pool/ghc
SHA256
4
0
Fork 1
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Accepting request 305686 from devel:languages:haskell

Browse Source 
+ add ghc-glibc-2.20_BSD_SOURCE.patch from fedora
- updated to 7.8.4 
+ removed D177.patch (in upstream)
+ removed D173.patch (in upstream)
+ removed ghc.git-e18525f.patch (in upstream)
+ refresh 0001-implement-native-code-generator-for-ppc64.patch
* A critical bug in the LLVM backend which would cause the compiler to generate incorrect code has been fixed (issue #9439).
* Several bugs in the code generator have been fixed for unregisterised platforms, such as 64bit PowerPC (issue #8819 and #8849).
* A bug that could cause GHC's constructor specialization pass (enabled by default at -O2, or via -fspec-constr) to loop forever and consume large amounts of memory has been fixed (issue #8960).
* A bug that would cause GHC to fail when attempting to determine GCC's version information in non-english locales has been fixed (issue #8825).
* A minor bug that allowed GHC to seemingly import (but not use) private data constructors has been fixed (issue #9006).
* A bug in the register allocator which would cause GHC to crash during compilation has been fixed (issue #9303).
* A bug that caused the compiler to panic on some input C-- code has been fixed (issue #9329).
* A few various minor deadlocks in the runtime system when using forkProcess have been fixed.
* A bug which made blocked STM transactions non-interruptible has been fixed (issue #9379).
* A bug in the compiler which broke pattern synonym imports across modules in Haddock has been fixed (issue #9417).
* A minor bug in the code generator in which the popCnt16# did not zero-extend its result has been fixed (issue #9435).
* A bug which caused the compiler to panic on pattern synonyms inside a class declaration has been fixed (issue #9705).
* A bug in the typechecker revolving around un-saturated type family applications has been fixed (issue #9433).
* Several bugs have been fixed causing problems with building GHC on ARM (issues #8951, #9620, #9336, and #9552).
* A bug in the typechecker that could cause an infinite loop when using superclasses in a cycle has been fixed (issue #9415).
* A bug causing corruption in signal handling with the single-threaded runtime system has been fixed (issue #9817).
* A bug that could cause compiled programs to crash due to use of overlapping type families has been fixed (issue #9371).
* A bug in the inliner that caused certain expressions within unboxed tuples to not be properly evaluated has been fixed (issue #9390).
* A bug that caused the compiler to not always properly detect LLVM tools (particularly on Windows) has been fixed (issue #7143).
* A bug that prevented GHC from deriving Generic1 instances for data families has been fixed (#9563).
* A bug that caused type inference to infer the incorrect type in the presence of certain type families and constraints has been fixed (issue #9316).

OBS-URL: https://build.opensuse.org/request/show/305686
OBS-URL: https://build.opensuse.org/package/show/openSUSE:Factory/ghc?expand=0&rev=26
This commit is contained in:
Stephan Kulow 2015-05-10 08:45:58 +00:00 committed by Git OBS Bridge
commit 9d761d8ba2
9 changed files with 63 additions and 231 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
0001-implement-native-code-generator-for-ppc64.patch
View File

