Page MenuHomePhabricator

Prashanth (Prashanth)
User

Projects

User does not belong to any projects.

User Details

User Since
Aug 11 2015, 7:39 PM (283 w, 4 d)

Recent Activity

Nov 7 2020

Prashanth added a comment to D89566: [LV] Epilogue Vectorization with Optimal Control Flow.

Just checking/summarising: is there consensus that this is the patch that is going to do epilogue vectorisation, and not D88819? Or does that depend on perf numbers as Florian suggested?

I think we should go with this, unless @fhahn finds that D88819 can outperform it (unlikely). There are more loops that can be transformed with this patch and I've listed other advantages in the description. @mivnay do you agree?

Are you blocking D88819 without running benchmarks? i.e. Without a comparison on any one of the architectures. For x264, D88819 might give better numbers....

Have you been following the conversations in D88819 and this revision? Blocking work has certainly not been my intention. I just resigned from D88819, in case people would like to move ahead with it. In terms of testing I've verified that x264 opportunity can be caught on POWER, however I do not have access to other architectures, which is why I would appreciate help on testing on other platforms.

Nov 7 2020, 9:17 AM · Restricted Project
Prashanth added a comment to D89566: [LV] Epilogue Vectorization with Optimal Control Flow.

I also think "blocking" is not the right terminology here. But I was asking about this because first we had D88819, then came this one, so we have 2 options. In the end it's pretty simple I guess, because we go for the one with the best code-gen, and this patch looks like the one with the most potential. But would be nice to get that confirmed.

Nov 7 2020, 9:14 AM · Restricted Project

Nov 6 2020

Prashanth added a comment to D89566: [LV] Epilogue Vectorization with Optimal Control Flow.

Just checking/summarising: is there consensus that this is the patch that is going to do epilogue vectorisation, and not D88819? Or does that depend on perf numbers as Florian suggested?

I think we should go with this, unless @fhahn finds that D88819 can outperform it (unlikely). There are more loops that can be transformed with this patch and I've listed other advantages in the description. @mivnay do you agree?

Are you blocking D88819 without running benchmarks? i.e. Without a comparison on any one of the architectures. For x264, D88819 might give better numbers....

Have you been following the conversations in D88819 and this revision? Blocking work has certainly not been my intention. I just resigned from D88819, in case people would like to move ahead with it. In terms of testing I've verified that x264 opportunity can be caught on POWER, however I do not have access to other architectures, which is why I would appreciate help on testing on other platforms.

Nov 6 2020, 10:46 AM · Restricted Project
Prashanth added a comment to D89566: [LV] Epilogue Vectorization with Optimal Control Flow.

Just checking/summarising: is there consensus that this is the patch that is going to do epilogue vectorisation, and not D88819? Or does that depend on perf numbers as Florian suggested?

I think we should go with this, unless @fhahn finds that D88819 can outperform it (unlikely). There are more loops that can be transformed with this patch and I've listed other advantages in the description. @mivnay do you agree?

Nov 6 2020, 10:04 AM · Restricted Project

Oct 27 2020

Prashanth added a comment to D89395: [Flang][OpenMP 4.5] Add semantic check for OpenMP shared and private clause.
Oct 27 2020, 10:25 AM · Restricted Project, Restricted Project, Restricted Project