- ghc-devs - Haskell.org

RE: [commit: ghc] master: Fix #8754 in a round-about way. (5023c91)
by Simon Peyton Jones 17 Feb '14

17 Feb '14

Austin, eek... you really need a Note [Need to initialise the hooks], a significant comment, and an explicit mention of this ticket. Otherwise someone in 2 yrs time will have no clue why that lonely call to defaultsHook is so important, nor will be confident enough to remove it when the hookery is done better. I have gotten into habit of writing a Note about virtually any fix. After all, it is by definition not obvious or we'd have got it right the first time. And I have too often re-invented the wheel when looking at Note-free code; or thanked the author when I read a Note and thought "ah, *that's* why!". I wrote up some of this here https://ghc.haskell.org/trac/ghc/wiki/Commentary/CodingStyle I don't think the commit log is enough. git annotate may or may not lead you to the right comment, depending on how heavily modified that code is. Thanks Simon | -----Original Message----- | From: ghc-commits [mailto:ghc-commits-bounces@haskell.org] On Behalf Of | git(a)git.haskell.org | Sent: 17 February 2014 07:49 | To: ghc-commits(a)haskell.org | Subject: [commit: ghc] master: Fix #8754 in a round-about way. (5023c91) | | Repository : ssh://git@git.haskell.org/ghc | | On branch : master | Link : | http://ghc.haskell.org/trac/ghc/changeset/5023c91780e90947680fe0640f7564 | a4f6448bea/ghc | | >--------------------------------------------------------------- | | commit 5023c91780e90947680fe0640f7564a4f6448bea | Author: Austin Seipp <austin(a)well-typed.com> | Date: Sun Feb 16 19:10:16 2014 -0600 | | Fix #8754 in a round-about way. | | For some reason on OS X, it seems like -Bsymbolic (which we use for | hooks into the RTS) isn't working, which results in #8754, where | stats | don't work because defaultHooks doesn't initialize the stats flag. | This | seems to work on Linux static/dynamically, but only on OS X | statically. | | After talking with Simon, really, the entire hooks thing is a bit | fragile. For now, we just work around it (since GHCi is dynamically | linked) by calling into the defaultHooks ourselves when GHC starts. | | Signed-off-by: Austin Seipp <austin(a)well-typed.com> | | | >--------------------------------------------------------------- | | 5023c91780e90947680fe0640f7564a4f6448bea | ghc/Main.hs | 4 ++++ | 1 file changed, 4 insertions(+) | | diff --git a/ghc/Main.hs b/ghc/Main.hs | index 868042b..1aa6553 100644 | --- a/ghc/Main.hs | +++ b/ghc/Main.hs | @@ -1,4 +1,5 @@ | {-# OPTIONS -fno-warn-incomplete-patterns -optc-DNON_POSIX_SOURCE #-} | +{-# LANGUAGE ForeignFunctionInterface #-} | | ----------------------------------------------------------------------- | ------ | -- | @@ -76,6 +77,7 @@ import Data.Maybe | | main :: IO () | main = do | + defaultsHook | hSetBuffering stdout LineBuffering | hSetBuffering stderr LineBuffering | GHC.defaultErrorHandler defaultFatalMessager defaultFlushOut $ do @@ | -818,3 +820,5 @@ unknownFlagsErr fs = throwGhcException $ UsageError $ | concatMap oneError fs | (case fuzzyMatch f (nub allFlags) of | [] -> "" | suggs -> "did you mean one of:\n" ++ unlines (map (" " ++) | suggs)) | + | +foreign import ccall safe "defaultsHook" defaultsHook :: IO () | | _______________________________________________ | ghc-commits mailing list | ghc-commits(a)haskell.org | http://www.haskell.org/mailman/listinfo/ghc-commits

2 1

Re: [commit: packages/base] master: Implement foldl with foldr (b63face)
by Simon Marlow 17 Feb '14

