Re: [Haskell-cafe] Parallel Profiling
Hi Amos, Thanks very much - I am taking a look. Dominic Steinitz domi...@steinitz.org http://idontgetoutmuch.wordpress.com On 30 Mar 2015, at 22:05, Amos Robinson wrote: > Hi Dominic, > > A few years ago we wrote a program for analysing DPH runs, dph-event-seer. It > provides a few general analyses like percent of time with N threads running, > time between wake-ups etc. You might find it interesting, but I haven't > actually looked at ghc-events-analyse, so I don't know what it provides. > > I'm sorry, but to compile it without DPH you'd have to modify it to remove > DphOps*. > > https://github.com/ghc/packages-dph/blob/master/dph-event-seer/src/Main.hs > > Amos > > On Tue, 31 Mar 2015 at 04:38 Dominic Steinitz wrote: > Does anyone know of any tools for analysing parallel program performance? > > I am trying to use threadscope but it keeps crashing with my 100M log file > and ghc-events-analyze is not going to help as I have many hundreds of > threads all carrying out the same computation. I think I’d like a library > that would allow me to construct my own analyses rather than display them via > GTK. There is ghc-events but that seems to be just for parsing the logs and I > couldn’t find anything that used it in the way I would like to (apart from > threadscope and ghc-events-analyze of course). > > Thanks > > Dominic Steinitz > domi...@steinitz.org > http://idontgetoutmuch.wordpress.com > > ___ > Haskell-Cafe mailing list > haskell-c...@haskell.org > http://mail.haskell.org/cgi-bin/mailman/listinfo/haskell-cafe ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Re: Binary bloat in 7.10
Mind you I'm just trying to come up with theories we can test, I'm not assigning blame. :) I'm not sure how to do the apples to apples comparison, but it sounds like some sleuthing Is In order. I dont have a 7.10 setup yet, but if someone can put a tarballed dist build folder for a 7.10 and the same for 7.8 for some lib that blows up,I'm happy to try to dig i n. Or I'll get things setup later this week to do the full comparison locally. On Apr 1, 2015 11:12 AM, "Edward Z. Yang" wrote: > Yes, this does seem like a potential culprit, although > we did do some measurements and I didn't think it was too bad. > Maybe we were wrong! > > Edward > > Excerpts from Jeremy's message of 2015-04-01 07:26:55 -0700: > > Carter Schonwald wrote > > > How much of this might be attributable to longer linker symbol names? > Ghc > > > 7.10 object code does have larger symbols! Is there a way to easily > > > tabulate that? > > > > That would explain why the hi files have also increased many-fold. Is > there > > any way to avoid the larger symbols? > > > ___ > Glasgow-haskell-users mailing list > Glasgow-haskell-users@haskell.org > http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users > ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Re: Binary bloat in 7.10
On Wed, Apr 01, 2015 at 02:30:49AM -0700, Jeremy wrote: > Why do the 7.10 libraries take up so much more space than 7.8? For example, > using the same build options and strip --strip-unneeded, 7.8 leaves me with That would be some kind of harsh april 1st joke, if everything compiled at that day gets a bloated data section by putting lots "april" strings into it. ;) Greetings, Daniel ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Re: Binary bloat in 7.10
Yes, this does seem like a potential culprit, although we did do some measurements and I didn't think it was too bad. Maybe we were wrong! Edward Excerpts from Jeremy's message of 2015-04-01 07:26:55 -0700: > Carter Schonwald wrote > > How much of this might be attributable to longer linker symbol names? Ghc > > 7.10 object code does have larger symbols! Is there a way to easily > > tabulate that? > > That would explain why the hi files have also increased many-fold. Is there > any way to avoid the larger symbols? > ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Re: Binary bloat in 7.10