@ -1809,16 +1809,16 @@ Index: ghc-7.8.3/compiler/nativeGen/RegAlloc/Linear/FreeRegs.hs
ArchAlpha -> panic "maxSpillSlots ArchAlpha"
ArchMipseb -> panic "maxSpillSlots ArchMipseb"
ArchMipsel -> panic "maxSpillSlots ArchMipsel"
Index: ghc-7.8.3/compiler/nativeGen/RegAlloc/Linear/Main.hs
Index: ghc-7.8.4/compiler/nativeGen/RegAlloc/Linear/Main.hs
===================================================================
--- ghc-7.8.3.orig/compiler/nativeGen/RegAlloc/Linear/Main.hs
+++ ghc-7.8.3/compiler/nativeGen/RegAlloc/Linear/Main.hs
@@ -207,7 +207,7 @@ linearRegAlloc dflags first_id block_liv
ArchSPARC -> linearRegAlloc' dflags (frInitFreeRegs platform :: SPARC.FreeRegs) first_id block_live sccs
ArchPPC -> linearRegAlloc' dflags (frInitFreeRegs platform :: PPC.FreeRegs) first_id block_live sccs
--- ghc-7.8.4.orig/compiler/nativeGen/RegAlloc/Linear/Main.hs
+++ ghc-7.8.4/compiler/nativeGen/RegAlloc/Linear/Main.hs
@@ -207,7 +207,7 @@ linearRegAlloc dflags entry_ids block_liv
ArchSPARC -> linearRegAlloc' dflags (frInitFreeRegs platform :: SPARC.FreeRegs) entry_ids block_live sccs
ArchPPC -> linearRegAlloc' dflags (frInitFreeRegs platform :: PPC.FreeRegs) entry_ids block_live sccs
ArchARM _ _ _ -> panic "linearRegAlloc ArchARM"
- ArchPPC_64 -> panic "linearRegAlloc ArchPPC_64"
+ ArchPPC_64 -> linearRegAlloc' dflags (frInitFreeRegs platform :: PPC.FreeRegs) first_id block_live sccs
+ ArchPPC_64 -> linearRegAlloc' dflags (frInitFreeRegs platform :: PPC.FreeRegs) entry_ids block_live sccs
ArchAlpha -> panic "linearRegAlloc ArchAlpha"
ArchMipseb -> panic "linearRegAlloc ArchMipseb"
ArchMipsel -> panic "linearRegAlloc ArchMipsel"

51
D173.patch
View File

@ -1,51 +0,0 @@
Index: ghc-7.8.3/compiler/cmm/PprC.hs
===================================================================
--- ghc-7.8.3.orig/compiler/cmm/PprC.hs
+++ ghc-7.8.3/compiler/cmm/PprC.hs
@@ -1220,8 +1220,9 @@ commafy xs = hsep $ punctuate comma xs
pprHexVal :: Integer -> Width -> SDoc
pprHexVal 0 _ = ptext (sLit "0x0")
pprHexVal w rep
- | w < 0 = parens (char '-' <> ptext (sLit "0x") <> go (-w) <> repsuffix rep)
- | otherwise = ptext (sLit "0x") <> go w <> repsuffix rep
+ | w < 0 = parens (char '-' <>
+ ptext (sLit "0x") <> intToDoc (-w) <> repsuffix rep)
+ | otherwise = ptext (sLit "0x") <> intToDoc w <> repsuffix rep
where
-- type suffix for literals:
-- Integer literals are unsigned in Cmm/C. We explicitly cast to
@@ -1236,10 +1237,33 @@ pprHexVal w rep
else panic "pprHexVal: Can't find a 64-bit type"
repsuffix _ = char 'U'
+ intToDoc :: Integer -> SDoc
+ intToDoc i = go (truncInt i)
+
+ -- We need to truncate value as Cmm backend does not drop
+ -- redundant bits to ease handling of negative values.
+ -- Thus the following Cmm code on 64-bit arch, like amd64:
+ -- CInt v;
+ -- v = {something};
+ -- if (v == %lobits32(-1)) { ...
+ -- leads to the following C code:
+ -- StgWord64 v = (StgWord32)({something});
+ -- if (v == 0xFFFFffffFFFFffffU) { ...
+ -- Such code is incorrect as it promotes both operands to StgWord64
+ -- and the whole condition is always false.
+ truncInt :: Integer -> Integer
+ truncInt i =
+ case rep of
+ W8 -> i `rem` (2^(8 :: Int))
+ W16 -> i `rem` (2^(16 :: Int))
+ W32 -> i `rem` (2^(32 :: Int))
+ W64 -> i `rem` (2^(64 :: Int))
+ _ -> panic ("pprHexVal/truncInt: C backend can't encode "
+ ++ show rep ++ " literals")
+
go 0 = empty
go w' = go q <> dig
where
(q,r) = w' `quotRem` 16
dig | r < 10 = char (chr (fromInteger r + ord '0'))
| otherwise = char (chr (fromInteger r - 10 + ord 'a'))
-

