+ case "${usergroup_phase}" in + local run_cmd + run_cmd=run_su + shift + run_su /opt/local/bin/bmake configure BATCH=1 DEPENDS_TARGET=/nonexistent PYTHON_VERSION_REQD=311 WRKLOG=/tmp/bulklog/ghc-9.6.3nb3/work.log + su pbulk -c '"$@"' make /opt/local/bin/bmake configure BATCH=1 DEPENDS_TARGET=/nonexistent PYTHON_VERSION_REQD=311 WRKLOG=/tmp/bulklog/ghc-9.6.3nb3/work.log => Bootstrap dependency digest>=20211023: found digest-20220214 => Checksum BLAKE2s OK for ghc-9.2.1-boot-hadrian.tar.gz => Checksum SHA512 OK for ghc-9.2.1-boot-hadrian.tar.gz => Checksum BLAKE2s OK for ghc-9.2.1-boot-x86_64-unknown-solaris2.tar.xz => Checksum SHA512 OK for ghc-9.2.1-boot-x86_64-unknown-solaris2.tar.xz => Checksum BLAKE2s OK for ghc-9.6.3-src.tar.xz => Checksum SHA512 OK for ghc-9.6.3-src.tar.xz ===> Installing dependencies for ghc-9.6.3nb3 ========================================================================== The supported build options for ghc are: llvm You can select which build options to use by setting PKG_DEFAULT_OPTIONS or the following variable. Its current value is shown: PKG_OPTIONS.ghc (not defined) ========================================================================== ========================================================================== The following variables will affect the build process of this package, ghc-9.6.3nb3. Their current value is shown below: * CURSES_DEFAULT = ncurses * PYTHON_VERSION_DEFAULT = 311 Based on these variables, the following variables have been set: * CURSES_TYPE = ncurses * PYPACKAGE = python311 You may want to abort the process now with CTRL-C and change the value of variables in the first group before continuing. Be sure to run `/opt/local/bin/bmake clean' after the changes. ========================================================================== => Tool dependency python311>=3.11: found python311-3.11.6nb4 => Tool dependency libiconv>=1.9.1: found libiconv-1.17 => Tool dependency ncurses>=6.0: found ncurses-6.4 => Tool dependency libtool-base>=2.4.2nb9: found libtool-base-2.4.7nb1 => Tool dependency mktools-[0-9]*: found mktools-20220614 => Tool dependency autoconf>=2.50: found autoconf-2.71nb2 => Tool dependency gmake>=3.81: found gmake-4.4.1 => Tool dependency cwrappers>=20150314: found cwrappers-20220403 => Tool dependency checkperms>=1.1: found checkperms-1.12 => Build dependency gcc12>=12.2.0: found gcc12-12.2.0 => Full dependency gcc12-libs>=12.2.0: found gcc12-libs-12.2.0 => Full dependency gmp>=5.0.1: found gmp-6.3.0 => Full dependency libiconv>=1.9.1nb4: found libiconv-1.17 => Full dependency libffi>=3.4.2nb3: found libffi-3.4.4 => Full dependency ncurses>=6.3nb1: found ncurses-6.4 ===> Overriding tools for ghc-9.6.3nb3 ===> Extracting for ghc-9.6.3nb3 => Substituting "ghc-wrapper-sh" in mk/install_script.sh ===> Patching for ghc-9.6.3nb3 => Applying pkgsrc patches for ghc-9.6.3nb3 => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-compiler_GHC_Driver_Pipeline_Execute.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-compiler_GHC_Driver_Pipeline_Execute.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-compiler_GHC_Driver_Pipeline_Execute.hs,v 1.1 2023/10/23 08:06:48 pho Exp $ | |This is pkgsrc-specific. Do not use -Xpreprocessor to pass options to |CPP. Our tool wrappers may remove some of them, and if that happens an |unpaired -Xpreprocessor will be left behind and confuses CC. | |--- compiler/GHC/Driver/Pipeline/Execute.hs.orig 2022-12-23 16:19:02.000000000 +0000 |+++ compiler/GHC/Driver/Pipeline/Execute.hs -------------------------- Patching file compiler/GHC/Driver/Pipeline/Execute.hs using Plan A... Hunk #1 succeeded at 464 (offset 91 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-compiler_GHC_SysTools_Process.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-compiler_GHC_SysTools_Process.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-compiler_GHC_SysTools_Process.hs,v 1.1 2023/10/23 08:06:48 pho Exp $ | |This is more like pkgsrc-specific than an upstream issue. Unless we |are on Windows, don't use "response file", the @FILE syntax while |invoking CC or LD. Our tool wrappers don't look into response files, |and therefore can't detect mistakes like depending on Cabal packages |that aren't declared as pkgsrc dependency. | |Note that this patch would break GHC if we didn't have its $ORIGIN use |eliminated because our wrappers hate relative rpaths. | |--- compiler/GHC/SysTools/Process.hs.orig 2022-12-23 16:19:02.000000000 +0000 |+++ compiler/GHC/SysTools/Process.hs -------------------------- Patching file compiler/GHC/SysTools/Process.hs using Plan A... Hunk #1 succeeded at 160. Hunk #2 succeeded at 199. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-configure.ac => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-configure.ac Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-configure.ac,v 1.1 2023/10/23 08:06:48 pho Exp $ | |Hunk #0: | Disable DTrace support on solaris2. | XXX: But why? | |--- configure.ac.orig 2022-12-23 16:19:02.000000000 +0000 |+++ configure.ac -------------------------- Patching file configure.ac using Plan A... Hunk #1 succeeded at 780 (offset -20 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_bindist_Makefile => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_bindist_Makefile Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-hadrian_bindist_Makefile,v 1.1 2023/10/23 08:06:48 pho Exp $ | |Don't run `ghc-pkg recache' while installing a binary distribution. See |also "do-install" in Makefile. | |--- hadrian/bindist/Makefile.orig 2023-01-12 03:34:04.854847389 +0000 |+++ hadrian/bindist/Makefile -------------------------- Patching file hadrian/bindist/Makefile using Plan A... Hunk #1 succeeded at 226 (offset 13 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_bootstrap_bootstrap.py => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_bootstrap_bootstrap.py Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-hadrian_bootstrap_bootstrap.py,v 1.1 2023/10/23 08:06:48 pho Exp $ | |bootstrap.py assumes that GHC always has a threaded RTS but our bootkits |don't. It is debatable at least as to whether this should be upstreamed. | |The hack to shake.cabal can be removed when [1] gets merged and GHC adopts |it in hadrian/bootstrap/*.json. | |[1]: https://github.com/ndmitchell/shake/pull/836 | |--- hadrian/bootstrap/bootstrap.py.orig 2023-09-21 11:30:31.000000000 +0000 |+++ hadrian/bootstrap/bootstrap.py -------------------------- Patching file hadrian/bootstrap/bootstrap.py using Plan A... Hunk #1 succeeded at 185. Hunk #2 succeeded at 501. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_bootstrap_src_Main.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_bootstrap_src_Main.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-hadrian_bootstrap_src_Main.hs,v 1.2 2023/11/02 15:29:31 pho Exp $ | |Fix build with cabal-install-parsers >= 0.6. This doesn't affect normal |builds. It's only for creating Hadrian bootstrap source archives. | |https://gitlab.haskell.org/ghc/ghc/-/merge_requests/11546 | |--- hadrian/bootstrap/src/Main.hs.orig 2023-09-21 11:30:31.000000000 +0000 |+++ hadrian/bootstrap/src/Main.hs -------------------------- Patching file hadrian/bootstrap/src/Main.hs using Plan A... Hunk #1 succeeded at 1. Hunk #2 succeeded at 87. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Builder.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Builder.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-hadrian_src_Builder.hs,v 1.2 2023/11/02 14:32:04 pho Exp $ | |Hunk #0, #2: | This is pkgsrc-specific. We build haddock as a dynamically-linked | executable but it doesn't contain an rpath to our not-yet-installed | shared libraries. This is usually not a problem because Hadrian embeds | relative rpaths (i.e. $ORIGIN) in dynamically-linked objects but we | patch Hadrian not to do that. | |Hunk #1: | | primop.txt contains Unicode characters. Don't assume the current locale | is "*.UTF-8". | | > _build/stage0/compiler/build/primops.txt: hGetContents: invalid argument (cannot decode byte sequence starting from 226) | | https://gitlab.haskell.org/ghc/ghc/-/merge_requests/11544 | |--- hadrian/src/Builder.hs.orig 2023-09-21 11:30:31.000000000 +0000 |+++ hadrian/src/Builder.hs -------------------------- Patching file hadrian/src/Builder.hs using Plan A... Hunk #1 succeeded at 31. Hunk #2 succeeded at 334. Hunk #3 succeeded at 384. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Oracles_Setting.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Oracles_Setting.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-hadrian_src_Oracles_Setting.hs,v 1.1 2023/10/23 08:06:48 pho Exp $ | |Hadrian's idea of rpath usability is overly pessimistic. If it's an ELF or |a Mach-O target we should assume it supports rpaths. | |Merge request: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/9767 | |--- hadrian/src/Oracles/Setting.hs.orig 2023-01-15 17:18:47.276886521 +0000 |+++ hadrian/src/Oracles/Setting.hs -------------------------- Patching file hadrian/src/Oracles/Setting.hs using Plan A... Hunk #1 succeeded at 274 (offset 5 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Rules_BinaryDist.