March 2003 - Glasgow-haskell-users

RE: ANNOUNCE: GHC vesrion 5.04.3 released
by Simon Marlow 09 May '03

09 May '03

> Actually, I find it difficult to extract the sometimes drastic > differences between releases from that document (just one example: > does this release include last week's bugfixes in the Network > module?). How do I find out about the various bugfixes between > releases, to decide whether or not to upgrade (and whether or not > upgrading will help with specific problems)? > > After all, the reason for patchlevel releases is that you've fixed > some bugs, so why be so secretive about what these fixes are?-) The release notes doesn't contain bugfixes, and the reason we don't have a list of them is because I'm too lazy to keep track of what bug fixes have been merged in, sorry! The Network fix wasn't merged in (just too late). Ok, I'll take a look at the diffs and summarise the fixes... here goes: - Literal strings could be duplicated sometimes - fixes for deriving Eq/Enum on large datatypes in GHCi - -no-hs-main has a sensible meaning in --make mode now (it forces linking even if there's no Main) - Fix for avoiding recompilation when there are stub files around (didn't pick up the stubs properly before) - Fixes for FFI decls containing type variables - MacOS X fixes for GHCi - Filenames with spaces don't cause problems in GHCi now - GHCi doesn't get confused if your code installs a signal handler for ^C - -pgmL now works as advertised - specifying filenames without the trailing .hs/.lhs works again in --make and GHCi. - Fix for "class ops that do not introduce foralls" (a cause of crashes with certain kinds of class methods). - Fix a crash when a class op is used as a record selector - Fix for deriving on records with fields with names with leading underscores. - Fixes for passing/returning FunPtr in the FFI - hsc2hs works in a binary distribution - Fixes for crashes in the biographical and retainer profilers. - Some signals were inadvertently blocked by the RTS - Uninitialised variable in the storage manager causes crashes - Fix for hp2ps -c - hsc2hs should now work on Windows - Fix off-by-one in Data.PackedString.splitWithPS - uninitialised elements in UArrays are now filled with zero. - Integer divide by zero is now consistently a catchable exception (previously it could just result in SIGFPE). - Fix for terminal buffering on Solaris - System.Console.GetOpt fixes - Fix for converting Integer to Int64/Word64 - Network.accept doesn't fail if the host fails to reverse-lookup. Cheers, Simon

5 7

RE: Histogram-building code (was: Re: Yet another weakly defined bug report)
by Simon Marlow 08 May '03

08 May '03

> One question though, about hFlush. I print out the status by > repeatedly putStr'ing "blah blah \r". With NoBuffering set, it works, > but when following the putStr with 'hFlush stdout', it doesn't (only > outputs very sporadically. I guess I'm misunderstanding the function > of hFlush, anybody care to elaborate?) It could be a bug. Can you provide us with a cut-down example? Cheers, Simon

3 2

RE: Patches for NetBSD port and automatic .hc bootstrap (long)
by Simon Marlow 01 Apr '03

01 Apr '03

> below find a set of patches that hopefully should fix the following > for ghc-5.04.3: > > 1. (re-)port to NetBSD/i386. > - the last GHC to work for NetBSD was 4.04, I think. Since then, > both GHC and NetBSD has changed a lot... > > 2. restore the ability to do an automatic bootstrap from .hc files. > - at least the updated distrib/hc-build script runs to completion > without errors for me now on NetBSD/i386, and the resulting > compiler seems to work fine (at least it is able to compile > itself once again). All this on a system with no previous GHC > installed. > > I think most of the patches should be more or less self-explanatory. > But please feel free to ask if something looks strange. Thanks, I've committed most of your patches... > I will just mention one problem I had, that I think needs a cleaner > solution. This is marked "XXX_UB" in the patches below. > > The problem is the readline library in NetBSD. There is one readline > in the base system (/usr/include and /usr/lib) but that is not the GNU > readline that GHC expects. Instead the readline in /usr/pkg/include > and /usr/pkg/lib must be used. The GCC in NetBSD does not include the > /usr/pkg paths by default, so we must pass explicit -I/-L/-R all the > time when using it. In particular this: > > -L/usr/pkg/lib -Wl,-R/usr/pkg/lib > > needs to go into hslibs/util/util.conf.in. And the same needs to go > into ghc/rts/rts.conf.in for libgmp. I had to invent two new configure > variables to make this work, ExtraLdOptsReadline and ExtraLdOptsGmp. > Sofar this is OK I think, other systems should be able to use these > too if they have readline or gmp installed in non-standard places. > > However, the problem is how to set these variables in the configure > script. Below you can see the hack I did for NetBSD, but that is not > very elegant. Maybe a better solution would be to add > --with-readline=... and --with-gmp=... options to the configure > script? This would at least partly solve the problem. Yes, I think this would be a better solution - at least more consistent with the way configure handles other such things. The right thing to do is to add --with-readline-includes=... --with-readline-libraries=... and --with-gmp-includes=... --with-gmp-libraries=... Incidentally, why are the -Wl,-R<dir> options needed? I've never come across that before, usually the runtime linker just finds the libraries in the place they were originally linked from. > I will also update the NetBSD package (port) with these patches. I > think I will try to keep the package booting from .hc files rather > than going the "first download small binary bootstrap kit" route that > you guys seem to have chosen for FreeBSD nowadays. It seems to me that > although booting from .hc files is very fragile business it has the > advantage that it makes the package less sensitive to the exact OS > versions involved (in particular the version the binary bootstrap kit > was built on). Note that HC files can also be platform dependent, since they depend in some cases on the values of symbols from config.h. We used to try to keep .hc files completely platform-independent, but I've come around to the conclusion that this isn't a realistic goal. As soon as you use a platform-dependent #ifdef or #define in the Haskell source (and we do that a fair bit in the libraries and even the compiler), the property is lost. So, while one set of .hc files might work for all the versions of NetBSD you want to build the compiler on, we don't guarantee it. Cheers, Simon