Carter Schonwald wrote > How much of this might be attributable to longer linker symbol names? Ghc > 7.10 object code does have larger symbols! Is there a way to easily > tabulate that? That would explain why the hi files have also increased many-fold. Is there any way to avoid the larger symbols? -- View this message in context: http://haskell.1045720.n5.nabble.com/Binary-bloat-in-7-10-tp5768067p5768095.html Sent from the Haskell - Glasgow-haskell-users mailing list archive at Nabble.com. ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Re: Binary bloat in 7.10
How much of this might be attributable to longer linker symbol names? Ghc 7.10 object code does have larger symbols! Is there a way to easily tabulate that? On Apr 1, 2015 9:40 AM, "Jeremy" wrote: > Roman Cheplyaka-2 wrote > > I'm not denying (or confirming) your claim, but it would look more > > legitimate if you compared the same version of Cabal compiled with > > different versions of GHC. > > > > At least some of this bloat could be because Cabal simply gained more > > code. > > I was going to prove you wrong by identifying packages which have barely > changed for 7.10 ... and found that those packages were a similar size to > their 7.8 versions. > > However, the size increase in other packages is huge, and "simply gained > more code" doesn't seem like an adequate explanation, with some package > more > than doubling. > > Here are the full results: > > 7.8: > > 34M Cabal-1.18.1.5 > 3.8Marray-0.5.0.0 > 50M base-4.7.0.2 > 52M bin > 368Kbin-package-db-0.0.0.0 > 2.7Mbinary-0.7.1.0 > 5.4Mbytestring-0.10.4.0 > 9.4Mcontainers-0.5.5.1 > 196Kdeepseq-1.3.0.2 > 608Kdirectory-1.2.1.0 > 740Kfilepath-1.3.0.2 > 105Mghc-7.8.4 > 2.7Mghc-prim-0.3.1.0 > 8.7Mhaskeline-0.7.1.2 > 3.4Mhoopl-3.10.0.1 > 1020K hpc-0.6.0.1 > 556Kinteger-gmp-0.5.1.0 > 680Kpretty-1.1.1.1 > 684Kprocess-1.2.0.0 > 1.6Mrts-1.0 > 13M template-haskell-2.9.0.0 > 1.4Mterminfo-0.4.0.0 > 6.1Mtime-1.4.2 > 4.4Mtransformers-0.3.0.0 > 5.2Munix-2.7.0.1 > > 7.10: > > 83M Cabal_HWT8QvVfJLn2ubvobpycJY > 3.7Marray_FaHmcBFfuRM8kmZLEY8D5S > 52M base_I5BErHzyOm07EBNpKBEeUv > 56M bin > 2.9Mbinar_EKE3c9Lmxb3DQpU0fPtru6 > 832Kbinpa_JNoexmBMuO8C771QaIy3YN > 5.7Mbytes_6vj5EoliHgNHISHCVCb069 > 11M conta_47ajk3tbda43DFWyeF3oHQ > 432Kdeeps_FpR4obOZALU1lutWnrBldi > 912Kdirec_3TcTyYedch32o1zTH2MR00 > 796Kfilep_5HhyRonfEZoDO205Wm9E4h > 113Mghc_EMlWrQ42XY0BNVbSrKixqY > 2.9Mghcpr_8TmvWUcS1U1IKHT0levwg3 > 8.9Mhaske_IlDhIe25uAn0WJY379Nu1M > 3.4Mhoopl_JxODiSRz1e84NbH6nnZuUk > 1.1Mhpc_CmUUQl5bURfBueJrdYfNs3 > 1.3Minteg_2aU3IZNMF9a7mQ0OzsZ0dS > 1.8Mprett_7jIfj8VCGFf1WS0tIQ1XSZ > 764Kproce_0hwN3CTKynhHQqQkChnSdH > 1.7Mrts > 19M templ_BVMCZyLwIlfGfcqqzyUAI8 > 1.4Mtermi_7qZwBlx3clR8sTBilJl253 > 6.2Mtime_Hh2clZW6in4HpYHx5bLtb7 > 7.3Mtrans_ALYlebOVzVI4kxbFX5SGhm > 5.4Munix_G4Yo1pNtYrk8nCq1cx8P9d > > > > > -- > View this message in context: > http://haskell.1045720.n5.nabble.com/Binary-bloat-in-7-10-tp5768067p5768083.html > Sent from the Haskell - Glasgow-haskell-users mailing list archive at > Nabble.com. > ___ > Glasgow-haskell-users mailing list > Glasgow-haskell-users@haskell.org > http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users > ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Re: Binary bloat in 7.10
