March 2009 - Glasgow-haskell-users

Solaris 8 and libm.so.2
by Ralph Crawford 17 Mar '09

17 Mar '09

Hi everyone. I'm trying to build GHC 6.8.3 on a Solaris 8 machine. I'd love to upgrade Solaris to 10, which I believe would solve the problem I'm having by making /usr/lib/libm.so.2 available, but that's not an option at this time. Here's the output of uname -a on the system. SunOS bwddev1 5.8 Generic_117350-53 sun4u sparc SUNW,Ultra-80 The GCC I'm using is gcc Target: sparc-sun-solaris2.8 Configured with: /usr/local/src/gnu/gcc-4.0.1/configure --with-as=/usr/ccs/bin/as --with-ld=/usr/ccs/bin/ld --disable-nls --prefix=/usr/local/gcc4 Thread model: posix gcc version 4.0.1 The GHC binary I'm bootstrapping with is 6.4.1. It works fine for compiling complex haskell programs and packages. I've already worked through a couple of problems to get to where I'm at now. What I had to do was... changed #define SIZET_FMT "d" to #define SIZET_FMT "d" in includes/mkDerivedConstances.c. change #include <stdint.h> to #include <inttypes.h> in includes/HsFFI.h added mk/build.mk containing SRC_HC_OPTS=-opta-mcpu=v9 My latest attempt at a build involved the following, where BOS_ROOT points at the root of my build tree, and $BOS_ROOT/gcc4 is where my gcc lives, and $BOS_ROOT/ghc is where the ghc binary lives. This is only the latest attempt. The libm.so.2 symbolic link and the LDFLAGS and LD_LIBRARY_PATH values were added on the most recent attempt. ln -s /usr/lib/libm.so.1 $BOS_ROOT/lib/libm.so.2 PATH=$BOS_ROOT/gcc4/bin:$BOS_ROOT/ghc/bin:$BOS_ROOT/bin:/usr/bin:/bin LD_LIBRARY_PATH=$BOS_ROOT/gcc4/lib:$BOS_ROOT/ghc/lib:$BOS_ROOT/lib:/usr/lib LDFLAGS=-static -L$BOS_ROOT/lib CFLAGS= export PATH LD_LIBRARY_PATH LDFLAGS CFLAGS fp_prog_ld_raw=/usr/ccs/bin/ld ./configure \ --prefix=$BOS_ROOT \ --with-ghc=$BOS_ROOT/ghc/bin/ghc \ --with-as=/usr/ccs/bin/as \ --with-ld=/usr/ccs/bin/ld \ --with-gcc=$BOS_ROOT/gcc4/bin/gcc \ --with-gmp-includes=$BOS_ROOT/include \ --with-gmp-libraries=$BOS_ROOT/lib \ --with-readline-includes=$BOS_ROOT/include \ --with-readline-libraries=$BOS_ROOT/lib Once I made these changes, the build goes on for a number of hours then dies with what's listed below. I understand that this is because the build is looking for libm.so.2, while the lib on Solaris 8 is /usr/lib/libm.so.1. No matter what I do - including the kludge above attempting to create a libm.so.2 symbolic link, I can't get rid of the errors. Any help at all would be very much appreciated. Please. Here's the error listing. make[3]: Entering directory `/winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/compiler' ../compiler/stage1/ghc-inplace -no-user-package-conf -opta-mcpu=v9 -package ghc -Istage2 -cpp -fglasgow-exts -fno-generics -Rghc-timing -I. -IcodeGen -InativeGen -Iparser -Rghc-timing -DGHCI -threaded -fforce-recomp -c main/Main.hs -o stage2/main/Main.o -ohi stage2/main/Main.hi <<ghc: 450131408 bytes, 1127 GCs, 6839050/13262620 avg/max bytes residency (8 samples), 35M in use, 0.00 INIT (0.01 elapsed), 13.34 MUT (40.52 elapsed), 8.42 GC (8.81 elapsed) :ghc>> ../compiler/stage1/ghc-inplace -no-user-package-conf -o stage2/ghc-6.8.3 -opta-mcpu=v9 -package ghc -Istage2 -cpp -fglasgow-exts -fno-generics -Rghc-timing -I. -IcodeGen -InativeGen -Iparser -Rghc-timing -DGHCI -threaded -fforce-recomp stage2/main/Main.o Undefined first referenced symbol in file cosf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__169.o) expf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__173.o) logf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__172.o) powf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__161.o) sinf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__170.o) tanf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__168.o) acosf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__166.o) asinf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__167.o) atanf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__165.o) coshf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__163.o) dlsym /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/rts/libHSrts_thr.a(Linker.thr_o) (symbol belongs to implicit dependency /usr/lib/libdl.so.1) sinhf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__164.o) tanhf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__162.o) sqrtf /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/libraries/base/dist/build/libHSbase-3.0.2.0.a(Float__171.o) dlopen /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/rts/libHSrts_thr.a(Linker.thr_o) (symbol belongs to implicit dependency /usr/lib/libdl.so.1) dlerror /winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/rts/libHSrts_thr.a(Linker.thr_o) (symbol belongs to implicit dependency /usr/lib/libdl.so.1) ld: fatal: Symbol referencing errors. No output written to stage2/ghc-6.8.3 collect2: ld returned 1 exit status <<ghc: 57428604 bytes, 11 GCs, 126372/126372 avg/max bytes residency (1 samples), 7M in use, 0.00 INIT (0.00 elapsed), 1.49 MUT (17.35 elapsed), 0.09 GC (0.10 elapsed) :ghc>> make[3]: *** [stage2/ghc-6.8.3] Error 1 make[3]: Leaving directory `/winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/compiler' make[2]: *** [stage2/ghc-6.8.3] Error 2 make[2]: Leaving directory `/winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3/compiler' make[1]: *** [stage2] Error 2 make[1]: Leaving directory `/winchester/bwd_home/rc4498/cvs_sandboxes/wip/bos-devel/bos/src/ghc-6.8.3' make: *** [bootstrap2] Error 2 -- --------------------------------------------------------------- My web host for www.ralphcrawford.org is... http://www.1and1.com/?k_id=6466911 They're good - I recommend them. I'm switching my email from stryder22204(a)yahoo.com to stryder100(a)gmail.com. Don't worry though I'll still get stuff from the stryder22204 address for about 6 months.

