Re: a patch for PR68695

[Kyrill Tkachov <kyrylo.tkachov@foss.arm.com>]

On 15/04/16 17:18, Jeff Law wrote:
> On 04/15/2016 05:06 AM, Kyrill Tkachov wrote:
>>
>> On 05/04/16 23:35, Segher Boessenkool wrote:
>>> On Tue, Apr 05, 2016 at 10:48:58AM +0100, Kyrill Tkachov wrote:
>>>> So for the test gcc.dg/pr10474.c on arm with -marm -O3 before this
>>>> patch we
>>>> perform shrink-wrapping:
>>>>      cmp    r0, #0
>>>>      bxeq    lr
>>>>      push    {r4, lr}
>>>>      mov    r4, r0
>>>>      ...
>>>>
>>>> And after the patch we don't:
>>>>      push    {r4, lr}
>>>>      subs    r4, r0, #0
>>>>      popeq    {r4, pc}
>>>>      ...
>>>>
>>>> The assembly after the "..." is identical.
>>>>
>>>> So the resulting code is indeed shorter, though there is an
>>>> extra stack push and pop on the early return path.
>>>> A similar effect appears on gcc.dg/ira-shrinkwrap-prep-2.c.
>>> The "new" code is better if there is no shrink-wrapping.  We can probably
>>> teach prepare_shrink_wrap to do the extra register move if that will
>>> allow
>>> us to wrap more.
>>>
>>>> Though if so, it looks like a shrink-wrapping deficiency exposed by
>>>> this patch, rather than caused by it.
>>> Yes, and mostly a testcase problem even.
>>>
>>>> Jakub, do you happen to have the before and after codegen for these
>>>> tests
>>>> on ppc64? I wonder if the effect is more clearcut there.
>>> RTL before shrink-wrapping would be useful, too.
>>
>> So what shall we do for these tests for GCC 6?
>> Add an XFAIL for arm and powerpc?
> We could just punt gcc-6 and focus on what we want for gcc-7 as this isn't a release critical issue.
>

This was resolved with:
https://gcc.gnu.org/ml/gcc-patches/2016-04/msg00724.html

Sorry, I should have replied to this thread...

Kyrill

> jeff