search for: d45842

There are at least 3 active proposals to add reassociative optimizations in IR: [1] D41574 <https://reviews.llvm.org/D41574>- a new pass for reassociation/factoring [2] D46336 <https://reviews.llvm.org/D46336> - enhance -instcombine to do more reassociation/factoring [3] D45842 <https://reviews.llvm.org/D45842> - add to the existing -reassociate pass to enable factoring Here's a basic motivating example that we fail to optimize: https://godbolt.org/g/64NmJM https://rise4fun.com/Alive/wec Currently, responsibility for reassociative transforms is split between -...

...east 3 active proposals to add reassociative optimizations > in IR: > [1] D41574 <https://reviews.llvm.org/D41574>- a new pass for > reassociation/factoring > [2] D46336 <https://reviews.llvm.org/D46336> - enhance -instcombine to do > more reassociation/factoring > [3] D45842 <https://reviews.llvm.org/D45842> - add to the existing > -reassociate pass to enable factoring > > Here's a basic motivating example that we fail to optimize: > https://godbolt.org/g/64NmJM > https://rise4fun.com/Alive/wec > > Currently, responsibility for reassociat...

...to add reassociative optimizations >> in IR: >> [1] D41574 <https://reviews.llvm.org/D41574>- a new pass for >> reassociation/factoring >> [2] D46336 <https://reviews.llvm.org/D46336> - enhance -instcombine to >> do more reassociation/factoring >> [3] D45842 <https://reviews.llvm.org/D45842> - add to the existing >> -reassociate pass to enable factoring >> >> Here's a basic motivating example that we fail to optimize: >> https://godbolt.org/g/64NmJM >> https://rise4fun.com/Alive/wec >> >> Currently, r...

...optimizer. It's optimizing code that no other pass does currently, and I don't see any other near-term proposal that gets us those optimizations. Omer, can you rebase that to trunk? I think a header has moved, so it doesn't build as-is. I'd like to know if it can catch the cases in D45842. If not, why not? If it can handle those easily, I'll abandon D45842. Also, I don't know if it's better to include that functionality as another iteration of the existing -reassociate or split it off as its own pass. But I think it should do the distributive simplifications that are cu...

...plexity of the instcombine pass. I do acknowledge that D46336 / D46595 (both in instcombine) is an improvement, which results in more folds that didn't happen before. But it makes the already-large inscombine even more complex, and i'm not sure how it affects the performance. Then there is D45842. That code is pretty simple, and takes ~one page. It does not do any change unless that opens the road for further simplification. (I hope that is what the D46336 / D46595 do, but i don't know.) As stated in https://reviews.llvm.org/D46336#1090588, the D45842 does allow instcombine to do //mos...

...gt;>>>> ​ >>>> Let me try one more time :) May we need multiple reassociate passes to >>>> fold different reassociative patterns? >>>> >>>> A longer version: If Sanjay wants a particular reassociative pattern to >>>> be folded (D45842), Omer wants another particular reassociative pattern >>>> to be folded (D41574), and I want yet another ​particular reassociative >>>> pattern to be folded (D46336), would we potentially need three >>>> different reassociate passes with each combined with instco...

...imizations >>> in IR: >>> [1] D41574 <https://reviews.llvm.org/D41574>- a new pass for >>> reassociation/factoring >>> [2] D46336 <https://reviews.llvm.org/D46336> - enhance -instcombine to >>> do more reassociation/factoring >>> [3] D45842 <https://reviews.llvm.org/D45842> - add to the existing >>> -reassociate pass to enable factoring >>> >>> Here's a basic motivating example that we fail to optimize: >>> https://godbolt.org/g/64NmJM >>> https://rise4fun.com/Alive/wec >>&g...

> On May 8, 2018, at 9:50 AM, Daniel Berlin via llvm-dev <llvm-dev at lists.llvm.org> wrote: > > 1. The reassociate pass that exists right now was *originally* (AFAIK) written to enable CSE/GVN to do better. Agreed. The original mindset included a (naive) belief that going with a canonical form was better than teaching redundancy elimination to handle abstractions (as a matter

.... > > This all quickly devolves into herbrand equivalence and it's variations. > > > ​ Let me try one more time :) May we need multiple reassociate passes to fold different reassociative patterns? A longer version: If Sanjay wants a particular reassociative pattern to be folded (D45842), Omer wants another particular reassociative pattern to be folded (D41574), and I want yet another ​particular reassociative pattern to be folded (D46336), would we potentially need three different reassociate passes with each combined with instcombine, rather than just one that may be able to som...

...s into herbrand equivalence and it's variations. >> >> >> ​ > Let me try one more time :) May we need multiple reassociate passes to > fold different reassociative patterns? > > A longer version: If Sanjay wants a particular reassociative pattern to be > folded (D45842), Omer wants another particular reassociative pattern to > be folded (D41574), and I want yet another ​particular reassociative > pattern to be folded (D46336), would we potentially need three > different reassociate passes with each combined with instcombine, rather > than just one tha...

...;> >>>> >>>> ​ >>> Let me try one more time :) May we need multiple reassociate passes to >>> fold different reassociative patterns? >>> >>> A longer version: If Sanjay wants a particular reassociative pattern to >>> be folded (D45842), Omer wants another particular reassociative pattern >>> to be folded (D41574), and I want yet another ​particular reassociative >>> pattern to be folded (D46336), would we potentially need three >>> different reassociate passes with each combined with instcombine, rathe...

On Wed, May 9, 2018 at 10:39 AM, Hiroshi Yamauchi <yamauchi at google.com> wrote: > > > On Tue, May 8, 2018 at 11:15 AM Daniel Berlin <dberlin at dberlin.org> wrote: > >> >> >> On Tue, May 8, 2018 at 10:38 AM, Hiroshi Yamauchi via llvm-dev < >> llvm-dev at lists.llvm.org> wrote: >> >>> ( >>> ​I came across this issue in

...ons. > > > ​ > Let me try one more time :) May we need multiple > reassociate passes to fold different reassociative patterns? > > A longer version: If Sanjay wants a particular > reassociative pattern to be folded (D45842), Omer wants > another particular reassociative pattern to be folded > (D41574), and I want yet another ​particular reassociative > pattern to be folded (D46336), would we potentially need > three different reassociate passes with each com...

...it's variations. >>> >>> >>> ​ >> Let me try one more time :) May we need multiple reassociate passes to >> fold different reassociative patterns? >> >> A longer version: If Sanjay wants a particular reassociative pattern to >> be folded (D45842), Omer wants another particular reassociative pattern >> to be folded (D41574), and I want yet another ​particular reassociative >> pattern to be folded (D46336), would we potentially need three >> different reassociate passes with each combined with instcombine, rather >> t...