3 2

ptrace for profiling?
by John Meacham 01 Apr '03

01 Apr '03

this may be a totally absurd request, but would it be possible to use ptrace or some other mechanism for poking a running program to cause a program to start generating profiling data or change what type of profiling data it is producing? I tend to view the profiling graph in real-time as I run my Haskell program interactively. every now and again I notice my program is behaving oddly but I am unable to reproduce the events that led to the odd behavior. It would be nice to be able to poke the program in such a way to cause it to start writing profiling frames to the stack for a bit. is this plausible? John -- --------------------------------------------------------------------------- John Meacham - California Institute of Technology, Alum. - john(a)foo.net ---------------------------------------------------------------------------

1 0

Patches for NetBSD port and automatic .hc bootstrap (long)
by Urban Boquist 31 Mar '03

31 Mar '03

Hi GHC maintainers, below find a set of patches that hopefully should fix the following for ghc-5.04.3: 1. (re-)port to NetBSD/i386. - the last GHC to work for NetBSD was 4.04, I think. Since then, both GHC and NetBSD has changed a lot... 2. restore the ability to do an automatic bootstrap from .hc files. - at least the updated distrib/hc-build script runs to completion without errors for me now on NetBSD/i386, and the resulting compiler seems to work fine (at least it is able to compile itself once again). All this on a system with no previous GHC installed. I think most of the patches should be more or less self-explanatory. But please feel free to ask if something looks strange. I will just mention one problem I had, that I think needs a cleaner solution. This is marked "XXX_UB" in the patches below. The problem is the readline library in NetBSD. There is one readline in the base system (/usr/include and /usr/lib) but that is not the GNU readline that GHC expects. Instead the readline in /usr/pkg/include and /usr/pkg/lib must be used. The GCC in NetBSD does not include the /usr/pkg paths by default, so we must pass explicit -I/-L/-R all the time when using it. In particular this: -L/usr/pkg/lib -Wl,-R/usr/pkg/lib needs to go into hslibs/util/util.conf.in. And the same needs to go into ghc/rts/rts.conf.in for libgmp. I had to invent two new configure variables to make this work, ExtraLdOptsReadline and ExtraLdOptsGmp. Sofar this is OK I think, other systems should be able to use these too if they have readline or gmp installed in non-standard places. However, the problem is how to set these variables in the configure script. Below you can see the hack I did for NetBSD, but that is not very elegant. Maybe a better solution would be to add --with-readline=... and --with-gmp=... options to the configure script? This would at least partly solve the problem. I will also update the NetBSD package (port) with these patches. I think I will try to keep the package booting from .hc files rather than going the "first download small binary bootstrap kit" route that you guys seem to have chosen for FreeBSD nowadays. It seems to me that although booting from .hc files is very fragile business it has the advantage that it makes the package less sensitive to the