llvm dev - Jan 2013 - [LLVMdev] Does loop vectorizer inquire about target's SIMD capabilities?

Nadav (or anyone who is familiar with the loop vectorizer),

Does the current loop vectorizer inquire about the SIMD capabilities of the
target architecture when it decides whether it is profitable to vectorize a
loop? I am asking this because I would like to have loop vectorization
disabled for targets that don't support SIMD instructions (for example,
standard mips32). Loop vectorization bloats the code size and prolongs
compilation time without any improvement to performance for such targets.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20130103/ab782fa5/attachment.html>

Hi Akira! 
> 
> Does the current loop vectorizer inquire about the SIMD capabilities of the
target architecture when it decides whether it is profitable to vectorize a
loop?
Yes, it uses a cost model to determine the profitability of vectorization. At
the moment only x86 provides the necessary hooks that are needed for calculating
the costs. We may need to change the cost defaults to prevent vectorization on
targets that don't implement the cost interface. If this is a problem for
you then I can do it soon.
 
> I am asking this because I would like to have loop vectorization disabled
for targets that don't support SIMD instructions (for example, standard
mips32).
> Loop vectorization bloats the code size and prolongs compilation time
without any improvement to performance for such targets.
> 
Yes. Also, notice that the loop vectorizer tries to be more conservative when
the 'optforsize' attribute is used.


Thanks,
Nadav

On 3 January 2013 21:53, Nadav Rotem <nrotem at apple.com> wrote:
> > I am asking this because I would like to have loop vectorization
> disabled for targets that don't support SIMD instructions (for example,
> standard mips32).
> > Loop vectorization bloats the code size and prolongs compilation time
> without any improvement to performance for such targets.
> >
>
> Yes. Also, notice that the loop vectorizer tries to be more conservative
> when the 'optforsize' attribute is used.
>
Isn't the vectorizer disabled by default? Or are you requesting that if the
user chooses -vectorize the compiler prints a warning (unused param since
no SIMD)?

cheers,
--renato
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20130103/d2254921/attachment.html>

Hi Nadav,

On Thu, Jan 3, 2013 at 1:53 PM, Nadav Rotem <nrotem at apple.com> wrote:
> Hi Akira!
>
> >
> > Does the current loop vectorizer inquire about the SIMD capabilities
of
> the target architecture when it decides whether it is profitable to
> vectorize a loop?
>
> Yes, it uses a cost model to determine the profitability of vectorization.
> At the moment only x86 provides the necessary hooks that are needed for
> calculating the costs. We may need to change the cost defaults to prevent
> vectorization on targets that don't implement the cost interface. If
this
> is a problem for you then I can do it soon.
>
>
I guess I can just implement all the vectorTargetTransformInfo::get*OpCost
functions since I will later need a cost model for mips-dsp anyway.

Would the code in LoopVectorizationCostModel::expectedCost work correctly
if those functions returned a large integer (max unsigned int)? I am
concerned about overflow.

> I am asking this because I would like to have loop vectorization disabled
> for targets that don't support SIMD instructions (for example, standard
> mips32).
> > Loop vectorization bloats the code size and prolongs compilation time
> without any improvement to performance for such targets.
> >
>
> Yes. Also, notice that the loop vectorizer tries to be more conservative
> when the 'optforsize' attribute is used.
>
>
> Thanks,
> Nadav
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20130103/94206511/attachment.html>