Roman Cheplyaka-2 wrote > I'm not denying (or confirming) your claim, but it would look more > legitimate if you compared the same version of Cabal compiled with > different versions of GHC. > > At least some of this bloat could be because Cabal simply gained more > code. I was going to prove you wrong by identifying packages which have barely changed for 7.10 ... and found that those packages were a similar size to their 7.8 versions. However, the size increase in other packages is huge, and "simply gained more code" doesn't seem like an adequate explanation, with some package more than doubling. Here are the full results: 7.8: 34M Cabal-1.18.1.5 3.8Marray-0.5.0.0 50M base-4.7.0.2 52M bin 368Kbin-package-db-0.0.0.0 2.7Mbinary-0.7.1.0 5.4Mbytestring-0.10.4.0 9.4Mcontainers-0.5.5.1 196Kdeepseq-1.3.0.2 608Kdirectory-1.2.1.0 740Kfilepath-1.3.0.2 105Mghc-7.8.4 2.7Mghc-prim-0.3.1.0 8.7Mhaskeline-0.7.1.2 3.4Mhoopl-3.10.0.1 1020K hpc-0.6.0.1 556Kinteger-gmp-0.5.1.0 680Kpretty-1.1.1.1 684Kprocess-1.2.0.0 1.6Mrts-1.0 13M template-haskell-2.9.0.0 1.4Mterminfo-0.4.0.0 6.1Mtime-1.4.2 4.4Mtransformers-0.3.0.0 5.2Munix-2.7.0.1 7.10: 83M Cabal_HWT8QvVfJLn2ubvobpycJY 3.7Marray_FaHmcBFfuRM8kmZLEY8D5S 52M base_I5BErHzyOm07EBNpKBEeUv 56M bin 2.9Mbinar_EKE3c9Lmxb3DQpU0fPtru6 832Kbinpa_JNoexmBMuO8C771QaIy3YN 5.7Mbytes_6vj5EoliHgNHISHCVCb069 11M conta_47ajk3tbda43DFWyeF3oHQ 432Kdeeps_FpR4obOZALU1lutWnrBldi 912Kdirec_3TcTyYedch32o1zTH2MR00 796Kfilep_5HhyRonfEZoDO205Wm9E4h 113Mghc_EMlWrQ42XY0BNVbSrKixqY 2.9Mghcpr_8TmvWUcS1U1IKHT0levwg3 8.9Mhaske_IlDhIe25uAn0WJY379Nu1M 3.4Mhoopl_JxODiSRz1e84NbH6nnZuUk 1.1Mhpc_CmUUQl5bURfBueJrdYfNs3 1.3Minteg_2aU3IZNMF9a7mQ0OzsZ0dS 1.8Mprett_7jIfj8VCGFf1WS0tIQ1XSZ 764Kproce_0hwN3CTKynhHQqQkChnSdH 1.7Mrts 19M templ_BVMCZyLwIlfGfcqqzyUAI8 1.4Mtermi_7qZwBlx3clR8sTBilJl253 6.2Mtime_Hh2clZW6in4HpYHx5bLtb7 7.3Mtrans_ALYlebOVzVI4kxbFX5SGhm 5.4Munix_G4Yo1pNtYrk8nCq1cx8P9d -- View this message in context: http://haskell.1045720.n5.nabble.com/Binary-bloat-in-7-10-tp5768067p5768083.html Sent from the Haskell - Glasgow-haskell-users mailing list archive at Nabble.com. ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Re: Binary bloat in 7.10
Roman Cheplyaka-2 wrote > I'm not denying (or confirming) your claim, but it would look more > legitimate if you compared the same version of Cabal compiled with > different versions of GHC. > > At least some of this bloat could be because Cabal simply gained more > code. Tricky to test that because of dependencies and global package db. I haven't measured the amount of code in Cabal, but I doubt it's increased that much, and there has been a big jump in the installed size of every library. -- View this message in context: http://haskell.1045720.n5.nabble.com/Binary-bloat-in-7-10-tp5768067p5768080.html Sent from the Haskell - Glasgow-haskell-users mailing list archive at Nabble.com. ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Increased memory usage with GHC 7.10.1
Forall hi, I just uprgaded both of my machines to use GHC 7.10.1. I keep sandboxed installations of GHC and this means I had to rebuild Agda and Idris because the binaries built with GHC 7.8.4 were stored inside deactivated 7.8.4 sandbox. Sadly, I had problems building both Agda and Idris due to GHC taking up all of available memory. With Idris the problematic module was Idris.ElabTerm (~2900LOC). The interesting part of the story is that when I do a clean build of Idris GHC consumes all of memory when compiling that module and I have to kill the build. But when I restart the build after killing GHC the module is compiled using a reasonable amount of memory and within reasonable time. With Agda the problematic module is Agda.TypeChecking.Serialise (~2000LOC). The trick with killing the build and restarting it didn't work in this case. I had to compile Agda with GHC 7.8.4 (which works without problems though the mentioned module still requires a lot of memory) and alter my setup so that Agda binary is not stored inside GHC sandbox. I wonder if any of you came across similar issues with GHC 7.10.1? Do we have any performance data that allows to compare memory usage and performance of GHC 7.10.1 with previous stable releases? All of the above happened on 64bit Debian Wheezy with 2GB of RAM. Janek --- Politechnika Łódzka Lodz University of Technology Treść tej wiadomości zawiera informacje przeznaczone tylko dla adresata. Jeżeli nie jesteście Państwo jej adresatem, bądź otrzymaliście ją przez pomyłkę prosimy o powiadomienie o tym nadawcy oraz trwałe jej usunięcie. This email contains information intended solely for the use of the individual to whom it is addressed. If you are not the intended recipient or if you have received this message in error, please notify the sender and delete it from your system. ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Re: Binary bloat in 7.10