exact OS versions involved (in particular the version the binary bootstrap kit was built on). Kinds regards, -- Urban Boquist <boquist(a)crt.se> P.S. Please CC: any replies to me, I am not subscribed to this list. diff -bu -r --exclude=configure ghc-5.04.3.orig/Makefile ghc-5.04.3/Makefile --- ghc-5.04.3.orig/Makefile Wed Mar 5 11:11:34 2003 +++ ghc-5.04.3/Makefile Sat Mar 29 00:38:06 2003 @@ -295,11 +295,11 @@ $(RM) -r $(ProjectNameShort)-$(ProjectVersion) $(LN_S) . $(ProjectNameShort)-$(ProjectVersion) $(FIND) $(ProjectNameShort)-$(ProjectVersion)/ghc/compiler \ - $(ProjectNameShort)-$(ProjectVersion)/ghc/driver \ + $(ProjectNameShort)-$(ProjectVersion)/ghc/utils \ $(ProjectNameShort)-$(ProjectVersion)/libraries \ $(ProjectNameShort)-$(ProjectVersion)/hslibs \ $ -name "*.hc" -o -name "*_hsc.[ch]" -o -name "*_stub.[ch]" $ -print > hc-files-to-go - for f in `$(FIND) $(ProjectNameShort)-$(ProjectVersion)/ghc/compiler $(ProjectNameShort)-$(ProjectVersion)/ghc/driver $(ProjectNameShort)-$(ProjectVersion)/libraries $(ProjectNameShort)-$(ProjectVersion)/hslibs -name "*.hsc" -print` ""; do \ + for f in `$(FIND) $(ProjectNameShort)-$(ProjectVersion)/ghc/compiler $(ProjectNameShort)-$(ProjectVersion)/ghc/utils $(ProjectNameShort)-$(ProjectVersion)/libraries $(ProjectNameShort)-$(ProjectVersion)/hslibs -name "*.hsc" -print` ""; do \ if test "x$$f" != "x" && test -e `echo "$$f" | sed 's/hsc$$/hs/g'`; then \ echo `echo "$$f" | sed 's/hsc$$/hs/g' ` >> hc-files-to-go ; \ fi; \ diff -bu -r --exclude=configure ghc-5.04.3.orig/configure.in ghc-5.04.3/configure.in --- ghc-5.04.3.orig/configure.in Mon Feb 17 12:27:20 2003 +++ ghc-5.04.3/configure.in Sat Mar 29 00:38:06 2003 @@ -718,6 +718,13 @@ fi AC_SUBST(HaveReadlineHeaders) +dnl ** XXX_UB: fix this in a better way! +ReadlineIncludePath= +if test x"$TargetOS_CPP" = x"netbsd"; then + ReadlineIncludePath=/usr/pkg/include +fi +AC_SUBST(ReadlineIncludePath) + dnl ** check for DOS include files AC_CHECK_HEADERS(dos.h conio.h io.h std.h) @@ -734,14 +741,14 @@ AC_CHECK_HEADERS(dlfcn.h dl.h) dnl ** check for farcalloc (in bcc) -AC_CHECK_HEADER(alloc.h,AC_CHECK_FUNCS(farcalloc)) +AC_CHECK_HEADER(alloc.h,[AC_CHECK_FUNCS(farcalloc)]) dnl ** check for valloc (in sunos, solaris, mips, amiga, next, minix, ultrix) -AC_CHECK_HEADER(malloc.h,AC_CHECK_FUNCS(valloc)) +AC_CHECK_HEADER(malloc.h,[AC_CHECK_FUNCS(valloc)]) dnl ** check for POSIX regex HavePosixRegex=NO -AC_CHECK_HEADER(regex.h,AC_CHECK_FUNC(regcomp, HavePosixRegex=YES)) +AC_CHECK_HEADER(regex.h,[AC_CHECK_FUNC(regcomp, HavePosixRegex=YES)]) AC_SUBST(HavePosixRegex) dnl ** how do we get a timezone name, and UTC offset ? @@ -942,6 +949,13 @@ AC_SUBST(HaveLibGmp) AC_SUBST(LibGmp) +dnl ** XXX_UB: fix this in a better way! +ExtraLdOptsGmp= +if test x"$TargetOS_CPP" = x"netbsd"; then + ExtraLdOptsGmp="-L/usr/pkg/lib -Wl,-R/usr/pkg/lib" +fi +AC_SUBST(ExtraLdOptsGmp) + dnl ** (Mac OS X only: check for HaskellSupport.framework) HaveFrameworkHaskellSupport=NO if test $HostPlatform = "powerpc-apple-darwin"; then @@ -1012,6 +1026,13 @@ AC_DEFINE(HAVE_READLINE_4, 0) AC_DEFINE(HAVE_READLINE_4_2, 0) fi + +dnl ** XXX_UB: fix this in a better way! +ExtraLdOptsReadline= +if test x"$TargetOS_CPP" = x"netbsd"; then + ExtraLdOptsReadline="-L/usr/pkg/lib -Wl,-R/usr/pkg/lib" +fi +AC_SUBST(ExtraLdOptsReadline) dnl ** check for math library FPTOOLS_CHECK_LIBM() diff -bu -r --exclude=configure ghc-5.04.3.orig/distrib/hc-build ghc-5.04.3/distrib/hc-build --- ghc-5.04.3.orig/distrib/hc-build Tue Jan 15 06:39:15 2002 +++ ghc-5.04.3/distrib/hc-build Sun Mar 30 17:22:08 2003 @@ -2,6 +2,7 @@ # Manuel M. T. Chakravarty <chak(a)acm.org>, June 2000 # Updated for GHC 5.00, Simon Marlow, March 2001 +# Updated for GHC 5.04.3, Urban Boquist, March 2003 # # Script to build GHC from .hc files (must be run in the fptools/ root # directory into which the source and .hc files must already have been @@ -32,29 +33,37 @@ # touch happy generated files; so that in non-bootstrapping mode for # installation, no attempt is made to call happy # -touch ghc/compiler/rename/ParseIface.hs touch ghc/compiler/parser/Parser.hs touch ghc/compiler/main/ParsePkgConf.hs touch hslibs/hssource/HsParser.hs # We don't have genprimopcode yet so don't try to use it touch ghc/compiler/prelude/primops.txt -touch ghc/lib/std/PrelPrimopWrappers.hs +touch libraries/base/GHC/PrimopWrappers.hs -echo "*** Building hsc..." +echo "*** Building compiler..." ./configure --enable-hc-boot $configopts + +# A couple of Makefiles test BootingFromHc before including +# boilerplate.mk, so we have to pass it explicitly: +MAKEFLAGS="BootingFromHc=YES" +export MAKEFLAGS + $MAKE -C glafp-utils boot all $MAKE -C ghc boot +$MAKE -C libraries boot all $MAKE -C hslibs boot all $MAKE -C ghc all +MAKEFLAGS= + echo "*** Building libraries..." # Reconfigure, using the newly-build ghc binary as our $(GHC), and # with hc bootstrapping disabled. -HappyCmd="$PWD/distrib/fake-happy" ./configure --with-ghc="$PWD/ghc/compiler/ghc-inplace" +HappyCmd="$PWD/distrib/fake-happy" ./configure --with-ghc="$PWD/ghc/compiler/ghc-inplace" $configopts -PRIMOP_BITS=primop-data-decl.hs-incl \ +PRIMOP_BITS="primop-data-decl.hs-incl \ primop-tag.hs-incl \ primop-list.hs-incl \ primop-has-side-effects.hs-incl \ @@ -64,28 +73,30 @@ primop-can-fail.hs-incl \ primop-strictness.hs-incl \ primop-usage.hs-incl \ - primop-primop-info.hs-incl + primop-primop-info.hs-incl" # The reconfigure step updates a few files, which can lead to # unnecessary recompilations. Touch a bunch of things here to avoid # having to recompile stuff that we've already built. -(cd ghc/compiler; touch $PRIMOP_BITS prelude/PrimOp.o main/Config.hs main/Config.o ghc-*) +(cd ghc/compiler; touch $PRIMOP_BITS parser/hschooks.o prelude/PrimOp.o main/Config.hs main/Config.o ghc-*) # Remove the old libraries. Don't use make clean, because we don't # want to delete the .hs files generated from the .hsc files, because # we don't have hsc2hs built yet. -find ghc/lib/std hslibs | grep '\.$o\|a$' | xargs rm -f +find libraries hslibs | grep '\.$o\|a$' | xargs rm -f # Do includes and RTS now $MAKE -C ghc/includes boot && $MAKE -C ghc/includes all $MAKE -C ghc/rts boot && $MAKE -C ghc/rts all # Now build a new set of libraries -$MAKE -C ghc/lib boot all +$MAKE -C libraries boot all $MAKE -C hslibs boot all -# Finally build ghc/utils, now that we have libraries -$MAKE -C ghc/utils boot all +# Finally build all of ghc/utils +$MAKE -C ghc/utils clean && $MAKE -C ghc/utils boot all -# At this point, the tree should be safe to do 'make install' in. +# Avoid relinking the compiler during 'make