80
D177.patch
View File

@ -1,80 +0,0 @@
Index: ghc-7.8.3/compiler/main/DriverPipeline.hs
===================================================================
--- ghc-7.8.3.orig/compiler/main/DriverPipeline.hs
+++ ghc-7.8.3/compiler/main/DriverPipeline.hs
@@ -1208,6 +1208,7 @@ runPhase (RealPhase (As with_cpp)) input
as_prog <- whichAsProg
let cmdline_include_paths = includePaths dflags
+ let pic_c_flags = picCCOpts dflags
next_phase <- maybeMergeStub
output_fn <- phaseOutputFilename next_phase
@@ -1221,6 +1222,9 @@ runPhase (RealPhase (As with_cpp)) input
= liftIO $ as_prog dflags
([ SysTools.Option ("-I" ++ p) | p <- cmdline_include_paths ]
+ -- See Note [-fPIC for assembler]
+ ++ map SysTools.Option pic_c_flags
+
-- We only support SparcV9 and better because V8 lacks an atomic CAS
-- instruction so we have to make sure that the assembler accepts the
-- instruction set. Note that the user can still override this
@@ -1262,6 +1266,8 @@ runPhase (RealPhase SplitAs) _input_fn d
osuf = objectSuf dflags
split_odir = base_o ++ "_" ++ osuf ++ "_split"
+ let pic_c_flags = picCCOpts dflags
+
-- this also creates the hierarchy
liftIO $ createDirectoryIfMissing True split_odir
@@ -1295,6 +1301,9 @@ runPhase (RealPhase SplitAs) _input_fn d
then [SysTools.Option "-mcpu=v9"]
else []) ++
+ -- See Note [-fPIC for assembler]
+ map SysTools.Option pic_c_flags ++
+
[ SysTools.Option "-c"
, SysTools.Option "-o"
, SysTools.FileOption "" (split_obj n)
@@ -2210,3 +2219,38 @@ haveRtsOptsFlags dflags =
isJust (rtsOpts dflags) || case rtsOptsEnabled dflags of
RtsOptsSafeOnly -> False
_ -> True
+
+-- Note [-fPIC for assembler]
+-- When compiling .c source file GHC's driver pipeline basically
+-- does the following two things:
+-- 1. ${CC} -S 'PIC_CFLAGS' source.c
+-- 2. ${CC} -x assembler -c 'PIC_CFLAGS' source.S
+--
+-- Why do we need to pass 'PIC_CFLAGS' both to C compiler and assembler?
+-- Because on some architectures (at least sparc32) assembler also choses
+-- relocation type!
+-- Consider the following C module:
+--
+-- /* pic-sample.c */
+-- int v;
+-- void set_v (int n) { v = n; }
+-- int get_v (void) { return v; }
+--
+-- $ gcc -S -fPIC pic-sample.c
+-- $ gcc -c pic-sample.s -o pic-sample.no-pic.o # incorrect binary
+-- $ gcc -c -fPIC pic-sample.s -o pic-sample.pic.o # correct binary
+--
+-- $ objdump -r -d pic-sample.pic.o > pic-sample.pic.o.od
+-- $ objdump -r -d pic-sample.no-pic.o > pic-sample.no-pic.o.od
+-- $ diff -u pic-sample.pic.o.od pic-sample.no-pic.o.od
+--
+-- Most of architectures won't show any difference in this test, but on sparc32
+-- the following assembly snippet:
+--
+-- sethi %hi(_GLOBAL_OFFSET_TABLE_-8), %l7
+--
+-- generates two kinds or relocations, only 'R_SPARC_PC22' is correct:
+--
+-- 3c: 2f 00 00 00 sethi %hi(0), %l7
+-- - 3c: R_SPARC_PC22 _GLOBAL_OFFSET_TABLE_-0x8
+-- + 3c: R_SPARC_HI22 _GLOBAL_OFFSET_TABLE_-0x8

