I've since run NoFib. You can find the results here: https://phabricator.haskell.org/D4244#119697 I wonder if you feel that any more notes are needed? The general idea of CPR remained the same, it's just the extension of the DmdResult lattice that needs some rationale as to why and when these new values are needed. I also wonder what the impact of "Slightly strengthen the strictness analysis" ( https://ghc.haskell.org/trac/ghc/wiki/NestedCPR/Akio2017#Changestothedemanda...) would be if regarded in isolation. On Tue, Jan 2, 2018 at 11:52 AM, Matthew Pickering < matthewtpickering@gmail.com> wrote:

I don't think anyone has run nofib on the rebased branch yet.

The Akio2017 subpage is a more accurate summary. Sebastian has also been adding notes to explain the more intricate parts.

Matt

...

Terrific!

What are the nofib results?

Can we have a couple of artificial benchmarks in cpranal/should_run that show substantial perf improvements because the nested CPR wins in some inner loop?

Is https://ghc.haskell.org/trac/ghc/wiki/NestedCPR still an accurate

On Fri, Dec 22, 2017 at 5:27 PM, Simon Peyton Jones wrote: summary of the idea? And the Akio2017 sub-page? It would be easier to review the code if the design documentation accurately described it.

...

I'll look in the new year. Thanks!

Simon

| -----Original Message----- | From: Matthew Pickering [mailto:matthewtpickering@gmail.com] | Sent: 22 December 2017 17:09 | To: GHC developers ; Simon Peyton Jones | ; Joachim Breitner ; | tkn.akio@gmail.com; Sebastian Graf | Subject: Nested CPR patch review | | Hi all, | | I recently resurrected akio's nested cpr branch and put it on

...

| for review. | | https://phabricator.haskell.org/D4244 | | Sebastian has kindly been going over it and ironed out a few kinks in

...

| last few days. He says now that he believes the patch is correct. | | Is there anything else which needs to be done before merging this

phabricator the patch?

...

| | Simon, would you perhaps be able to give the patch a look over? | | Cheers, | | Matt

Yes, I apologise for being tardy. It's on my stack. Just swamped. Simon | -----Original Message----- | From: Matthew Pickering [mailto:matthewtpickering@gmail.com] | Sent: 29 January 2018 09:35 | To: Sebastian Graf | Cc: Simon Peyton Jones ; GHC developers ; Joachim Breitner ; | tkn.akio@gmail.com | Subject: Re: Nested CPR patch review | | Would one of you have time to review this patch now? | | Is there something more needed? | | Matt | | On Sun, Jan 7, 2018 at 10:11 AM, Sebastian Graf | wrote: | > I've since run NoFib. You can find the results here: | > https://phabricator.haskell.org/D4244#119697 | > | > I wonder if you feel that any more notes are needed? The general | idea | > of CPR remained the same, it's just the extension of the DmdResult | > lattice that needs some rationale as to why and when these new | values are needed. | > | > I also wonder what the impact of "Slightly strengthen the strictness | > analysis" | > | (https://ghc.haskell.org/trac/ghc/wiki/NestedCPR/Akio2017#Changestothe | > demandanalyzer) | > would be if regarded in isolation. | > | > On Tue, Jan 2, 2018 at 11:52 AM, Matthew Pickering | > wrote: | >> | >> I don't think anyone has run nofib on the rebased branch yet. | >> | >> The Akio2017 subpage is a more accurate summary. Sebastian has also | >> been adding notes to explain the more intricate parts. | >> | >> Matt | >> | >> On Fri, Dec 22, 2017 at 5:27 PM, Simon Peyton Jones | >> wrote: | >> > Terrific! | >> > | >> > What are the nofib results? | >> > | >> > Can we have a couple of artificial benchmarks in | cpranal/should_run | >> > that show substantial perf improvements because the nested CPR | wins | >> > in some inner loop? | >> > | >> > Is https://ghc.haskell.org/trac/ghc/wiki/NestedCPR still an | accurate | >> > summary of the idea? And the Akio2017 sub-page? It would be | easier to | >> > review the code if the design documentation accurately described | it. | >> > | >> > I'll look in the new year. Thanks! | >> > | >> > Simon | >> > | >> > | -----Original Message----- | >> > | From: Matthew Pickering [mailto:matthewtpickering@gmail.com] | >> > | Sent: 22 December 2017 17:09 | >> > | To: GHC developers ; Simon Peyton Jones | >> > | ; Joachim Breitner | >> > | ; tkn.akio@gmail.com; Sebastian Graf | >> > | | >> > | Subject: Nested CPR patch review | >> > | | >> > | Hi all, | >> > | | >> > | I recently resurrected akio's nested cpr branch and put it on | >> > phabricator | >> > | for review. | >> > | | >> > | https://phabricator.haskell.org/D4244 | >> > | | >> > | Sebastian has kindly been going over it and ironed out a few | >> > | kinks in | >> > the | >> > | last few days. He says now that he believes the patch is | correct. | >> > | | >> > | Is there anything else which needs to be done before merging | >> > | this | >> > patch? | >> > | | >> > | Simon, would you perhaps be able to give the patch a look | over? | >> > | | >> > | Cheers, | >> > | | >> > | Matt | > | >