install': +(cd ghc/compiler; touch parser/hschooks.o ghc-*) +# At this point, the tree should be safe to do 'make install' in. diff -bu -r --exclude=configure ghc-5.04.3.orig/ghc/Makefile ghc-5.04.3/ghc/Makefile --- ghc-5.04.3.orig/ghc/Makefile Tue Jul 2 16:15:22 2002 +++ ghc-5.04.3/ghc/Makefile Sat Mar 29 00:38:06 2003 @@ -18,10 +18,10 @@ # before the rest to have a config.h, etc. # # If we're booting from .hc files, swap the order -# we descend into compiler/ and lib/. +# we descend into subdirs - to boot utils must be before driver. # ifeq "$(BootingFromHc)" "YES" -SUBDIRS = includes utils rts docs compiler driver +SUBDIRS = includes rts docs compiler utils driver else ifneq "$(ILXized)" "YES" SUBDIRS = includes utils driver docs compiler rts diff -bu -r --exclude=configure ghc-5.04.3.orig/ghc/compiler/Makefile ghc-5.04.3/ghc/compiler/Makefile --- ghc-5.04.3.orig/ghc/compiler/Makefile Fri Jun 14 10:23:57 2002 +++ ghc-5.04.3/ghc/compiler/Makefile Sat Mar 29 00:38:06 2003 @@ -4,6 +4,7 @@ TOP = .. # Use GHC for compiling C bits (NB. must be before boilerplate include) +# NB. this requires BootingFromHc to be set on the make command line! # ifneq "$(BootingFromHc)" "YES" UseGhcForCc = YES diff -bu -r --exclude=configure ghc-5.04.3.orig/ghc/driver/mangler/ghc-asm.lprl ghc-5.04.3/ghc/driver/mangler/ghc-asm.lprl --- ghc-5.04.3.orig/ghc/driver/mangler/ghc-asm.lprl Sun Dec 8 15:29:50 2002 +++ ghc-5.04.3/ghc/driver/mangler/ghc-asm.lprl Sat Mar 29 00:38:06 2003 @@ -197,7 +197,7 @@ $T_hsc_cc_PAT = '\.string.*\)(hsc|cc) (.*)\\\\t(.*)"'; } - $T_DOT_WORD = '\.(long|value|byte|zero)'; + $T_DOT_WORD = '\.(long|value|word|byte|zero)'; $T_DOT_GLOBAL = '\.globl'; $T_HDR_literal = "\.section\t\.rodata\n"; # or just use .text??? (WDP 95/11) $T_HDR_misc = "\.text\n\t\.align 4\n"; diff -bu -r --exclude=configure ghc-5.04.3.orig/ghc/mk/paths.mk ghc-5.04.3/ghc/mk/paths.mk --- ghc-5.04.3.orig/ghc/mk/paths.mk Mon Apr 1 15:57:10 2002 +++ ghc-5.04.3/ghc/mk/paths.mk Sat Mar 29 00:38:06 2003 @@ -34,6 +34,7 @@ GHC_DRIVER_DIR = $(GHC_TOP)/$(GHC_DRIVER_DIR_REL) GHC_PKG_DIR = $(GHC_TOP)/$(GHC_PKG_DIR_REL) GHC_GENPRIMOP_DIR = $(GHC_TOP)/$(GHC_GENPRIMOP_DIR_REL) +GHC_MANGLER_DIR = $(GHC_TOP)/$(GHC_MANGLER_DIR_REL) GHC_LIB_DIR = $(FPTOOLS_TOP)/libraries diff -bu -r --exclude=configure ghc-5.04.3.orig/ghc/rts/Linker.c ghc-5.04.3/ghc/rts/Linker.c --- ghc-5.04.3.orig/ghc/rts/Linker.c Sun Oct 13 19:50:54 2002 +++ ghc-5.04.3/ghc/rts/Linker.c Sat Mar 29 00:38:06 2003 @@ -56,7 +56,7 @@ #include <sys/mman.h> #endif -#if defined(linux_TARGET_OS) || defined(solaris2_TARGET_OS) || defined(freebsd_TARGET_OS) +#if defined(linux_TARGET_OS) || defined(solaris2_TARGET_OS) || defined(freebsd_TARGET_OS) || defined(netbsd_TARGET_OS) # define OBJFORMAT_ELF #elif defined(cygwin32_TARGET_OS) || defined (mingw32_TARGET_OS) # define OBJFORMAT_PEi386 diff -bu -r --exclude=configure ghc-5.04.3.orig/ghc/rts/Makefile ghc-5.04.3/ghc/rts/Makefile --- ghc-5.04.3.orig/ghc/rts/Makefile Sun Oct 13 19:50:55 2002 +++ ghc-5.04.3/ghc/rts/Makefile Sat Mar 29 00:38:06 2003 @@ -21,6 +21,7 @@ # set of suffix rules for compiling C code, using $(HC) rather than $(CC) # and prepending "-optc" to $(CC_OPTS). NB. must be done before including # boilerplate.mk below. +# NB. this requires BootingFromHc to be set on the make