14
_constraints
View File

@ -2,9 +2,19 @@
<hardware>
<disk>
<size unit="G">12</size>
</disk>
</disk>
<physicalmemory>
<size unit="G">4</size>
<size unit="G">4</size>
</physicalmemory>
</hardware>
<overwrite>
<conditions>
<arch>ppc64le</arch>
</conditions>
<hardware>
<memory>
<size unit="G">8</size>
</memory>
</hardware>
</overwrite>
</constraints>

3
ghc-7.8.3-src.tar.xz
View File

@ -1,3 +0,0 @@
version https://git-lfs.github.com/spec/v1
oid sha256:b0cd96a549ba3b5e512847a4a8cd1a3174e4b2b75dadfc41c568fb812887b958
size 9160092

3
ghc-7.8.4-src.tar.xz Normal file
View File

@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:c319cd94adb284177ed0e6d21546ed0b900ad84b86b87c06a99eac35152982c4
size 9128576

31
ghc.changes
View File

@ -20,6 +20,37 @@ Mon Feb 23 16:48:10 UTC 2015 - peter.trommler@ohm-hochschule.de
- drop ghc-cabal-unversion-docdir.patch
* ghc-rpm-macros follows upstream ghc again
-------------------------------------------------------------------
Wed Feb 4 11:57:35 UTC 2015 - mimi.vx@gmail.com
+ add ghc-glibc-2.20_BSD_SOURCE.patch from fedora
- updated to 7.8.4
+ removed D177.patch (in upstream)
+ removed D173.patch (in upstream)
+ removed ghc.git-e18525f.patch (in upstream)
+ refresh 0001-implement-native-code-generator-for-ppc64.patch
* A critical bug in the LLVM backend which would cause the compiler to generate incorrect code has been fixed (issue #9439).
* Several bugs in the code generator have been fixed for unregisterised platforms, such as 64bit PowerPC (issue #8819 and #8849).
* A bug that could cause GHC's constructor specialization pass (enabled by default at -O2, or via -fspec-constr) to loop forever and consume large amounts of memory has been fixed (issue #8960).
* A bug that would cause GHC to fail when attempting to determine GCC's version information in non-english locales has been fixed (issue #8825).
* A minor bug that allowed GHC to seemingly import (but not use) private data constructors has been fixed (issue #9006).
* A bug in the register allocator which would cause GHC to crash during compilation has been fixed (issue #9303).
* A bug that caused the compiler to panic on some input C-- code has been fixed (issue #9329).
* A few various minor deadlocks in the runtime system when using forkProcess have been fixed.
* A bug which made blocked STM transactions non-interruptible has been fixed (issue #9379).
* A bug in the compiler which broke pattern synonym imports across modules in Haddock has been fixed (issue #9417).
* A minor bug in the code generator in which the popCnt16# did not zero-extend its result has been fixed (issue #9435).
* A bug which caused the compiler to panic on pattern synonyms inside a class declaration has been fixed (issue #9705).
* A bug in the typechecker revolving around un-saturated type family applications has been fixed (issue #9433).
* Several bugs have been fixed causing problems with building GHC on ARM (issues #8951, #9620, #9336, and #9552).
* A bug in the typechecker that could cause an infinite loop when using superclasses in a cycle has been fixed (issue #9415).
* A bug causing corruption in signal handling with the single-threaded runtime system has been fixed (issue #9817).
* A bug that could cause compiled programs to crash due to use of overlapping type families has been fixed (issue #9371).
* A bug in the inliner that caused certain expressions within unboxed tuples to not be properly evaluated has been fixed (issue #9390).
* A bug that caused the compiler to not always properly detect LLVM tools (particularly on Windows) has been fixed (issue #7143).
* A bug that prevented GHC from deriving Generic1 instances for data families has been fixed (#9563).
* A bug that caused type inference to infer the incorrect type in the presence of certain type families and constraints has been fixed (issue #9316).
-------------------------------------------------------------------
Tue Feb 3 13:35:25 UTC 2015 - mimi.vx@gmail.com