17 Feb '14

This worries me a bit. If foldl isn't inlined, I get a less efficient version, so it has to be inlined everywhere. So -O0 code gets worse, and binary sizes for -O1+ get bigger - foldl, sum, and product are now INLINE. What I'm arguing is that we should have more flexibility to *not* inline things (INLINABLE is much better than INLINE), and when not inlining things we should be calling an efficient version of the function. This is why map is not defined in terms of foldr, for instance. Cheers, Simon On 10/02/2014 13:52, git(a)git.haskell.org wrote: > Repository : ssh://git@git.haskell.org/base > > On branch : master > Link : http://ghc.haskell.org/trac/ghc/changeset/b63facef165b957183b65604ef99b2b85… > >> --------------------------------------------------------------- > > commit b63facef165b957183b65604ef99b2b8574747a5 > Author: Joachim Breitner <mail(a)joachim-breitner.de> > Date: Tue Jan 28 14:31:05 2014 +0100 > > Implement foldl with foldr > > together with the call arity analysis and the following patch (about inlining > maximum), we get nice benefits from fusing foldl and foldl' with good > producers: > > Min -0.1% -74.5% -6.8% -8.3% -50.0% > Max +0.2% 0.0% +38.5% +38.5% 0.0% > Geometric Mean -0.0% -4.1% +7.7% +7.7% -0.8% > > Because this depends on a compiler optimisation, we have to watch out for cases > where this is not an improvements, and whether they occur in the wild. > > >> --------------------------------------------------------------- > > b63facef165b957183b65604ef99b2b8574747a5 > Data/List.hs | 34 +++++++++------------------------- > GHC/List.lhs | 13 +++++++------ > 2 files changed, 16 insertions(+), 31 deletions(-) > > diff --git a/Data/List.hs b/Data/List.hs > index 130ceb2..4796055 100644 > --- a/Data/List.hs > +++ b/Data/List.hs > @@ -1,5 +1,5 @@ > {-# LANGUAGE Trustworthy #-} > -{-# LANGUAGE CPP, NoImplicitPrelude, MagicHash #-} > +{-# LANGUAGE CPP, NoImplicitPrelude, ScopedTypeVariables, MagicHash #-} > > ----------------------------------------------------------------------------- > -- | > @@ -989,10 +989,11 @@ unfoldr f b = > -- ----------------------------------------------------------------------------- > > -- | A strict version of 'foldl'. > -foldl' :: (b -> a -> b) -> b -> [a] -> b > -foldl' f z0 xs0 = lgo z0 xs0 > - where lgo z [] = z > - lgo z (x:xs) = let z' = f z x in z' `seq` lgo z' xs > +foldl' :: forall a b . (b -> a -> b) -> b -> [a] -> b > +foldl' k z0 xs = foldr (\(v::a) (fn::b->b) (z::b) -> z `seq` fn (k z v)) (id :: b -> b) xs z0 > +-- Implementing foldl' via foldr is only a good idea if the compiler can optimize > +-- the resulting code (eta-expand the recursive "go"), so this needs -fcall-arity! > +-- Also see #7994 > > -- | 'foldl1' is a variant of 'foldl' that has no starting value argument, > -- and thus must be applied to non-empty lists. > @@ -1008,32 +1009,15 @@ foldl1' _ [] = errorEmptyList "foldl1'" > -- ----------------------------------------------------------------------------- > -- List sum and product > > -{-# SPECIALISE sum :: [Int] -> Int #-} > -{-# SPECIALISE sum :: [Integer] -> Integer #-} > -{-# INLINABLE sum #-} > -{-# SPECIALISE product :: [Int] -> Int #-} > -{-# SPECIALISE product :: [Integer] -> Integer #-} > -{-# INLINABLE product #-} > --- We make 'sum' and 'product' inlinable so that we get specialisations > --- at other types. See, for example, Trac #7507. > - > -- | The 'sum' function computes the sum of a finite list of numbers. > sum :: (Num a) => [a] -> a > -- | The 'product' function computes the product of a finite list of numbers. > product :: (Num a) => [a] -> a > -#ifdef USE_REPORT_PRELUDE > + > +{-# INLINE sum #-} > sum = foldl (+) 0 > +{-# INLINE product #-} > product = foldl (*) 1 > -#else > -sum l = sum' l 0 > - where > - sum' [] a = a > - sum' (x:xs) a = sum' xs (a+x) > -product l = prod l 1 > - where > - prod [] a = a > - prod (x:xs) a = prod xs (a*x) > -#endif > > -- ----------------------------------------------------------------------------- > -- Functions on strings > diff --git a/GHC/List.lhs b/GHC/List.lhs > index b7b78c7..e004ded 100644 > --- a/GHC/List.lhs > +++ b/GHC/List.lhs > @@ -1,6 +1,6 @@ > \begin{code} > {-# LANGUAGE Trustworthy #-} > -{-# LANGUAGE CPP, NoImplicitPrelude, MagicHash #-} > +{-# LANGUAGE CPP, NoImplicitPrelude, ScopedTypeVariables, MagicHash #-} > {-# OPTIONS_HADDOCK hide #-} > > ----------------------------------------------------------------------------- > @@ -178,11 +178,12 @@ filterFB c p x r | p x = x `c` r > -- can be inlined, and then (often) strictness-analysed, > -- and hence the classic space leak on foldl (+) 0 xs > > -foldl :: (b -> a -> b) -> b -> [a] -> b > -foldl f z0 xs0 = lgo z0 xs0 > - where > - lgo z [] = z > - lgo z (x:xs) = lgo (f z x) xs > +foldl :: forall a b. (b -> a -> b) -> b -> [a] -> b > +{-# INLINE foldl #-} > +foldl k z0 xs = foldr (\(v::a) (fn::b->b) (z::b) -> fn (k z v)) (id :: b -> b) xs z0 > +-- Implementing foldl via foldr is only a good idea if the compiler can optimize > +-- the resulting code (eta-expand the recursive "go"), so this needs -fcall-arity! > +-- Also see #7994 > > -- | 'scanl' is similar to 'foldl', but returns a list of successive > -- reduced values from the left: > > _______________________________________________ > ghc-commits mailing list > ghc-commits(a)haskell.org > http://www.haskell.org/mailman/listinfo/ghc-commits >