command line! ifneq "$(BootingFromHc)" "YES" UseGhcForCc = YES endif @@ -126,6 +127,10 @@ ifeq "$(HaveLibMingwEx)" "YES" PACKAGE_CPP_OPTS += -DHAVE_LIBMINGWEX endif + +# have to get ExtraLdOptsGmp in through CPP to rts.conf.in (see hslibs/util/Makefile) +comma = , +PACKAGE_CPP_OPTS += -DExtraLdOptsGmp='$(patsubst %,"%"$(comma),$(ExtraLdOptsGmp))' #----------------------------------------------------------------------------- # Include the Front panel code? diff -bu -r --exclude=configure ghc-5.04.3.orig/ghc/rts/Stats.c ghc-5.04.3/ghc/rts/Stats.c --- ghc-5.04.3.orig/ghc/rts/Stats.c Wed Feb 6 02:21:40 2002 +++ ghc-5.04.3/ghc/rts/Stats.c Sat Mar 29 00:38:06 2003 @@ -749,7 +749,7 @@ fprintf(sf, "<<ghc: %llu bytes, ", GC_tot_alloc*sizeof(W_)); fprintf(sf, "%d GCs, %ld/%ld avg/max bytes residency (%ld samples), %luM in use, %.2f INIT (%.2f elapsed), %.2f MUT (%.2f elapsed), %.2f GC (%.2f elapsed) :ghc>>\n", total_collections, - AvgResidency*sizeof(W_)/ResidencySamples, + (ResidencySamples==0) ? 0 : AvgResidency*sizeof(W_)/ResidencySamples, MaxResidency*sizeof(W_), ResidencySamples, (unsigned long)(mblocks_allocated * MBLOCK_SIZE / (1024L * 1024L)), diff -bu -r --exclude=configure ghc-5.04.3.orig/ghc/rts/rts.conf.in ghc-5.04.3/ghc/rts/rts.conf.in --- ghc-5.04.3.orig/ghc/rts/rts.conf.in Mon Feb 10 11:18:31 2003 +++ ghc-5.04.3/ghc/rts/rts.conf.in Sat Mar 29 00:38:06 2003 @@ -67,6 +67,7 @@ so we force it to be included with special options to ld. */ extra_ld_opts = [ + ExtraLdOptsGmp #ifdef LEADING_UNDERSCORE "-u", "_GHCziBase_Izh_static_info" , "-u", "_GHCziBase_Czh_static_info" diff -bu -r --exclude=configure ghc-5.04.3.orig/ghc/utils/Makefile ghc-5.04.3/ghc/utils/Makefile --- ghc-5.04.3.orig/ghc/utils/Makefile Tue Sep 11 13:13:22 2001 +++ ghc-5.04.3/ghc/utils/Makefile Sat Mar 29 00:38:06 2003 @@ -6,7 +6,7 @@ SUBDIRS = hp2ps stat2resid unlit else ifeq "$(BootingFromHc)" "YES" -SUBDIRS = hp2ps parallel stat2resid prof unlit +SUBDIRS = ghc-pkg unlit else SUBDIRS = hasktags ghc-pkg hp2ps hsc2hs parallel stat2resid prof unlit genprimopcode endif diff -bu -r --exclude=configure ghc-5.04.3.orig/hslibs/util/Makefile ghc-5.04.3/hslibs/util/Makefile --- ghc-5.04.3.orig/hslibs/util/Makefile Tue Apr 23 11:31:39 2002 +++ ghc-5.04.3/hslibs/util/Makefile Sat Mar 29 00:38:06 2003 @@ -23,16 +23,25 @@ PACKAGE_DEPS := $(filter-out posix, $(PACKAGE_DEPS)) endif +# hsc2hs needs to find the correct readline includes: +EXTRA_HSC2HS_OPTS += -I$(ReadlineIncludePath) + # yeuch, have to get LibsReadline in through CPP to util.conf.in comma = , PACKAGE_CPP_OPTS += -DLibsReadline='$(patsubst %,$(comma)"%",$(LibsReadline))' +# ... same with ExtraLdOptsReadline: +PACKAGE_CPP_OPTS += -DExtraLdOptsReadline='$(patsubst %,"%"$(comma),$(ExtraLdOptsReadline))' + # Remove Readline.hs if it is not wanted or headers are not available. ifeq "$(GhcLibsWithReadline)" "YES" ifneq "$(ReadlineIncludePath)" "" SRC_HC_OPTS += -I$(ReadlineIncludePath) endif STUBOBJS += Readline_stub.$(way_)o + ifeq "$(BootingFromHc)" "YES" + all :: $(STUBOBJS) + endif CLEAN_FILES += Readline_stub.