71
ghc.git-e18525f.patch
View File

@ -1,71 +0,0 @@
From: Sergei Trofimovich <slyfox@gentoo.org>
Date: Thu, 4 Sep 2014 14:50:45 +0000 (+0300)
Subject: pprC: declare extern cmm primitives as functions, not data
X-Git-Url: https://git.haskell.org/ghc.git/commitdiff_plain/e18525fae273f4c1ad8d6cbe1dea4fc074cac721
pprC: declare extern cmm primitives as functions, not data
Summary:
The commit fixes incorrect code generation of
integer-gmp package on ia64 due to C prototypes mismatch.
Before the patch prototypes for "foreign import prim" were:
StgWord poizh[];
After the patch they became:
StgFunPtr poizh();
Long story:
Consider the following simple example:
{-# LANGUAGE MagicHash, GHCForeignImportPrim, UnliftedFFITypes #-}
module M where
import GHC.Prim -- Int#
foreign import prim "poizh" poi# :: Int# -> Int#
Before the patch unregisterised build generated the
following 'poizh' reference:
EI_(poizh); /* StgWord poizh[]; */
FN_(M_poizh_entry) {
// ...
JMP_((W_)&poizh);
}
After the patch it looks this way:
EF_(poizh); /* StgFunPtr poizh(); */
FN_(M_poizh_entry) {
// ...
JMP_((W_)&poizh);
}
On ia64 it leads to different relocation types being generated:
incorrect one:
addl r14 = @ltoffx(poizh#)
ld8.mov r14 = [r14], poizh# ; r14 = address-of 'poizh#'
correct one:
addl r14 = @ltoff(@fptr(poizh#)), gp ; r14 = address-of-thunk 'poizh#'
ld8 r14 = [r14]
'@fptr(poizh#)' basically instructs assembler to creates
another obect consisting of real address to 'poizh' instructions
and module address. That '@fptr' object is used as a function "address"
This object is different for every module referencing 'poizh' symbol.
All indirect function calls expect '@fptr' object. That way
call site reads real destination address and set destination
module address in 'gp' register from '@fptr'.
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
---
Index: ghc-7.8.3/compiler/cmm/CLabel.hs
===================================================================
--- ghc-7.8.3.orig/compiler/cmm/CLabel.hs
+++ ghc-7.8.3/compiler/cmm/CLabel.hs
@@ -801,6 +801,7 @@ labelType (CmmLabel _ _ CmmClosure)
labelType (CmmLabel _ _ CmmCode) = CodeLabel
labelType (CmmLabel _ _ CmmInfo) = DataLabel
labelType (CmmLabel _ _ CmmEntry) = CodeLabel
+labelType (CmmLabel _ _ CmmPrimCall) = CodeLabel
labelType (CmmLabel _ _ CmmRetInfo) = DataLabel
labelType (CmmLabel _ _ CmmRet) = CodeLabel
labelType (RtsLabel (RtsSelectorInfoTable _ _)) = DataLabel

