blob: edd527ba731c9e8bb75df6b03ada6c52e258f026 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
|
# $Id: PKGBUILD 167880 2012-10-03 15:19:04Z tdziedzic $
# Maintainer: Thomas Dziedzic <gostrc@gmail.com>
# Maintainer: Vesa Kaihlavirta <vesa@archlinux.org>
# Special note for devs looking to upgrade this package:
# ghc places a unique hash for each library when it is built.
# Libraries depend on versions specified by those hashes.
# This implies that all libraries need to be rebuilt when ghc is rebuilt.
pkgname=ghc
pkgver=7.6.1
pkgrel=1
pkgdesc='The Glasgow Haskell Compiler'
arch=('i686' 'x86_64')
url='http://www.haskell.org/ghc/'
license=('custom')
# libffi might be needed in the future: http://hackage.haskell.org/trac/ghc/ticket/4496
depends=('perl' 'gmp>=5.0' 'gcc') # 'libffi')
makedepends=('ghc' 'perl' 'libxslt' 'docbook-xsl')
checkdepends=('python2')
install='ghc.install'
# keep this otherwise a lot of tests will show as unexpected failures... need to investigate why using -jn for n > 1 does this...
#options=('!makeflags')
# missing rtl?
provides=('haskell-array=0.4.0.1'
'haskell-base=4.6.0.0'
'haskell-binary=0.5.1.1'
'haskell-bin-package-db=0.0.0.0'
'haskell-bytestring=0.10.0.0'
'haskell-containers=0.5.0.0'
'haskell-deepseq=1.3.0.1'
'haskell-directory=1.2.0.0'
'haskell-filepath=1.3.0.1'
'haskell-ghc-prim=0.3.0.0'
'haskell-haskell2010=1.1.1.0'
'haskell-haskell98=2.0.0.2'
'haskell-hoopl=3.9.0.0'
'haskell-hpc=0.6.0.0'
'haskell-integer-gmp=0.5.0.0'
'haskell-old-locale=1.0.0.5'
'haskell-old-time=1.1.0.1'
'haskell-pretty=1.1.1.0'
'haskell-process=1.1.0.2'
'haskell-template-haskell=2.8.0.0'
'haskell-time=1.4.0.1'
'haskell-unix=2.6.0.0'
'haskell-cabal=1.16.0')
replaces=('haskell-array<0.4.0.1'
'haskell-base<4.6.0.0'
'haskell-binary<0.5.1.1'
'haskell-bin-package-db<0.0.0.0'
'haskell-bytestring<0.10.0.0'
'haskell-containers<0.5.0.0'
'haskell-deepseq<1.3.0.1'
'haskell-directory<1.2.0.0'
'haskell-filepath<1.3.0.1'
'haskell-ghc-prim<0.3.0.0'
'haskell-haskell2010<1.1.1.0'
'haskell-haskell98<2.0.0.2'
'haskell-hoopl<3.9.0.0'
'haskell-hpc<0.6.0.0'
'haskell-integer-gmp<0.5.0.0'
'haskell-old-locale<1.0.0.5'
'haskell-old-time<1.1.0.1'
'haskell-pretty<1.1.1.0'
'haskell-process<1.1.0.2'
'haskell-template-haskell<2.8.0.0'
'haskell-time<1.4.0.1'
'haskell-unix<2.6.0.0'
'haskell-cabal<1.16.0')
source=("http://www.haskell.org/ghc/dist/${pkgver}/ghc-${pkgver}-src.tar.bz2"
"http://www.haskell.org/ghc/dist/${pkgver}/ghc-${pkgver}-testsuite.tar.bz2"
'build.mk')
md5sums=('e9d892d2970550ec5e70130b3621f6b6'
'0d93c458e3b445a7e524ef6ae459ce11'
'c367ef26300648ee9b8aca3dee5c9669')
build() {
cd ghc-${pkgver}
cp ${srcdir}/build.mk mk/build.mk
./configure \
--prefix=/usr
make
}
check() {
# keep this otherwise a lot of tests will show as unexpected failures... need to investigate why using -jn for n > 1 does this...
unset MAKEFLAGS
cd ghc-${pkgver}
# python2 rename
sed -e 's/PYTHON = python/&2/' -i testsuite/mk/boilerplate.mk
sed -e 's_#!/usr/bin/env python_&2_' -i testsuite/timeout/calibrate testsuite/timeout/timeout.py
# upstream known failures on both i686 and x86_64:
# ghci.debugger/scripts break011 [bad stderr] (ghci)
# perf/compiler parsing001 [stat not good enough] (normal)
# simplCore/should_compile spec-inline [stderr mismatch] (optasm)
# upstream known failures on i686 only:
# numrun012 test fails on i686 works on x86_64 - http://hackage.haskell.org/trac/ghc/ticket/5856
make test
# enabling threads to speed up
# causes more failures than make test, need to investigate
# upstream known failures on both i686 and x86_64:
# perf/compiler parsing001 [stat not good enough] (normal)
# simplCore/should_compile spec-inline [stderr mismatch] (optasm)
# upstream known failures on i686 only:
# numrun012 test fails on i686 works on x86_64 - http://hackage.haskell.org/trac/ghc/ticket/5856
#make THREADS=9 test
# will most likely cause lots of failures
# enable this when upstream has 0 test failures from a simple 'make test'
# zero unexpected failures on all tier 1 platforms - http://hackage.haskell.org/trac/ghc/ticket/5757
#make fulltest
}
package() {
cd ghc-${pkgver}
make DESTDIR=${pkgdir} install
install -d ${pkgdir}/usr/share/licenses/ghc
install -m644 LICENSE \
${pkgdir}/usr/share/licenses/ghc
}
|