[ch] $(STUBOBJS) else EXCLUDED_SRCS += Readline.hsc Readline_stub.c diff -bu -r --exclude=configure ghc-5.04.3.orig/hslibs/util/util.conf.in ghc-5.04.3/hslibs/util/util.conf.in --- ghc-5.04.3.orig/hslibs/util/util.conf.in Sun Feb 17 16:22:11 2002 +++ ghc-5.04.3/hslibs/util/util.conf.in Sat Mar 29 00:38:06 2003 @@ -35,5 +35,8 @@ ], extra_ghc_opts = [], extra_cc_opts = [], - extra_ld_opts = [] + extra_ld_opts = [ + ExtraLdOptsReadline + "" /* handle final comma... */ + ] } diff -bu -r --exclude=configure ghc-5.04.3.orig/mk/bootstrap.mk ghc-5.04.3/mk/bootstrap.mk --- ghc-5.04.3.orig/mk/bootstrap.mk Tue May 14 10:25:46 2002 +++ ghc-5.04.3/mk/bootstrap.mk Sat Mar 29 00:38:06 2003 @@ -123,10 +123,22 @@ HC_BOOT_LIBS = -lHStext -lHStext_cbits -lHSutil -lHSposix -lHSposix_cbits -lHSconcurrent -lHSlang -lHSlang_cbits -lHShaskell98 -lHSbase -lHSbase_cbits -lHSrts -lgmp -lm $(EXTRA_HC_BOOT_LIBS) +# XXX_UB: if we build ghc/utils with hc-bootstrapped compiler+libs, +# the link fails, libHSrts uses a (new) symbol from libHSbase. +# Maybe we should insert a ranlib somewhere? +# Workaround the problem: +HC_BOOT_LIBS += -lHSbase + ifeq "$(GhcLibsWithReadline)" "YES" HC_BOOT_LIBS += $(patsubst %, -l%, $(LibsReadline)) +HC_BOOT_LIBS += $(ExtraLdOptsReadline) +ifneq "$(ReadlineIncludePath)" "" +HC_BOOT_CC_OPTS += -I$(ReadlineIncludePath) +endif endif +HC_BOOT_LIBS += $(ExtraLdOptsGmp) + ifeq "$(HaveLibDL)" "YES" HC_BOOT_LIBS += -ldl endif @@ -149,7 +161,7 @@ $(CC) -x c $< -o $@ -S -O $(HC_BOOT_CC_OPTS) -I. `echo $(patsubst -monly-%-regs, -DSTOLEN_X86_REGS=%, $(filter -monly-%-regs, $($*_HC_OPTS))) | sed 's/^$$/-DSTOLEN_X86_REGS=4/'` %.s : %.raw_s - $(FPTOOLS_TOP)/$(GHC_MANGLER_DIR)/$(GHC_MANGLER) $< $@ $(patsubst -monly-%-regs, %, $(filter -monly-%-regs, $($*_HC_OPTS))) + $(GHC_MANGLER) $< $@ $(patsubst -monly-%-regs, %, $(filter -monly-%-regs, $($*_HC_OPTS))) %.o : %.s $(CC) -c -o $@ $< diff -bu -r --exclude=configure ghc-5.04.3.orig/mk/config.mk.in ghc-5.04.3/mk/config.mk.in --- ghc-5.04.3.orig/mk/config.mk.in Wed Feb 12 22:41:18 2003 +++ ghc-5.04.3/mk/config.mk.in Sat Mar 29 00:38:06 2003 @@ -338,10 +338,13 @@ # Libraries needed for linking with readline LibsReadline=@LibsReadline@ +# Additional linker flags needed for linking with readline +ExtraLdOptsReadline=@ExtraLdOptsReadline@ + # Include path to readline.h # (no path == in standard include path) # -ReadlineIncludePath= +ReadlineIncludePath=@ReadlineIncludePath@ # Math library LIBM=@LIBM@ @@ -678,6 +681,9 @@ # HaveLibGmp = @HaveLibGmp@ LibGmp = @LibGmp@ + +# Additional linker flags needed for linking with gmp (when HaveLibGmp=YES) +ExtraLdOptsGmp=@ExtraLdOptsGmp@ #----------------------------------------------------------------------------- # Mingwex Library diff -bu -r --exclude=configure ghc-5.04.3.orig/mk/package.mk ghc-5.04.3/mk/package.mk --- ghc-5.04.3.orig/mk/package.mk Tue Jul 23 12:19:50 2002 +++ ghc-5.04.3/mk/package.mk Sat Mar 29 00:38:06 2003 @@ -17,8 +17,10 @@ | sed 's/^#.*$$//g' >$@ boot all :: $(PACKAGE).conf.inplace $(PACKAGE).conf.installed +ifneq "$(BootingFromHc)" "YES" $(GHC_PKG_INPLACE) --update-package <$(PACKAGE).conf.inplace $(GHC_PKG_INPLACE) -f $(GHC_DRIVER_DIR)/package.conf --update-package <$(PACKAGE).conf.installed +endif CLEAN_FILES += $(PACKAGE).conf.installed $(PACKAGE).conf.inplace @@ -31,7 +33,7 @@ SRC_HSC2HS_OPTS += -I. -ifeq "$(NON_HS_PACKAGE)" "" +ifeq "$(NON_HS_PKG)" "" SRC_HC_OPTS += -package-name $(PACKAGE) SRC_HC_OPTS += $(GhcLibHcOpts) SRC_HC_OPTS += $(patsubst %, -package %, $(PACKAGE_DEPS))

