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 the context of > D46336 <https://reviews.llvm.org/D46336>. > > Thanks, Sanjay, for starting this discussion.) > > If > we will > move > reassociation, > or keep additional ones > , > out of instcombine, > open questions for me would be > : > > > 1. Since -reassociate isn't a fixed point pass, >This is fixable, fwiw, without fixpointing it.> we might need to repeat "-instcombine -reassociate" multiple times to > fold > down to what we want (relating to my comment here > <https://reviews.llvm.org/D46336#1087082>). I assumed this isn't not what > we want to do > ? My impression is we don't do a fixed-point with passes? >Well, i mean there is no practical difference between passes that we fixpoint externally and fixpoint internally.> >2.> Since -reassociate needs to come up with one operand order (at least > currently as the only reassociate pass), would there exist a single, > unique operand order that would enable all reassociative/commutative > foldings that we want? >In what way? Are you asking whether there is a single reassociation order that makes all foldings occur in the same operation or something? I don't feel like i understand what you are asking.> If not, could there be conflicts among different reassociation orders > depending on what folding we want? >We actually do try to have a single canonical form at various points in the compiler. We may want to canonicalize differently in different parts of the compiler, but it would be nice to at least be self consistent.> > > On Tue, May 8, 2018 at 9:19 AM Sanjay Patel <spatel at rotateright.com> > wrote: > >> 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 >> -reassociate and -instcombine. I don't know if that split is intentional or >> just how the code evolved. Debug stats for test-suite compiled with -O3 >> show: >> 78K "reassociations" by -instcombine. >> 58K "reassociated" by -reassociate >> >> A debug stat with D45842 showed that that transform fired 1.3K times. >> >> Keep in mind that instcombine runs 1st, runs to fixed-point, and runs 8 >> times in the -O2 pipeline. Reassociate runs 1 time and doesn't run to >> fixed-point. Some transforms are unique to 1 pass or the other, but there >> is some overlapping functionality already there. >> >> So the questions are: >> 1. How do we solve these remaining reassociation optimizations? >> 2. Do we want to consolidate existing reassociation transforms in 1 pass >> or is there value in splitting the folds across multiple passes? >> >> >> >> > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev > >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20180508/f9ae6ca9/attachment.html>
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 the context of >> D46336 <https://reviews.llvm.org/D46336>. >> >> Thanks, Sanjay, for starting this discussion.) >> >> If >> we will >> move >> reassociation, >> or keep additional ones >> , >> out of instcombine, >> open questions for me would be >> : >> >> >> 1. Since -reassociate isn't a fixed point pass, >> > > This is fixable, fwiw, without fixpointing it. >How?> >> we might need to repeat "-instcombine -reassociate" multiple times to >> fold >> down to what we want (relating to my comment here >> <https://reviews.llvm.org/D46336#1087082>). I assumed this isn't not >> what we want to do >> ? My impression is we don't do a fixed-point with passes? >> > > Well, i mean there is no practical difference between passes that we > fixpoint externally and fixpoint internally. > I had the following in mind: Does the pass manager support fixpointing externally? Is there any performance difference? Are people okay with that in general? But if there is no practical difference, I don't see any problem with that :)> >> >> > 2. >> Since -reassociate needs to come up with one operand order (at least >> currently as the only reassociate pass), would there exist a single, >> unique operand order that would enable all reassociative/commutative >> foldings that we want? >> > > In what way? > Are you asking whether there is a single reassociation order that makes > all foldings occur in the same operation or something? > I don't feel like i understand what you are asking. >Does this rephrase help: with the motivating examples (like and-of-shifts or bit check patterns) from the above differentials in mind, can we come up with a single reassociation order that solves all those and all the others that may come up in the future? Would we need different reassociation orders to fold different patterns?