3 4

Re: [commit: packages/base] master: Implement foldl with foldr (b63face)
by Simon Marlow 17 Feb '14

17 Feb '14

This worries me a bit. If foldl isn't inlined, I get a less efficient version, so it has to be inlined everywhere. So -O0 code gets worse, and binary sizes for -O1+ get bigger - foldl, sum, and product are now INLINE. What I'm arguing is that we should have more flexibility to *not* inline things (INLINABLE is much better than INLINE), and when not inlining things we should be calling an efficient version of the function. This is why map is not defined in terms of foldr, for instance. Cheers, Simon On 10/02/2014 13:52, git(a)git.haskell.org wrote: > Repository : ssh://git@git.haskell.org/base > > On branch : master > Link : http://ghc.haskell.org/trac/ghc/changeset/b63facef165b957183b65604ef99b2b85… > >> --------------------------------------------------------------- > > commit b63facef165b957183b65604ef99b2b8574747a5 > Author: Joachim Breitner <mail(a)joachim-breitner.de> > Date: Tue Jan 28 14:31:05 2014 +0100 > > Implement foldl with foldr > > together with the call arity analysis and the following patch (about inlining > maximum), we get nice benefits from fusing foldl and foldl' with good > producers: > > Min -0.1% -74.5% -6.8% -8.3% -50.0% > Max +0.2% 0.0% +38.5% +38.5% 0.0% > Geometric Mean -0.0% -4.1% +7.7% +7.7% -0.8% > > Because this depends on a compiler optimisation, we have to watch out for cases > where this is not an improvements, and whether they occur in the wild. > > >> --------------------------------------------------------------- > > b63facef165b957183b65604ef99b2b8574747a5 > Data/List.hs | 34 +++++++++------------------------- > GHC/List.lhs | 13 +++++++------ > 2 files changed, 16 insertions(+), 31 deletions(-) > > diff --git a/Data/List.hs b/Data/List.hs > index 130ceb2..4796055 100644 > --- a/Data/List.hs > +++ b/Data/List.hs > @@ -1,5 +1,5 @@ > {-# LANGUAGE Trustworthy #-} > -{-# LANGUAGE CPP, NoImplicitPrelude, MagicHash #-} > +{-# LANGUAGE CPP, NoImplicitPrelude, ScopedTypeVariables, MagicHash #-} > > ----------------------------------------------------------------------------- > -- | > @@ -989,10 +989,11 @@ unfoldr f b = > -- ----------------------------------------------------------------------------- > > -- | A strict version of 'foldl'. > -foldl' :: (b -> a -> b) -> b -> [a] -> b > -foldl' f z0 xs0 = lgo z0 xs0 > - where lgo z [] = z > - lgo z (x:xs) = let z' = f z x in z' `seq` lgo z' xs > +foldl' :: forall a b . (b -> a -> b) -> b -> [a] -> b > +foldl' k z0 xs = foldr (\(v::a) (fn::b->b) (z::b) -> z `seq` fn (k z v)) (id :: b -> b) xs z0 > +-- Implementing foldl' via foldr is only a good idea if the compiler can optimize > +-- the resulting code (eta-expand