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Rules_BinaryDist.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-hadrian_src_Rules_BinaryDist.hs,v 1.1 2023/10/23 08:06:48 pho Exp $ | |Hunk #0, #1, #3: | Distinguish bootstrapping bindists from regular ones. This is | pkgsrc-specific. | |Hunk #2: | Do not run `ghc-pkg recache' while creating a binary distribution. We | don't want a package cache to be installed because we do it in our | INSTALL script. Also it's not possible to run ghc-pkg in the first | place without setting LD_LIBRARY_PATH since we don't build relocatable | executables. This is pkgsrc-specific. | |Hunk #4: | We want our bootkits to be as small as possible, even though `xz -9e' | is very slow and consumes about 680 MiB of memory. This is | pkgsrc-specific. | |--- hadrian/src/Rules/BinaryDist.hs.orig 2022-12-23 16:19:02.000000000 +0000 |+++ hadrian/src/Rules/BinaryDist.hs -------------------------- Patching file hadrian/src/Rules/BinaryDist.hs using Plan A... Hunk #1 succeeded at 113. Hunk #2 succeeded at 136. Hunk #3 succeeded at 194. Hunk #4 succeeded at 270 (offset 1 line). Hunk #5 succeeded at 323 (offset 1 line). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Settings_Builders_Ghc.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Settings_Builders_Ghc.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-hadrian_src_Settings_Builders_Ghc.hs,v 1.1 2023/10/23 08:06:48 pho Exp $ | |Don't use rpaths relative to $ORIGIN. Use the final installation path. This |is pkgsrc-specific. | |--- hadrian/src/Settings/Builders/Ghc.hs.orig 2022-12-23 16:19:02.000000000 +0000 |+++ hadrian/src/Settings/Builders/Ghc.hs -------------------------- Patching file hadrian/src/Settings/Builders/Ghc.hs using Plan A... Hunk #1 succeeded at 5. Hunk #2 succeeded at 107 (offset 2 lines). Hunk #3 succeeded at 124 (offset 2 lines). Hunk #4 succeeded at 138 (offset 2 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Settings_Default.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Settings_Default.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-hadrian_src_Settings_Default.hs,v 1.1 2023/10/23 08:06:48 pho Exp $ | |Do not build haddock with the stage-0 compiler. That's not only pointless |but also impossible in our case because our bootkits in the pre-Hadrian era |lack the xhtml package, which haddock depends on. | |Probably not eligible for upstreaming though, because regular GHC |distributions of course have xhtml. | |--- hadrian/src/Settings/Default.hs.orig 2023-01-11 15:35:59.835839905 +0000 |+++ hadrian/src/Settings/Default.hs -------------------------- Patching file hadrian/src/Settings/Default.hs using Plan A... Hunk #1 succeeded at 89 (offset 3 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Settings_Packages.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-hadrian_src_Settings_Packages.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-hadrian_src_Settings_Packages.hs,v 1.2 2023/11/02 15:29:31 pho Exp $ | |Hunk #0, #1: | Don't assume we always build the threaded RTS. This can only be | upstreamed after they update its Haddock to a version containing this | change: https://github.com/haskell/haddock/pull/1575 | |Hunk #2: | Hadrian does something unholy on i386 to gain speed but it seems to be | incompatible with LLD. Disable the speed hack to work around a linkage | failure. Ideally we should do this by detecting the type of linker but | not the OS. | |--- hadrian/src/Settings/Packages.hs.orig 2023-09-21 11:30:31.000000000 +0000 |+++ hadrian/src/Settings/Packages.hs -------------------------- Patching file hadrian/src/Settings/Packages.hs using Plan A... Hunk #1 succeeded at 33. Hunk #2 succeeded at 164. Hunk #3 succeeded at 451. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_base_configure => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_base_configure Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_base_configure,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Ensure libiconv can be found at runtime. | |XXX: I don't think this is needed because rpath flags are handled by |Cabal. LDFLAGS is not used anyway. (pho@) | |--- libraries/base/configure.orig 2021-10-28 20:43:03.000000000 +0000 |+++ libraries/base/configure -------------------------- Patching file libraries/base/configure using Plan A... Hunk #1 succeeded at 4308 (offset 99 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_base_GHC_Event_KQueue.hsc => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_base_GHC_Event_KQueue.hsc Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_base_GHC_Event_KQueue.hsc,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Suppress linker warnings about compatibility syscall wrappers by using |"capi" instead of "ccall". | |--- libraries/base/GHC/Event/KQueue.hsc.orig 2019-12-26 18:05:56.276488590 +0000 |+++ libraries/base/GHC/Event/KQueue.hsc -------------------------- Patching file libraries/base/GHC/Event/KQueue.hsc using Plan A... Hunk #1 succeeded at 299 (offset 15 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_base_System_CPUTime_Posix_Times.hsc => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_base_System_CPUTime_Posix_Times.hsc Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_base_System_CPUTime_Posix_Times.hsc,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Suppress linker warnings about compatibility syscall wrappers by using |"capi" instead of "ccall". | |--- libraries/base/System/CPUTime/Posix/Times.hsc.orig 2020-01-04 12:08:39.515726533 +0000 |+++ libraries/base/System/CPUTime/Posix/Times.hsc -------------------------- Patching file libraries/base/System/CPUTime/Posix/Times.hsc using Plan A... Hunk #1 succeeded at 29 (offset 2 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_base_System_Environment_ExecutablePath.hsc => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_base_System_Environment_ExecutablePath.hsc Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_base_System_Environment_ExecutablePath.hsc,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Support executablePath on Solaris. | |Merge request: |https://gitlab.haskell.org/ghc/ghc/-/merge_requests/9768 | |--- libraries/base/System/Environment/ExecutablePath.hsc.orig 2023-01-14 16:23:19.153537486 +0000 |+++ libraries/base/System/Environment/ExecutablePath.hsc -------------------------- Patching file libraries/base/System/Environment/ExecutablePath.hsc using Plan A... Hunk #1 succeeded at 47 (offset 9 lines). Hunk #2 succeeded at 107 (offset 9 lines). Hunk #3 succeeded at 168 (offset 4 lines). Hunk #4 succeeded at 192 (offset 9 lines). Hunk #5 succeeded at 198 (offset 4 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_base_System_Environment.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_base_System_Environment.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_base_System_Environment.hs,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Suppress linker warnings about compatibility syscall wrappers by using |"capi" instead of "ccall". | |--- libraries/base/System/Environment.hs.orig 2018-02-25 20:02:28.000000000 +0000 |+++ libraries/base/System/Environment.hs -------------------------- Patching file libraries/base/System/Environment.hs using Plan A... Hunk #1 succeeded at 1. Hunk #2 succeeded at 250 (offset 3 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_Cabal_Cabal_src_Distribution_Simple_GHC.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_Cabal_Cabal_src_Distribution_Simple_GHC.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_Cabal_Cabal_src_Distribution_Simple_GHC.hs,v 1.2 2023/11/01 17:53:53 pho Exp $ | |Hunk #0, #1, #2: | Revert the change introduced in https://github.com/haskell/cabal/issues/1660 | | Do not use @rpath on Darwin, instead use the final destination path as a | library install name. This is pkgsrc-specific. | | See also https://github.com/haskell/cabal/pull/7076 | And https://github.com/haskell/cabal/issues/7339 | |Hunk #2: | Enable Cabal's rpath overriding functionality on all the platforms known | to use ELF. | | When a Cabal package which defines both a library and an executable to be | built, GHC by default embeds an rpath to the *build* directory into the | executable. This may sound like a bug in GHC but is actually not, because | GHC doesn't know the fact that the said library is going to be installed | elsewhere. To overcome this issue, Cabal has a functionality to disable | the default rpath handling of GHC and construct the correct set of | rpaths... but only on certain platforms for some reason. | | Pull request: https://github.com/haskell/cabal/pull/7382 | |--- libraries/Cabal/Cabal/src/Distribution/Simple/GHC.hs.orig 2023-09-21 11:31:40.000000000 +0000 |+++ libraries/Cabal/Cabal/src/Distribution/Simple/GHC.hs -------------------------- Patching file libraries/Cabal/Cabal/src/Distribution/Simple/GHC.hs using Plan A... Hunk #1 succeeded at 836. Hunk #2 succeeded at 900. Hunk #3 succeeded at 1812. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_ghc-boot_GHC_BaseDir.