
> I've already checked the patch in to 4.6 too -- the tests newly enabled
> by it all pass there as well.

Cool thanks.

Ramana

>
> Bye,
> Ulrich
>
> --
>  Dr. Ulrich Weigand
>  GNU Toolchain for Linux on System z and Cell BE
>  Ulrich.Weigand@de.ibm.com
>

Patch

Index: gcc/testsuite/lib/target-supports.exp
===================================================================
--- gcc/testsuite/lib/target-supports.exp	(revision 184398)
+++ gcc/testsuite/lib/target-supports.exp	(working copy)
@@ -3430,7 +3430,8 @@ 
 	     || [istarget ia64-*-*]
 	     || [istarget i?86-*-*]
 	     || [istarget spu-*-*]
-	     || [istarget x86_64-*-*] } {
+	     || [istarget x86_64-*-*]
+	     || ([istarget arm*-*-*] && [check_effective_target_arm_neon_ok]) } {
 	   set et_vect_cond_saved 1
 	}
     }