This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: optimization question
On Tue, May 19, 2009 at 2:02 AM, Daniel Berlin <dberlin@dberlin.org> wrote:
> On Sat, May 16, 2009 at 5:49 AM, Richard Guenther
> <richard.guenther@gmail.com> wrote:
>> On Sat, May 16, 2009 at 11:41 AM, VandeVondele Joost <vondele@pci.uzh.ch> wrote:
>>>
>>>>>> I think it is useful to have a bugzilla here.
>>>>>
>>>>> will do.
>>>
>>> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40168
>>>
>>>>
>>>> Btw, complete unrolling is also hindred by the artificial limit of
>>>> maximally
>>>> unrolling 16 iterations. ?Your inner loops iterate 27 times. ?Also by the
>>>> artificial limit of the maximal unrolled size.
>>>>
>>>> With --param max-completely-peel-times=27 --param
>>>> max-completely-peeled-insns=666
>>>>
>>>> (values for trunk) the loops are unrolled at -O3.
>>>
>>> hmmm. but leading to slower code.
>>
>> Not for me - but well, the main issue is the memsets which the
>> frontend generates.
>>
> We should be able to ignore memset for scalarizing/etc, no?
Sure - we can start to play tricks here. Like not marking things
addressable (and thus aliased) if the address is taken but it
doesn't "escape" to any local or global pointer. Which will probably
be one thing necessary to make scalarization "ignore" memset.
There is lots of low-hanging stuff left regarding to escape analysis
in the face of builtin functions.
Richard.