I think better now. Simon
| -----Original Message----- | From: Crutcher Dunnavant [mailto:[email protected]] | Sent: 30 October 2009 16:57 | To: Simon Peyton-Jones | Cc: Ian Lynagh; BuildBot Collator | Subject: Re: Looking at GHC bugs | | So, actually there's another blow-up. This seems to be saying that the | ASSERT2() is broken, but I don't see how. | | compiler/coreSyn/CoreUnfold.lhs:1073:4: parse error on input `if' | | 1069 #ifdef DEBUG | 1070 let dump_doc = vcat [ppr dc, ppr dc_univ_tyvars, ppr dc_ex_tyvars, | 1071 ppr arg_tys, ppr dc_args, ppr | _dc_univ_args, | 1072 ppr ex_args, ppr val_args] | 1073 ASSERT2( coreEqType from_ty (mkTyConApp dc_tc _dc_univ_args), | dump_doc ) | 1074 ASSERT2( all isTypeArg (ex_args ++ co_args), dump_doc ) | 1075 ASSERT2( equalLength val_args arg_tys, dump_doc ) | 1076 #endif | | On Fri, Oct 30, 2009 at 9:46 AM, Simon Peyton-Jones | <[email protected]> wrote: | > Ah yes. I see that Ian has pushed a patch. So you are good to go. | > | > Obviously no one else is building with 6.8, but it is supposed to work | > | > S | > | > | -----Original Message----- | > | From: Crutcher Dunnavant [mailto:[email protected]] | > | Sent: 30 October 2009 16:05 | > | To: Simon Peyton-Jones | > | Cc: Simon Marlow; [email protected]; Don Stewart; Ian Lynagh; BuildBot | > | Collator | > | Subject: Re: Looking at GHC bugs | > | | > | Okay, looks like it was just this line: | > | | > | $ darcs diff compiler/specialise/SpecConstr.lhs | > | 55a56 | > | > #if __GLASGOW_HASKELL__ > 609 | > | 56a58,60 | > | > #else | > | > import Data.Generics ( Data, Typeable ) | > | > #endif | > | | > | I don't know enough about darcs to send this as a patch to anyone yet. | > | | > | On Fri, Oct 30, 2009 at 3:22 AM, Simon Peyton-Jones | > | <[email protected]> wrote: | > | > Does anyone have any idea what is going on here? | > | > | > | > The import of Data.Data is in utils/Serialized, but it looks like this: | > | > | > | > #if __GLASGOW_HASKELL__ > 609 | > | > import Data.Data | > | > #else | > | > import Data.Generics | > | > #endif | > | > | > | > If you are compiling with GHC 6.8, then __GLASGOW_HASKELL__ should be set to | 608, | > | so you should have imported Data.Generics. Why doesn't that happen? I wonder | what | > | __GLASGOW_HASKELL__ is set to? It's set on the command line, so you should be | able | > | to see it. | > | > | > | > Anyway, that's the place to look. | > | > | > | > Simon | > | > | > | > | -----Original Message----- | > | > | From: Crutcher Dunnavant [mailto:[email protected]] | > | > | Sent: 29 October 2009 22:35 | > | > | To: Simon Marlow | > | > | Cc: Simon Peyton-Jones; [email protected]; Don Stewart; Ian Lynagh; | > | BuildBot | > | > | Collator | > | > | Subject: Re: Looking at GHC bugs | > | > | | > | > | Meh. I'm becoming somewhat frustrated. I can schedule a day for this | > | > | every week, but every time I come back to this, something has changed | > | > | to rot my build. At least I'm getting to it faster every time, but I'm | > | > | not actually getting any _work_ done on these bugs. | > | > | | > | > | I've attached my current problems. I really can schedule one day a | > | > | week on this (my Google 20% time), and I could maybe provide some | > | > | productive work if I could get enough hand holding to get over the | > | > | learning hump with the build system. | > | > | | > | > | This is the