> On Thu, Feb 27, 2014 at 2:53 PM, Andreas Schwab wrote:
> > Jason Merrill writes:
> >
> >> Hmm, I wonder why we aren't devirtualizing that call on ia64. Does it
> >> work with -O3?
> >
> > That doesn't change anything fundamentally.
>
> I think the vtable lookup sequence is different and nobod
On Thu, Feb 27, 2014 at 04:00:22PM +0100, Richard Biener wrote:
> On Thu, Feb 27, 2014 at 3:51 PM, Jason Merrill wrote:
> > On 02/27/2014 09:03 AM, Richard Biener wrote:
> >>>
> >>> Jason Merrill writes:
>
> Hmm, I wonder why we aren't devirtualizing that call on ia64.
> >>
> >> I think
On Thu, Feb 27, 2014 at 3:51 PM, Jason Merrill wrote:
> On 02/27/2014 09:03 AM, Richard Biener wrote:
>>>
>>> Jason Merrill writes:
Hmm, I wonder why we aren't devirtualizing that call on ia64.
>>
>> I think the vtable lookup sequence is different and nobody cared to adjust
>>
>> the gi
On 02/27/2014 09:03 AM, Richard Biener wrote:
Jason Merrill writes:
Hmm, I wonder why we aren't devirtualizing that call on ia64.
I think the vtable lookup sequence is different and nobody cared to adjust
the gimple matcher to also match the ia64 sequence.
Ah. So xfail on ia64?
Jason
On Thu, Feb 27, 2014 at 2:53 PM, Andreas Schwab wrote:
> Jason Merrill writes:
>
>> Hmm, I wonder why we aren't devirtualizing that call on ia64. Does it
>> work with -O3?
>
> That doesn't change anything fundamentally.
I think the vtable lookup sequence is different and nobody cared to adjust
Jason Merrill writes:
> Hmm, I wonder why we aren't devirtualizing that call on ia64. Does it
> work with -O3?
That doesn't change anything fundamentally.
Andreas.
--
Andreas Schwab, SUSE Labs, sch...@suse.de
GPG Key fingerprint = 0196 BAD8 1CE9 1970 F4BE 1748 E4D4 88E3 0EEA B9D7
"And now f
Hmm, I wonder why we aren't devirtualizing that call on ia64. Does it
work with -O3?
Jason
Jason Merrill writes:
> diff --git a/gcc/testsuite/g++.dg/opt/devirt4.C
> b/gcc/testsuite/g++.dg/opt/devirt4.C
> new file mode 100644
> index 000..26e8ee6
> --- /dev/null
> +++ b/gcc/testsuite/g++.dg/opt/devirt4.C
> @@ -0,0 +1,16 @@
> +// PR c++/53808
> +// Devirtualization + inlining should
> But this caused bug 60347: turns out that we shouldn't do this
> unless the vtable (and thus the contents of the vtable) are used.
The ipa-devirt type inheritance builder will use any vtable it finds in
DECL_BINFO
of types that it knows about. It starts with types of virtual methods and
virtua
On 02/25/2014 01:53 PM, Jason Merrill wrote:
The primary bug under discussion in 53808 has been fixed separately, but
it also pointed out that once devirtualization resolves the delete to
use the bar destructor, we ought to be able to inline that destructor.
So if we're devirtualizing, always add
10 matches
Mail list logo