the recursive "go"), so this needs -fcall-arity! > +-- Also see #7994 > > -- | 'foldl1' is a variant of 'foldl' that has no starting value argument, > -- and thus must be applied to non-empty lists. > @@ -1008,32 +1009,15 @@ foldl1' _ [] = errorEmptyList "foldl1'" > -- ----------------------------------------------------------------------------- > -- List sum and product > > -{-# SPECIALISE sum :: [Int] -> Int #-} > -{-# SPECIALISE sum :: [Integer] -> Integer #-} > -{-# INLINABLE sum #-} > -{-# SPECIALISE product :: [Int] -> Int #-} > -{-# SPECIALISE product :: [Integer] -> Integer #-} > -{-# INLINABLE product #-} > --- We make 'sum' and 'product' inlinable so that we get specialisations > --- at other types. See, for example, Trac #7507. > - > -- | The 'sum' function computes the sum of a finite list of numbers. > sum :: (Num a) => [a] -> a > -- | The 'product' function computes the product of a finite list of numbers. > product :: (Num a) => [a] -> a > -#ifdef USE_REPORT_PRELUDE > + > +{-# INLINE sum #-} > sum = foldl (+) 0 > +{-# INLINE product #-} > product = foldl (*) 1 > -#else > -sum l = sum' l 0 > - where > - sum' [] a = a > - sum' (x:xs) a = sum' xs (a+x) > -product l = prod l 1 > - where > - prod [] a = a > - prod (x:xs) a = prod xs (a*x) > -#endif > > -- ----------------------------------------------------------------------------- > -- Functions on strings > diff --git a/GHC/List.lhs b/GHC/List.lhs > index b7b78c7..e004ded 100644 > --- a/GHC/List.lhs > +++ b/GHC/List.lhs > @@ -1,6 +1,6 @@ > \begin{code} > {-# LANGUAGE Trustworthy #-} > -{-# LANGUAGE CPP, NoImplicitPrelude, MagicHash #-} > +{-# LANGUAGE CPP, NoImplicitPrelude, ScopedTypeVariables, MagicHash #-} > {-# OPTIONS_HADDOCK hide #-} > > ----------------------------------------------------------------------------- > @@ -178,11 +178,12 @@ filterFB c p x r | p x = x `c` r > -- can be inlined, and then (often) strictness-analysed, > -- and hence the classic space leak on foldl (+) 0 xs > > -foldl :: (b -> a -> b) -> b -> [a] -> b > -foldl f z0 xs0 = lgo z0 xs0 > - where > - lgo z [] = z > - lgo z (x:xs) = lgo (f z x) xs > +foldl :: forall a b. (b -> a -> b) -> b -> [a] -> b > +{-# INLINE foldl #-} > +foldl k z0 xs = foldr (\(v::a) (fn::b->b) (z::b) -> fn (k z v)) (id :: b -> b) xs z0 > +-- Implementing foldl via foldr is only a good idea if the compiler can optimize > +-- the resulting code (eta-expand the recursive "go"), so this needs -fcall-arity! > +-- Also see #7994 > > -- | 'scanl' is similar to 'foldl', but returns a list of successive > -- reduced values from the left: > > _______________________________________________ > ghc-commits mailing list > ghc-commits(a)haskell.org > http://www.haskell.org/mailman/listinfo/ghc-commits >