27
ghc.spec
View File

@ -19,14 +19,13 @@
%global unregisterised_archs ppc64le s390 s390s
Name: ghc
Version: 7.8.3
Version: 7.8.4
Release: 0
Url: http://haskell.org/ghc/dist/%{version}/%{name}-%{version}-src.tar.bz2
Summary: The Glorious Glasgow Haskell Compiler
License: BSD-3-Clause
Group: Development/Languages/Other
ExclusiveArch: %{ix86} x86_64 ppc ppc64 ppc64le
BuildRequires: binutils-devel
BuildRequires: gcc
BuildRequires: ghc-bootstrap >= 7.4
@ -35,6 +34,9 @@ BuildRequires: glibc-devel
BuildRequires: gmp-devel
BuildRequires: libelf-devel
BuildRequires: libffi-devel
%ifnarch ppc
BuildRequires: llvm
%endif
BuildRequires: ncurses-devel
BuildRequires: pkg-config
BuildRequires: xz
@ -61,12 +63,6 @@ Patch12: D349.patch
Patch13: integer-gmp.patch
# PATCH-FIX-UPSTREAM ghc-7.8.2-cgen-constify.patch peter.trommler@ohm-hochschule.de - Make constant strings constant in C backend to save data segment space. This is a gentoo patch.
Patch14: ghc-7.8.2-cgen-constify.patch
# PATCH-FIX-UPSTREAM D173.patch peter.trommler@ohm-hochschule.de -- Fix C backend. Backport of upstream fix for 7.10. See https://phabricator.haskell.org/D173.
Patch15: D173.patch
# PATCH-FIX-UPSTREAM D177.patch peter.trommler@ohm-hochschule.de -- Pass PIC flags to assembler. See https://phabricator.haskell.org/D177.
Patch16: D177.patch
# PATCH-FIX-UPSTREAM ghc.git-e18525f.patch peter.trommler@ohm-hochschule.de -- Declare extern cmm primitives as functions not data. Backport of upstream fix for 7.10. See https://git.haskell.org/ghc.git/commitdiff_plain/e18525fae273f4c1ad8d6cbe1dea4fc074cac721.
Patch17: ghc.git-e18525f.patch
# PATCH-FIX-UPSTREAM D560.patch peter.trommler@ohm-hochschule.de -- Fix loading of PIC register. See https://phabricator.haskell.org/D560.
Patch18: D560.patch
# PATCH-FEATURE-UPSTREAM 0001-implement-native-code-generator-for-ppc64.patch peter.trommler@ohm-hochschule.de -- Implement native code generator for ppc64. Haskell Trac #9863.
@ -116,9 +112,9 @@ To install all of GHC install package ghc.
%global ghc_pkg_c_deps ghc-compiler = %{ghc_version_override}-%{release}
%if %{defined ghclibdir}
%ghc_lib_subpackage Cabal 1.18.1.3
%ghc_lib_subpackage Cabal 1.18.1.5
%ghc_lib_subpackage array 0.5.0.0
%ghc_lib_subpackage -c gmp-devel,libffi-devel base 4.7.0.1
%ghc_lib_subpackage -c gmp-devel,libffi-devel base 4.7.0.2
%ghc_lib_subpackage binary 0.7.1.0
%ghc_lib_subpackage bytestring 0.10.4.0
%ghc_lib_subpackage containers 0.5.5.1
@ -167,9 +163,6 @@ except the ghc library, which is installed by the toplevel ghc metapackage.
%patch12 -p1
%patch13 -p1
%patch14 -p1
%patch15 -p1
%patch16 -p1
%patch17 -p1
%patch18 -p1
%patch19 -p1
%patch20 -p1
@ -345,13 +338,13 @@ rm testghc/*
# groups under a single name 'runhaskell'. Either these tools should be
# disentangled from the Haskell implementations or all implementations should
# have the same set of tools. *sigh*
"%_sbindir/update-alternatives" --install %{_bindir}/runhaskell runhaskell %{_bindir}/runghc 500
"%_sbindir/update-alternatives" --install %{_bindir}/hsc2hs hsc2hs %{_bindir}/hsc2hs-ghc 500
"%{_sbindir}/update-alternatives" --install %{_bindir}/runhaskell runhaskell %{_bindir}/runghc 500
"%{_sbindir}/update-alternatives" --install %{_bindir}/hsc2hs hsc2hs %{_bindir}/hsc2hs-ghc 500
%preun compiler
if test "$1" = 0; then
"%_sbindir/update-alternatives" --remove runhaskell %{_bindir}/runghc
"%_sbindir/update-alternatives" --remove hsc2hs %{_bindir}/hsc2hs-ghc
"%{_sbindir}/update-alternatives" --remove runhaskell %{_bindir}/runghc
"%{_sbindir}/update-alternatives" --remove hsc2hs %{_bindir}/hsc2hs-ghc
fi
%files