3 2

testing 6.10.2-candidate
by Serge D. Mechveliani 17 Mar '09

17 Mar '09

I have tested ghc-6.10.1.20090314 on Debian Linux, i386-unknown, on making from source by ghc-6.10.1, making itself from source, DoCon. It looks all right. ----------------- Serge Mechveliani mechvel(a)botik.ru

2 1

Developer Wiki & Buildbots
by Philip K.F. Hölzenspies 17 Mar '09

17 Mar '09

Dear all, It seems that the front page of the developer wiki is rather out-of-date. Considering that 6.10.2 is now at rc1-stage, I was rather hoping to find some updated notes on release plans for 6.10.2 and what will be in 6.12, but those notices are still at 6.8.3 and 6.10, resp. On another matter: I've been trying to contact Ian on rejoining the buildbot network, but haven't heard from him in response to any of messages. Ian, I figured I may have - somehow - ended up in your spam-box, so I thought it a good idea to go through the mailing list. Could you maybe check your spam-box for my e-mails? Regards, Philip

2 1

ghci finding include files exported from other packages?
by Conal Elliott 17 Mar '09

17 Mar '09

The applicative-numbers package [1] provides an include file. With ghci, the include file isn't being found, though with cabal+ghc it is found. My test source is just two lines: {-# LANGUAGE CPP #-} #include "ApplicativeNumeric-inc.hs" I'd sure appreciate it if someone could take a look at the .cabal file [2] and tell me if I'm doing something wrong. And/or point me to one or more working examples of cabal packages that export include files that are then findable via ghci. - Conal [1] http://hackage.haskell.org/packages/archive/applicative-numbers [2] http://hackage.haskell.org/packages/archive/applicative-numbers/0.0.2/appli…

3 10

Link errors
by Lingappan, Loganathan 15 Mar '09

15 Mar '09

Hi, If I include import Text.Regex.Posix ((=~)) into a Haskell code, I get the following link error: FindBBUsage.o:fake:(.text+0x44d): undefined reference to `__stginit_regexzmposixzm0zi72zi0zi3_TextziRegexziPosix_' collect2: ld returned 1 exit status Any ideas on how to fix this? I am using GHC version 6.10.1 on Windows XP. Thanks, Logo