1 0

Re: 7.8.1-candidate fail
by Jan Stolarek 17 Feb '14

17 Feb '14

I had the same problem on Debian Squeeze: https://ghc.haskell.org/trac/ghc/ticket/8666 What is your distro? CCing ghc-devs. Janek Dnia czwartek, 6 lutego 2014, Sergei Meshveliani napisał: > Dear GHC team, > > I am trying to test ghc-7.8.20140130-src.tar.bz2 > > I make it from source with ghc-7.6.3 on Debian Linux (64 bit). > > ./configure looks all right. > > And `make' reports after 40 minutes: > > ------------------------------------------------------- > ... > ... > "inplace/bin/ghc-stage1" -optc-Ilibraries/integer-gmp/. > -optc-I'/home/mechvel/g..... > ... > ... > /usr/bin/ld: libraries/integer-gmp/gmp/objs/aors.o: relocation > R_X86_64_32 against `__gmpz_sub' can not be used when making a shared > object; recompile with -fPIC > libraries/integer-gmp/gmp/objs/aors.o: could not read symbols: Bad value > collect2: ld returned 1 exit status > make[1]: *** > [libraries/integer-gmp/dist-install/build/libHSinteger-gmp-0.5.1.0\ > -ghc7.8.20140130.so] Error 1 > ------------------------------------------------------- > > > What might this mean? Need I to install a fresher libgmp ? > > Thanks, > > ------ > Sergei > > _______________________________________________ > Glasgow-haskell-users mailing list > Glasgow-haskell-users(a)haskell.org > http://www.haskell.org/mailman/listinfo/glasgow-haskell-users

2 2

Re: Overlapping names in GHC: ConLike
by Carter Schonwald 15 Feb '14

15 Feb '14

BijectionLike, InvertableCon ReversableConstructor InvertableLike Dualable DualLike DuelFoil -- I kid basically something about how patternSynonyms and normal type constructors have this "bjiective" (ish?) quality On Fri, Feb 14, 2014 at 11:10 PM, Dr. ÉRDI Gergő <gergo(a)erdi.hu> wrote: > Can we force some more elements of that lazy stream? > On Feb 15, 2014 12:09 PM, "Carter Schonwald" <carter.schonwald(a)gmail.com> > wrote: > >> ConstructorAble ? (I'm happy to come up with a whole list of >> candidates). :-) >> >> On Friday, February 14, 2014, Dr. ERDI Gergo <gergo(a)erdi.hu> wrote: >> >>> On Thu, 13 Feb 2014, Carter Schonwald wrote: >>> >>> Caseable? PatternLike? >>>> >>> >>> The reason I went with ConLike originally (apart from not noticing the >>> clash) was that (bidirectional) pattern synonyms can also be used in >>> expressions, so they behave like constructors in both contexts. >> >>