> > >> If not, could there be conflicts among different reassociation orders >> depending on what folding we want? >> > > We actually do try to have a single canonical form at various points in > the compiler. > > We may want to canonicalize differently in different parts of the > compiler, but it would be nice to at least be self consistent. > >> >> >> On Tue, May 8, 2018 at 9:19 AM Sanjay Patel <spatel at rotateright.com> >> wrote: >> >>> 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 >>> -reassociate and -instcombine. I don't know if that split is intentional or >>> just how the code evolved. Debug stats for test-suite compiled with -O3 >>> show: >>> 78K "reassociations" by -instcombine. >>> 58K "reassociated" by -reassociate >>> >>> A debug stat with D45842 showed that that transform fired 1.3K times. >>> >>> Keep in mind that instcombine runs 1st, runs to fixed-point, and runs 8 >>> times in the -O2 pipeline. Reassociate runs 1 time and doesn't run to >>> fixed-point. Some transforms are unique to 1 pass or the other, but there >>> is some overlapping functionality already there. >>> >>> So the questions are: >>> 1. How do we solve these remaining reassociation optimizations? >>> 2. Do we want to consolidate existing reassociation transforms in 1 pass >>> or is there value in splitting the folds across multiple passes? >>> >>> >>> >>> >> _______________________________________________ >> LLVM Developers mailing list >> llvm-dev at lists.llvm.org >> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev >> >> >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20180509/e8646ee2/attachment-0001.html>
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 the context of >>> D46336 <https://reviews.llvm.org/D46336>. >>> >>> Thanks, Sanjay, for starting this discussion.) >>> >>> If >>> we will >>> move >>> reassociation, >>> or keep additional ones >>> , >>> out of instcombine, >>> open questions for me would be >>> : >>> >>> >>> 1. Since -reassociate isn't a fixed point pass, >>> >> >> This is fixable, fwiw, without fixpointing it. >> > > How? >Depends on specifically which part you would like to know about ;)> > >> >>> we might need to repeat "-instcombine -reassociate" multiple times to >>> fold >>> down to what we want (relating to my comment here >>> <https://reviews.llvm.org/D46336#1087082>). I assumed this isn't not >>> what we want to do >>> ? My impression is we don't do a fixed-point with passes? >>> >> >> Well, i mean there is no practical difference between passes that we >> fixpoint externally and fixpoint internally. >> > > I had the following in mind: Does the pass manager support fixpointing > externally? Is there any performance difference? Are people okay with that > in general? > > But if there is no practical difference, I don't see any problem with > that :) > > >> >>> >>> >> 2. >>> Since -reassociate needs to come up with one operand order (at least >>> currently as the only reassociate pass), would there exist a single, >>> unique operand order that would enable all reassociative/commutative >>> foldings that we want? >>> >> >> In what way? >> Are you asking whether there is a single reassociation order that makes >> all foldings occur in the same operation or something? >> I don't feel like i understand what you are asking. >> > > Does this rephrase help: with the motivating examples (like and-of-shifts > or bit check patterns) from the above differentials in mind, can we come up > with a single reassociation order that solves all those and all the > others that may come up in the future? Would we need different reassociation > orders to fold different patterns? >It doesn't quite help. When stated that generally, there can be no such ordering at all, that's easy to prove. It is a statically undecidable problem. There is however, a different question and answer to a few related problems that maybe you are really asking? 1. Is there a way to determine and apply the a maximal or nearly-maximal set of folds/graph transforms that could be applied to a given set of code in a sane and principled way -> yes (see, e.g., http://www.cs.cornell.edu/~ross/publications/eqsat/) 2. Is there a way to determine all expressions in the program as it exists that are equivalent or equivalent under constant time constant folding/reassociation, in a reasonable time bound -> yes (not a single easy link, happy to talk about it) Your original question is basically equivalent to Is there a way to determine all expressions in the program as it exists that are equivalent or could be made equivalent through any type of folding that one can think up? The answer to that is "no", it's provable that this is not statically decidable, so the time bound doesn't matter :) You have to limit the possible folding/evaluation you apply in various ways to make this decidable, and then further limit it to make the time bound reasonable. This all quickly devolves into herbrand equivalence and it's variations. -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20180509/946f6ad4/attachment-0001.html>