On 01/04/15 12:30, Jeremy wrote: > Why do the 7.10 libraries take up so much more space than 7.8? For example, > using the same build options and strip --strip-unneeded, 7.8 leaves me with > > 15M libHSCabal-1.18.1.5.a > 17M HSCabal-1.18.1.5.o > > whereas 7.10 balloons to > > 23M HSCabal-1.22.2.0-HWT8QvVfJLn2ubvobpycJY.o > 53M libHSCabal-1.22.2.0-HWT8QvVfJLn2ubvobpycJY.a I'm not denying (or confirming) your claim, but it would look more legitimate if you compared the same version of Cabal compiled with different versions of GHC. At least some of this bloat could be because Cabal simply gained more code. Roman ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Re: Binary bloat in 7.10
Karel Gardas wrote > 7.10.1 should IIRC support some kind of DWARF debugging information and > IIRC it was mentioned and decided on ghc devel that the libraries will > ship with some DWARF to easy debugging > > -- but takes me lightly on it and verify if this is the case since I may > be completely off and this may apply to GHC HEAD and not to 7.10.x Stripped all debugging, didn't help. -- View this message in context: http://haskell.1045720.n5.nabble.com/Binary-bloat-in-7-10-tp5768067p5768077.html Sent from the Haskell - Glasgow-haskell-users mailing list archive at Nabble.com. ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Re: Binary bloat in 7.10
7.10.1 should IIRC support some kind of DWARF debugging information and IIRC it was mentioned and decided on ghc devel that the libraries will ship with some DWARF to easy debugging -- but takes me lightly on it and verify if this is the case since I may be completely off and this may apply to GHC HEAD and not to 7.10.x Karel On 04/ 1/15 11:30 AM, Jeremy wrote: Why do the 7.10 libraries take up so much more space than 7.8? For example, using the same build options and strip --strip-unneeded, 7.8 leaves me with 15M libHSCabal-1.18.1.5.a 17M HSCabal-1.18.1.5.o whereas 7.10 balloons to 23M HSCabal-1.22.2.0-HWT8QvVfJLn2ubvobpycJY.o 53M libHSCabal-1.22.2.0-HWT8QvVfJLn2ubvobpycJY.a -- View this message in context: http://haskell.1045720.n5.nabble.com/Binary-bloat-in-7-10-tp5768067.html Sent from the Haskell - Glasgow-haskell-users mailing list archive at Nabble.com. ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Re: Binary bloat in 7.10
It's not just binaries, even hi files have ballooned. (I should note that (stripped) executables appear to be unaffected.) -- View this message in context: http://haskell.1045720.n5.nabble.com/Binary-bloat-in-7-10-tp5768067p5768072.html Sent from the Haskell - Glasgow-haskell-users mailing list archive at Nabble.com. ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users
Binary bloat in 7.10
Why do the 7.10 libraries take up so much more space than 7.8? For example, using the same build options and strip --strip-unneeded, 7.8 leaves me with 15M libHSCabal-1.18.1.5.a 17M HSCabal-1.18.1.5.o whereas 7.10 balloons to 23M HSCabal-1.22.2.0-HWT8QvVfJLn2ubvobpycJY.o 53M libHSCabal-1.22.2.0-HWT8QvVfJLn2ubvobpycJY.a -- View this message in context: http://haskell.1045720.n5.nabble.com/Binary-bloat-in-7-10-tp5768067.html Sent from the Haskell - Glasgow-haskell-users mailing list archive at Nabble.com. ___ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/glasgow-haskell-users