1 0

[PATCH] rts: unrust 'libbfd' debug symbols parser
by slyich＠gmail.com 15 Feb '14

15 Feb '14

From: Sergei Trofimovich <slyfox(a)gentoo.org> Signed-off-by: Sergei Trofimovich <slyfox(a)gentoo.org> --- configure.ac | 3 ++- rts/Printer.c | 10 ++++++++-- rts/RtsStartup.c | 6 ++++++ 3 files changed, 16 insertions(+), 3 deletions(-) diff --git a/configure.ac b/configure.ac index e7fbc7f..e47979c 100644 --- a/configure.ac +++ b/configure.ac @@ -801,7 +801,8 @@ fi dnl ** check whether this machine has BFD and libiberty installed (used for debugging) dnl the order of these tests matters: bfd needs libiberty AC_CHECK_LIB(iberty, xmalloc) -AC_CHECK_LIB(bfd, bfd_uncompress_section_contents) +dnl 'bfd_init' is a rare non-macro in libbfd +AC_CHECK_LIB(bfd, bfd_init) dnl ################################################################ dnl Check for libraries diff --git a/rts/Printer.c b/rts/Printer.c index ca9ca49..ce02b02 100644 --- a/rts/Printer.c +++ b/rts/Printer.c @@ -48,6 +48,9 @@ void printPtr( StgPtr p ) raw = lookupGHCName(p); if (raw != NULL) { printZcoded(raw); + /* it can be just a C symbol, like 'stg_returnToStackTop' */ + debugBelch("<%s>", raw); + debugBelch("[%p]", p); } else { debugBelch("%p", p); } @@ -794,7 +797,7 @@ static void printZcoded( const char *raw ) disabling this for now. */ #ifdef USING_LIBBFD - +#include "../mk/config.h" /* silly BFD's requirement */ #include <bfd.h> /* Fairly ad-hoc piece of code that seems to filter out a lot of @@ -863,7 +866,10 @@ extern void DEBUG_LoadSymbols( char *name ) for( i = 0; i != number_of_symbols; ++i ) { symbol_info info; bfd_get_symbol_info(abfd,symbol_table[i],&info); - /*debugBelch("\t%c\t0x%x \t%s\n",info.type,(nat)info.value,info.name); */ + if (0) + { + debugBelch("\t%c\t0x%x \t%s\n",info.type,(nat)info.value,info.name); + } if (isReal(info.type, info.name)) { num_real_syms += 1; } diff --git a/rts/RtsStartup.c b/rts/RtsStartup.c index aa7306f..5bdef94 100644 --- a/rts/RtsStartup.c +++ b/rts/RtsStartup.c @@ -19,6 +19,7 @@ #include "RtsFlags.h" #include "RtsUtils.h" #include "Prelude.h" +#include "Printer.h" /* DEBUG_LoadSymbols */ #include "Schedule.h" /* initScheduler */ #include "Stats.h" /* initStats */ #include "STM.h" /* initSTM */ @@ -162,6 +163,11 @@ hs_init_ghc(int *argc, char **argv[], RtsConfig rts_config) rts_config.rts_opts_enabled, rts_config.rts_opts, rts_config.rts_hs_main); } +#ifdef DEBUG + /* load debugging symbols */ + DEBUG_LoadSymbols((*argv)[0]); +#endif /* DEBUG */ + /* Initialise the stats department, phase 1 */ initStats1(); -- 1.8.5.2

2 1

[PATCH] includes/stg/SMP.h: use 'NOSMP' instead of never defined 'WITHSMP'
by slyich＠gmail.com 15 Feb '14

15 Feb '14

From: Sergei Trofimovich <slyfox(a)gentoo.org> Signed-off-by: Sergei Trofimovich <slyfox(a)gentoo.org> --- includes/stg/SMP.h | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/includes/stg/SMP.h b/includes/stg/SMP.h index 01663dd..8d819c7 100644 --- a/includes/stg/SMP.h +++ b/includes/stg/SMP.h @@ -154,7 +154,7 @@ xchg(StgPtr p, StgWord w) : "r" (w), "r" (p) : "memory" ); -#elif !defined(WITHSMP) +#elif defined(NOSMP) result = *p; *p = w; #else @@ -225,7 +225,7 @@ cas(StgVolatilePtr p, StgWord o, StgWord n) : "cc","memory"); return result; -#elif !defined(WITHSMP) +#elif defined(NOSMP) StgWord result; result = *p; if (result == o) { @@ -313,7 +313,7 @@ write_barrier(void) { __asm__ __volatile__ ("" : : : "memory"); #elif arm_HOST_ARCH && !defined(arm_HOST_ARCH_PRE_ARMv7) __asm__ __volatile__ ("dmb st" : : : "memory"); -#elif !defined(WITHSMP) +#elif defined(NOSMP) return; #else #error memory barriers unimplemented on this architecture @@ -332,7 +332,7 @@ store_load_barrier(void) { __asm__ __volatile__ ("membar #StoreLoad" : : : "memory"); #elif arm_HOST_ARCH && !defined(arm_HOST_ARCH_PRE_ARMv7) __asm__ __volatile__ ("dmb" : : : "memory"); -#elif !defined(WITHSMP) +#elif defined(NOSMP) return; #else #error memory barriers unimplemented on this architecture @@ -352,7 +352,7 @@ load_load_barrier(void) { __asm__ __volatile__ ("" : : : "memory"); #elif arm_HOST_ARCH && !defined(arm_HOST_ARCH_PRE_ARMv7) __asm__ __volatile__ ("dmb" : : : "memory"); -#elif !defined(WITHSMP) +#elif defined(NOSMP) return; #else #error memory barriers unimplemented on this architecture -- 1.8.5.2

3 2

[PATCH] compiler/ghc.mk: restore GhcHcOpts variable handling
by slyich＠gmail.com 15 Feb '14