tail of my history: | > | > | 2554 rm -rf ghc | > | > | 2555 tar -xjvf ghc-HEAD-2009-10-23-ghc-corelibs-testsuite.tar.bz2 | > | > | 2556 cd ghc | > | > | 2557 darcs pull -a | > | > | 2558 chmod +x ./darcs-all | > | > | 2559 ./darcs-all --help | > | > | 2560 ./darcs-all get -a --nofib | > | > | 2561 vim darcs-all | > | > | 2562 ./darcs-all --nofib get -a | > | > | 2563 ./darcs-all --nofib pull -a | > | > | 2564 sh boot | > | > | 2565 cp ../ghc.old/mk/build.mk mk/ | > | > | 2566 sh boot | > | > | 2567 ./configure | > | > | 2568 make | > | > | | > | > | And this gives me the following build error: | > | > | | > | > | compiler/specialise/SpecConstr.lhs:56:7: | > | > | Could not find module `Data.Data': | > | > | Use -v to see a list of the files searched for. | > | > | <<ghc: 999358456 bytes, 177 GCs, 6793216/14557184 avg/max bytes | > | > | residency (4 samples), 74M in use, 0.01 INIT (0.00 elapsed), 1.47 MUT | > | > | (6.04 elapsed), 0.43 GC (0.45 elapsed) :ghc>> | > | > | make[1]: *** [compiler/stage1/build/.depend-v] Error 1 | > | > | make[1]: *** Deleting file `compiler/stage1/build/.depend-v' | > | > | make: *** [all] Error 2 | > | > | | > | > | This is a new error, which wasn't present last week. I'm on an older | > | > | Ubuntu box, so my bootstrap GHC is old. | > | > | $ ghc --version | > | > | The Glorious Glasgow Haskell Compilation System, version 6.8.2 | > | > | | > | > | Running the failing command with -v, I get the following list of | > | > | failed search locations: | > | > | compiler/specialise/SpecConstr.lhs:56:7: | > | > | Could not find module `Data.Data': | > | > | locations searched: | > | > | compiler/nativeGen/Data/Data.hs | > | > | compiler/nativeGen/Data/Data.lhs | > | > | compiler/basicTypes/Data/Data.hs | > | > | compiler/basicTypes/Data/Data.lhs | > | > | compiler/cmm/Data/Data.hs | > | > | compiler/cmm/Data/Data.lhs | > | > | compiler/codeGen/Data/Data.hs | > | > | compiler/codeGen/Data/Data.lhs | > | > | compiler/coreSyn/Data/Data.hs | > | > | compiler/coreSyn/Data/Data.lhs | > | > | compiler/cprAnalysis/Data/Data.hs | > | > | compiler/cprAnalysis/Data/Data.lhs | > | > | compiler/deSugar/Data/Data.hs | > | > | compiler/deSugar/Data/Data.lhs | > | > | compiler/ghci/Data/Data.hs | > | > | compiler/ghci/Data/Data.lhs | > | > | compiler/hsSyn/Data/Data.hs | > | > | compiler/hsSyn/Data/Data.lhs | > | > | compiler/iface/Data/Data.hs | > | > | compiler/iface/Data/Data.lhs | > | > | compiler/main/Data/Data.hs | > | > | compiler/main/Data/Data.lhs | > | > | compiler/parser/Data/Data.hs | > | > | compiler/parser/Data/Data.lhs | > | > | compiler/prelude/Data/Data.hs | > | > | compiler/prelude/Data/Data.lhs | > | > | compiler/profiling/Data/Data.hs | > | > | compiler/profiling/Data/Data.lhs | > | > | compiler/rename/Data/Data.hs | > | > | compiler/rename/Data/Data.lhs | > | > | compiler/simplCore/Data/Data.hs | > | > | compiler/simplCore/Data/Data.lhs | > | > | compiler/simplStg/Data/Data.hs | > | > | compiler/simplStg/Data/Data.lhs | > | > | compiler/specialise/Data/Data.hs | > | > | compiler/specialise/Data/Data.lhs | > | > | compiler/stgSyn/Data/Data.hs | > | > | compiler/stgSyn/Data/Data.lhs | > | > | compiler/stranal/Data/Data.hs | > | > | compiler/stranal/Data/Data.lhs | > | > | compiler/typecheck/Data/Data.hs | > | > | compiler/typecheck/Data/Data.lhs | > | > | compiler/types/Data/Data.hs | > | > | compiler/types/Data/Data.lhs | > | > | compiler/utils/Data/Data.hs | > | > | compiler/utils/Data/Data.lhs | > | > | compiler/vectorise/Data/Data.hs | > | > | compiler/vectorise/Data/Data.lhs | > | > | compiler/stage1/build/Data/Data.hs | > | > | compiler/stage1/build/Data/Data.lhs | > | > | compiler/stage1/build/autogen/Data/Data.hs | > | > | compiler/stage1/build/autogen/Data/Data.lhs | > | > | | > | > | | > | > | On Wed, Oct 21, 2009 at 1:30 AM, Simon Marlow <[email protected]> wrote: | > | > | > On 21/10/2009 09:26, Simon Peyton-Jones wrote: | > | > | >> | > | > | >> | Are the instructions on the wiki for nofib accurate? I've not been | > | > | >> | able to get it to compile, let alone run. I kinda feel like an idiot | > | > | >> | > | > | >> I think so. I do | > | > | >> cd nofib | > | > | >> make | > | > | > | > | > | > You need to do 'make boot' before 'make'. | > | > | > | > | > | > Cheers, | > | > | > Simon | > | > | > | > | > | > | > | > | >> | > | > | >> If that doesn't not work, send email to cvs-ghc | > | > | >> | > | > | >> Simon | > | > | >> | > | > | >> | :P | > | > | >> | | > | > | >> | On Tue, Oct 20, 2009 at 8:28 AM, Simon Peyton-Jones | > | > | >> |<[email protected]> wrote: | > | > | >> |> Crutcher | > | > | >> |> | > | > | >> |> | In the meantime, I'm looking at maybe this one: | > | > | >> |> | Int / Word / IntN / WordN are unequally optimized | > | > | >> |> | http://hackage.haskell.org/trac/ghc/ticket/3055 | > | > | >> |> | > | > | >> |> Great. There's bunch of other arithmetic-related bugs: | > | > | >> |> #1434 | > | > | >> |> #2271 | > | > | >> |> #2281 | > | > | >> |> #3070 | > | > | >> |> #2269 | > | > | >> |> | > | > | >> |> These would be good to look at as a group. Some are assigned to Don | > | > | >> Stewart, and | > | > | >> | he may be able to give you some useful context and perhaps some code, | > | > | >> but he's too | > | > | >> | busy to fix them. Likewise #2281 and Ian Lynagh | > | > | >> |> | > | > | >> |> Simon | > | > | >> |> | > | > | >> |> | -----Original Message----- | > | > | >> |> | From: Crutcher Dunnavant [mailto:[email protected]] | > | > | >> |> | Sent: 16 October 2009 21:31 | > | > | >> |> | To: Simon Peyton-Jones; Simon Marlow; [email protected] | > | > | >> |> | Subject: Looking at GHC bugs | > | > | >> |> | | > | > | >> |> | Hi. The last time I cycled through these, I spent some time looking | > | > | >> at | > | > | >> |> | #698. While that's certainly an interesting bug, I don't think I'm | > | > | >> up | > | > | >> |> | to speed on it yet, so I'm gonna shelve it until I understand more | > | > | >> |> | about the garbage collector. I'd like to actually do something | > | > | >> useful | > | > | >> |> | in the short term. | > | > | >> |> | | > | > | >> |> | So, I'm looking at the bug list yet again, but I'd really | appreciate | > | > | >> |> | being given a more focused task. You'd get some work out of me, and | > | > | >> |> | I'd learn something. | > | > | >> |> | | > | > | >> |> | In the meantime, I'm looking at maybe this one: | > | > | >> |> | Int / Word / IntN / WordN are unequally optimized | > | > | >> |> | http://hackage.haskell.org/trac/ghc/ticket/3055 | > | > | >> |> | > | > | >> |> | > | > | >> | > | > | > | > | > | > | > | > | > | > | > | > _______________________________________________ Cvs-ghc mailing list [email protected] http://www.haskell.org/mailman/listinfo/cvs-ghc