2 2

Bug in deriving Data Typeable?
by Bernd Brassel 13 Mar '09

13 Mar '09

Hi folks, I have been surprised by the derived instances of Data/Typeable when using the combinator ext1Q. First look at the following definition: > useExt1 :: Data a => a -> () > useExt1 = undefined `ext1Q` (\ (Just _) -> ()) > testExt1 :: () > testExt1 = useExt1 (Just ()) As I expected, testExt1 yields () But when I define my own version of Maybe and derive the Data/Typeable instances > data MyMaybe a = MyJust a | MyNothing deriving (Data,Typeable) the corresponding test > useExt1' :: Data a => a -> () > useExt1' = undefined `ext1Q` (\ (MyJust _) -> ()) > testExt1' :: () > testExt1' = useExt1' (MyJust ()) yields *** Exception: Prelude.undefined All of this happens with both ghc 6.10.1 and 6.8.2. Is this a bug in the derived instances, e.g., in dataCast1? If so is there a concise tutorial telling me how to write correct instances for the data type I'm interested in (which is not MyMaybe)? Thanks for your time! Bernd

2 3

ThreadScope: Request for features for the performance tuning of parallel and concurrent Haskell programs
by Satnam Singh 12 Mar '09

12 Mar '09

Donnie Jones, Simon Marlow and I have been working on infrastructure for logging run-time events and a graphical viewer program called ThreadScope. Hopefully these features will make it into the next release of GHC. We hope the event-log viewer ThreadScope will be useful for the performance tuning of parallel and concurrent Haskell programs. You can see a few screen shots at the program's website http://raintown.org/threadscope Before making the release I thought it would be an idea to ask people what other features people would find useful or performance tuning. So if you have any suggestions please do let us know! Cheers, Satnam Singh ________________________________ Satnam Singh Microsoft 7 JJ Thomson Avenue Cambridge CB3 0FB United Kingdom Email: satnams(a)microsoft.com<mailto:satnams@microsoft.com> UK tel: +44 1223 479905 Fax: +44 1223 479 999 UK mobile: +44 7979 648412 USA cell: 206 330 1580 USA tel: 206 219 9024 URL: http://research.microsoft.com/~satnams Live Messenger: satnam(a)raintown.org<mailto:satnam@raintown.org>

5 5

Cygwin version
by Tuomo Valkonen 11 Mar '09

11 Mar '09

I want a _real_ cygwin version of darcs. The non-deterministic pseudo-cygwin *nix/Windows hybrid currently available has just too many problems integrating into cygwin, that I want to use as my TeXing and minor coding environment. A real cygwin version of darcs would seem to depend on a real cygwin version of GHC. Is there any easy way to compile one? Otherwise I may have to abandon darcs (and Haskell software in general) for Mercurial. (Thanks to the over-bearing cabal and resulting hsc2hs etc. build problems with conventional Makefiles, I have already pretty much already abandoned my own Haskell projects.) -- Stop Gnomes and other pests! Purchase Windows today!

6 15

a possibility to redefine built-in GHCi commands
by Peter Hercek 11 Mar '09

11 Mar '09

Hi GHCi users, I would like to be able to redefine the built-in GHCi commands. The idea is that when searching for a command the user defined commands would be searched first and only then the built-in commands would be searched. If user wants to invoke a built-in command regardless of user defined commands he/she would need to start it with two colons (instead of one). It is an user interface change which may break some scripts, but it would allow to provide different default behavior. For example: * when I use GhciExt I want all my ":continue" commands to be actually ":x :continue" * it would allow to specify different order of searching for abbreviated commands as the default one * it would allow to specify different default switches for builtin commands Would such a change be merged upstream if I would provide a patch? Peter.

2 2

Reliability of trace?
by Colin Paul Adams 10 Mar '09

10 Mar '09

I'm trying to use Debug.Trace to debug some tree-walking that I've written. It seems to me that I am missing some traces on intermediate function calls. I guess that ghc is re-arranging the code in such a way that some of these intermediate calls disappear. Anyway of stopping this? I already specify -O0. -- Colin Adams Preston Lancashire

2 2