[PATCH] [LoopVectorize] Move LoopAccessAnalysis to its own module
Adam Nemet
anemet at apple.com
Fri Jan 30 23:45:15 PST 2015
> On Jan 30, 2015, at 10:33 AM, Hal Finkel <hfinkel at anl.gov> wrote:
>
> ----- Original Message -----
>> From: "Adam Nemet" <anemet at apple.com>
>> To: anemet at apple.com, hfinkel at anl.gov, aschwaighofer at apple.com
>> Cc: llvm-commits at cs.uiuc.edu
>> Sent: Friday, January 30, 2015 12:31:44 PM
>> Subject: Re: [PATCH] [LoopVectorize] Move LoopAccessAnalysis to its own module
>>
>> ================
>> Comment at: include/llvm/Analysis/LoopAccessAnalysis.h:37
>> @@ +36,3 @@
>> +/// check that a group of pointers do not overlap.
>> +struct RuntimePointerCheck {
>> + RuntimePointerCheck() : Need(false) {}
>> ----------------
>> hfinkel wrote:
>>> Should this really be a public class? It lacks a proper interface,
>>> and the meaning of the dependency set ids is not really defined.
>> My hope is that in the long run this becomes private to LAA. The
>> problem right now is that the actual code emission happens in
>> InnerLoopVectorizer::addRuntimeCheck which uses an instance of this
>> class.
>>
>> My plan is to move addRuntimeCheck to LAA as well of course and I see
>> no reason why this would have to remain public after that.
>>
>> I can move this class inside LAA. It would of course still be public
>> but at least it would tie it better to LAA in the interim. What do
>> you think?
>
> I think that would be better.
The new review (http://reviews.llvm.org/D7312) covers this. In terms of patch ordering it’s right before the move to the new module.
Thanks again for all your reviews!
Adam
> -Hal
>
>>
>> http://reviews.llvm.org/D7285 <http://reviews.llvm.org/D7285>
>>
>> EMAIL PREFERENCES
>> http://reviews.llvm.org/settings/panel/emailpreferences/ <http://reviews.llvm.org/settings/panel/emailpreferences/>
>>
>>
>>
>
> --
> Hal Finkel
> Assistant Computational Scientist
> Leadership Computing Facility
> Argonne National Laboratory
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20150130/76497caf/attachment.html>
More information about the llvm-commits
mailing list