1 0

RE: profiling question
by Simon Marlow 28 Mar '03

28 Mar '03

> I think this question was asked by someone else as part of another > thread back in January, but I couldn't find an answer on the list > archive. What do the zeros for "entries" mean in a time profiling > report? If I want to know why "getAppropriatePreds" takes up so much > time (14.7%) in my code, does the zero mean an overflowed > count or that > the cost centre has never been entered? I'm using 5.04.2 on Linux, and > my program is compiled with -O -prof -auto-all. > > compose InfrequentCheck 1137 12280580 2.1 2.5 > 2.1 2.5 > align InfrequentCheck 1127 0 3.7 > 6.2 8.3 > 6.7 > align2 InfrequentCheck 1139 8577943 0.4 0.0 > 0.4 0.0 > lookupModeArgs InfrequentCheck 1136 86747940 4.2 0.4 > 4.2 0.4 > getAppropriatePreds InfrequentCheck 1126 0 14.7 6.6 > 14.7 6.6 I think the best answer to give here is that the entries count is not at all reliable. Please feel free to file a bug, we'll get around to it next time I have the profiling system paged in ;-) Cheers, SImon

1 0

H/Direct CVS snapshot needed
by Лукин Константин Валерьевич 27 Mar '03

27 Mar '03

Hi, How can I obtain H/Direct CVS snapshot? I can't checkout H/Direct due to corporate firewall. Kostya Lukin

1 0

irg.ssu.ac.kr
by TrafficMagnet Reseller 27 Mar '03

27 Mar '03

Become a TrafficMagnet Reseller Today! ------------------------------------------------------------------------ It's Free! Click at http://p1j2m3a4.pdhost.com/pdsvr/www/r?1000032249.1871.9.zE6XlichEfakn6 to Sign Up now! ------------------------------------------------------------------------ - Place any of these banners on your website, and you will earn $10 every time you refer a sale to TrafficMagnet! - Start making more money from your web space by reselling TrafficMagnet to your own customers and site visitors. - Get the HTML code for our banners and text links by signing up as a Reseller today! Visit http://www.trafficmagnet.com to find out more! Not interested in becoming a CoolStats Reseller? To be taken off our mailing list, please follow the instructions here: http://p1j2m3a4.pdhost.com/pdsvr/www/optoutredirect?UC=Lead&UI=14498163

1 0

profiling question
by Amanda Clare 27 Mar '03

27 Mar '03

I think this question was asked by someone else as part of another thread back in January, but I couldn't find an answer on the list archive. What do the zeros for "entries" mean in a time profiling report? If I want to know why "getAppropriatePreds" takes up so much time (14.7%) in my code, does the zero mean an overflowed count or that the cost centre has never been entered? I'm using 5.04.2 on Linux, and my program is compiled with -O -prof -auto-all. compose InfrequentCheck 1137 12280580 2.1 2.5 2.1 2.5 align InfrequentCheck 1127 0 3.7 6.2 8.3 6.7 align2 InfrequentCheck 1139 8577943 0.4 0.0 0.4 0.0 lookupModeArgs InfrequentCheck 1136 86747940 4.2 0.4 4.2 0.4 getAppropriatePreds InfrequentCheck 1126 0 14.7 6.6 14.7 6.6 Amanda

1 0

RE: Fix in CVS ndeed Posix compile problem
by Simon Marlow 26 Mar '03

26 Mar '03

> SunOS 5.8 > Reading specs from > /usr/local/lib/gcc-lib/sparc-sun-solaris2.8/2.95.3/specs > gcc version 2.95.3 20010315 (release) > In the posix related libaries there are two forign call definition. > > foreign import ccall unsafe "getpwuid_r" > c_getpwuid_r :: CUid -> Ptr CPasswd -> > CString -> CSize -> Ptr (Ptr CPasswd) -> IO CInt > > foreign import ccall unsafe "getpwnam_r" > c_getpwnam_r :: CString -> Ptr CPasswd -> > CString -> CSize -> Ptr (Ptr CPasswd) -> IO CInt > > But the gcc says there are too many argument. > There are two ways to go around. > I change the foriegn call definition "getpwuid_r" ==> > "__posix_getpwuid_r" > I change the foriegn call definition "getpwnam_r" ==> > "__posix_getpwnam_r" Thanks, I'll look into this. Cheers, Simon

2 1