[llvm] r194459 - Change data structure to memorize computed result in ScalarEvolution

[Hal Finkel]

----- Original Message -----
> 
> 
> 
> 
> On Nov 12, 2013, at 3:05 AM, Chandler Carruth < chandlerc at google.com
> > wrote:
> 
> 
> 
> 
> 
> On Tue, Nov 12, 2013 at 2:52 AM, Wan, Xiaofei < xiaofei.wan at intel.com
> > wrote:
> 
> 
> 
> 
> 
> 
> Sorry, we discussed it privately to get quick response, see emails
> attached.
> 
> 
> Sure.
> 
> 
> Andy, in the future if code review is taking place over email, please
> keep it on the list. It is much healthier for the community and
> avoids confusion like this.
> 
> 
> Wan Xiaofei, when stuff is discussed in person or for whatever reason
> accidentally ends up off-list, it is useful to add at least a note
> to the commit log or to the original review thread for other members
> of the community just so we can have a heads up or summary of the
> discussion.
> 
> 
> My first review was accidentally off-list as I didn’t realize I was
> replying to an off-list message. Sorry.

For the record, this was likewise my fault.

 -Hal

> 
> 
> -Andy
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Thanks
> 
> Wan Xiaofei
> 
> From: Chandler Carruth [mailto: chandlerc at google.com ]
> Sent: Tuesday, November 12, 2013 5:56 PM
> To: Wan, Xiaofei
> Cc: Commit Messages and Patches for LLVM
> Subject: Re: [llvm] r194459 - Change data structure to memorize
> computed result in ScalarEvolution
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> On Tue, Nov 12, 2013 at 1:40 AM, Wan Xiaofei < xiaofei.wan at intel.com
> > wrote:
> 
> 
> Author: xiaofei.wan
> Date: Tue Nov 12 03:40:41 2013
> New Revision: 194459
> 
> URL: http://llvm.org/viewvc/llvm-project?rev=194459&view=rev
> Log:
> Change data structure to memorize computed result in ScalarEvolution
> 
> Replace std::map with SmallVector to memorize the cached result since
> SCEV usually belongs to little Loop/BB
> Linear scan on SmallVector is faster than std::map.
> 
> Code reviewer : Andrew Trick.
> 
> 
> Please revert this.
> 
> 
> 
> 
> 
> Andy never reviewed this on the mailing list. Hal gave one comment,
> which you replied to, and no updates to the code review thread have
> occurred. There was no code review or LGTM. =/ I really like the
> idea of this patch, but you can't just bypass the code review
> process like this!
> 
> 
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
> 

-- 
Hal Finkel
Assistant Computational Scientist
Leadership Computing Facility
Argonne National Laboratory