15 Feb '14

From: Sergei Trofimovich <slyfox(a)gentoo.org> Signed-off-by: Sergei Trofimovich <slyfox(a)gentoo.org> --- compiler/ghc.mk | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/compiler/ghc.mk b/compiler/ghc.mk index 4977e28..d902b7f 100644 --- a/compiler/ghc.mk +++ b/compiler/ghc.mk @@ -667,9 +667,9 @@ compiler_stage2_CONFIGURE_OPTS += --disable-library-for-ghci compiler_stage3_CONFIGURE_OPTS += --disable-library-for-ghci # after build-package, because that sets compiler_stage1_HC_OPTS: -compiler_stage1_HC_OPTS += $(GhcStage1HcOpts) -compiler_stage2_HC_OPTS += $(GhcStage2HcOpts) -compiler_stage3_HC_OPTS += $(GhcStage3HcOpts) +compiler_stage1_HC_OPTS += $(GhcHcOpts) $(GhcStage1HcOpts) +compiler_stage2_HC_OPTS += $(GhcHcOpts) $(GhcStage2HcOpts) +compiler_stage3_HC_OPTS += $(GhcHcOpts) $(GhcStage3HcOpts) ifneq "$(BINDIST)" "YES" -- 1.8.5.2

2 1

Overlapping names in GHC: ConLike
by Joachim Breitner 15 Feb '14

15 Feb '14

Hi, while reading some code I noticed that we use the term “ConLike” for two different things: From Haskell up to the desugarer, this is a data type to treat data constructors and pattern synonyms together; afterwards, in the world of Core, ConLike is an attribute of expression (exprIsConLik) similar to exprIsHNF. The former use is newer, introduced by the pattern synonym patch. In the long run, there is a risk of confusion, maybe we should rename ConLike module and data type. Any suggestions for a good name? Greetings, Joachim -- Joachim “nomeata” Breitner mail(a)joachim-breitner.de • http://www.joachim-breitner.de/ Jabber: nomeata(a)joachim-breitner.de • GPG-Key: 0x4743206C Debian Developer: nomeata(a)debian.org

3 3

RC2 status
by Austin Seipp 14 Feb '14

14 Feb '14

Hello all! I'm sure some of you are wondering about RC2. Please take a look at this page, which gives the basic overview: https://ghc.haskell.org/trac/ghc/wiki/Status/GHC-7.8/RC1 This currently lists all the active tickets filed against RC1. It lists several categories, but the only one really relevant is 'new' for those reading. Most of the bugs in 'new' will be fixed shortly. I will then merge them (and a lot of others) to 7.8 shortly. This includes #8735, #8736, #8760, and #8770 (and also #8719 and #8748 in 'patch' status.) After talking with Simon, I will also cook up a fix for #8754. So it's closer than it looks! There are two outliers: * #8745 & #8773 - After Richard and Joachim hashed out the details (fairly quickly,) it seems this will go into 7.8 as well. Richard - I suspect you will likely commit #8773 soon? Afterwords, I can still take care of #8745 to enable GND for SafeHaskell, and these can both be merged. Quite easy. * #8696 - this one is the main blocker for RC2. And after I'm done brushing the other stuff up, it will be my main focus as I haven't had time to think about it closely. Reid Barton has already been fantastically helpful with his test case and thoughtful analysis, and perhaps someone else would look. I'd really appreciate extra eyes to help diagnose a possible fix or even a patch - the human parallelism could certainly help. So that's it: most of these are very basic, and I expect to be done with many quite soon. The last bit here is the sticky one, so I encourage people to look if they have a chance, please! Finally, a quick note: https://ghc.haskell.org/trac/ghc/wiki/Status/GHC-7.8/RC1 has content which is generated dynamically. You can automatically make another ticket appear there by: * Setting the version to '7.8.1-rc1'. * Setting the milestone to '7.8.1'. This will make sure it's clearly on my radar. The bugs have slowed down however, which is a good thing: so I imagine this is quite representative of the final ones. Let me know if there are any questions or more importantly if I've missed anything. -- Regards, Austin Seipp, Haskell Consultant Well-Typed LLP, http://www.well-typed.com/

3 2