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_ghc-boot_GHC_BaseDir.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_ghc-boot_GHC_BaseDir.hs,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Make getBaseDir less platform-dependent by making use of base-4.17 |`executablePath` when possible, and falling back on `getExecutablePath` |when it's not available. | |Merge request: |https://gitlab.haskell.org/ghc/ghc/-/merge_requests/9768 | |--- libraries/ghc-boot/GHC/BaseDir.hs.orig 2023-09-21 11:30:31.000000000 +0000 |+++ libraries/ghc-boot/GHC/BaseDir.hs -------------------------- Patching file libraries/ghc-boot/GHC/BaseDir.hs using Plan A... Hunk #1 succeeded at 12. Hunk #2 succeeded at 24. Hunk #3 succeeded at 45. Hunk #4 succeeded at 73. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_process_cbits_posix_fork__exec.c => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_process_cbits_posix_fork__exec.c Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_process_cbits_posix_fork__exec.c,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Ensure _POSIX_PTHREAD_SEMANTICS is defined on SunOS. | |Pull request: |https://github.com/haskell/process/pull/274 | |--- libraries/process/cbits/posix/fork_exec.c.orig 2023-08-21 03:50:41.000000000 +0000 |+++ libraries/process/cbits/posix/fork_exec.c -------------------------- Patching file libraries/process/cbits/posix/fork_exec.c using Plan A... Hunk #1 succeeded at 1. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_terminfo_configure.ac => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_terminfo_configure.ac Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_terminfo_configure.ac,v 1.2 2023/11/02 15:29:31 pho Exp $ | |Don't use "==" as it's a Bash-ism: |https://github.com/judah/terminfo/commit/09fe6ebe02d439c507c0bad6e500c9212f56e8e3 | |--- libraries/terminfo/configure.ac.orig 2023-10-23 02:48:49.253512773 +0000 |+++ libraries/terminfo/configure.ac -------------------------- Patching file libraries/terminfo/configure.ac using Plan A... Hunk #1 succeeded at 18. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_time_lib_Data_Time_Clock_Internal_CTimespec.hsc => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_time_lib_Data_Time_Clock_Internal_CTimespec.hsc Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_time_lib_Data_Time_Clock_Internal_CTimespec.hsc,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Suppress linker warnings about compatibility symbols. | |--- libraries/time/lib/Data/Time/Clock/Internal/CTimespec.hsc.orig 2020-01-04 02:21:30.506303962 +0000 |+++ libraries/time/lib/Data/Time/Clock/Internal/CTimespec.hsc -------------------------- Patching file libraries/time/lib/Data/Time/Clock/Internal/CTimespec.hsc using Plan A... Hunk #1 succeeded at 3 (offset 2 lines). Hunk #2 succeeded at 26 (offset -4 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_time_lib_Data_Time_Clock_Internal_CTimeval.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_time_lib_Data_Time_Clock_Internal_CTimeval.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_time_lib_Data_Time_Clock_Internal_CTimeval.hs,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Suppress linker warnings about compatibility syscall wrappers by using |"capi" instead of "ccall". In Haskell FFI, "ccall" is actually an |interface to C ABI rather than C API. That is, GHC generates direct |references to the symbol even if it's actually defined as a cpp macro |or something like that, because GHC knows nothing about those macros |in foreign headers. Hence the following warnings: | | .../libHStime-1.4.0.1.a(CTimeval.o): In function `s1Kp_info': | (.text+0x2f): warning: warning: reference to compatibility | gettimeofday(); include to generate correct reference | |In other words, you can safely use "ccall" only when you are sure the |symbol you want to import is actually a symbol in the ABI sense, which |is not always the case for the POSIX API. | |--- libraries/time/lib/Data/Time/Clock/Internal/CTimeval.hs.orig 2021-10-28 20:42:10.000000000 +0000 |+++ libraries/time/lib/Data/Time/Clock/Internal/CTimeval.hs -------------------------- Patching file libraries/time/lib/Data/Time/Clock/Internal/CTimeval.hs using Plan A... Hunk #1 succeeded at 2 (offset 1 line). Hunk #2 succeeded at 23 (offset 1 line). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_Env_ByteString.hsc => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_Env_ByteString.hsc Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_unix_System_Posix_Env_ByteString.hsc,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Suppress linker warnings about compatibility symbols by using "capi" |instead of "ccall". | |--- libraries/unix/System/Posix/Env/ByteString.hsc.orig 2022-12-23 16:19:54.000000000 +0000 |+++ libraries/unix/System/Posix/Env/ByteString.hsc -------------------------- Patching file libraries/unix/System/Posix/Env/ByteString.hsc using Plan A... Hunk #1 succeeded at 153. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_Env_PosixString.hsc => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_Env_PosixString.hsc Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_unix_System_Posix_Env_PosixString.hsc,v 1.1 2023/10/23 17:00:08 pho Exp $ | |Suppress linker warnings about compatibility symbols by using "capi" |instead of "ccall". | |--- libraries/unix/System/Posix/Env/PosixString.hsc.orig 2023-10-23 10:24:23.010158238 +0000 |+++ libraries/unix/System/Posix/Env/PosixString.hsc -------------------------- Patching file libraries/unix/System/Posix/Env/PosixString.hsc using Plan A... Hunk #1 succeeded at 146. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_Env.hsc => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_Env.hsc Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_unix_System_Posix_Env.hsc,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Suppress linker warnings about compatibility symbols by using "capi" |instead of "ccall". | |--- libraries/unix/System/Posix/Env.hsc.orig 2023-09-21 11:31:41.000000000 +0000 |+++ libraries/unix/System/Posix/Env.hsc -------------------------- Patching file libraries/unix/System/Posix/Env.hsc using Plan A... Hunk #1 succeeded at 141. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_Signals.hsc => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_Signals.hsc Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_unix_System_Posix_Signals.hsc,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Suppress linker warnings about compatibility symbols. | |--- libraries/unix/System/Posix/Signals.hsc.orig 2023-09-21 11:31:41.000000000 +0000 |+++ libraries/unix/System/Posix/Signals.hsc -------------------------- Patching file libraries/unix/System/Posix/Signals.hsc using Plan A... Hunk #1 succeeded at 848. Hunk #2 succeeded at 860. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_User_ByteString.hsc => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_User_ByteString.hsc Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_unix_System_Posix_User_ByteString.hsc,v 1.1 2023/10/24 06:38:45 pho Exp $ | |Suppress linker warnings about compatibility symbols by using "capi" |instead of "ccall". | |--- libraries/unix/System/Posix/User/ByteString.hsc.orig 2023-10-24 03:45:34.747217813 +0000 |+++ libraries/unix/System/Posix/User/ByteString.hsc -------------------------- Patching file libraries/unix/System/Posix/User/ByteString.hsc using Plan A... Hunk #1 succeeded at 522. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_User.hsc => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-libraries_unix_System_Posix_User.hsc Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-libraries_unix_System_Posix_User.hsc,v 1.1 2023/10/23 17:00:08 pho Exp $ | |Suppress linker warnings about compatibility symbols by using "capi" |instead of "ccall". | |--- libraries/unix/System/Posix/User.hsc.orig 2023-10-23 10:26:54.941601450 +0000 |+++ libraries/unix/System/Posix/User.hsc -------------------------- Patching file libraries/unix/System/Posix/User.hsc using Plan A... Hunk #1 succeeded at 605. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-m4_fptools__set__haskell__platform__vars.m4 => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-m4_fptools__set__haskell__platform__vars.m4 Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-m4_fptools__set__haskell__platform__vars.m4,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Pretend the toolchain doesn't support .note.GNU-stack on NetBSD. This is a |temporary workaround and shouldn't live forever. See: |https://mail-index.netbsd.org/tech-toolchain/2023/01/15/msg004188.html | |--- m4/fptools_set_haskell_platform_vars.m4.orig 2023-01-15 14:19:38.638967614 +0000 |+++ m4/fptools_set_haskell_platform_vars.m4 -------------------------- Patching file m4/fptools_set_haskell_platform_vars.m4 using Plan A... Hunk #1 succeeded at 195 (offset 15 lines). Hunk #2 succeeded at 226 (offset 15 lines). done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-rts_linker_MMap.c => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-rts_linker_MMap.c Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-rts_linker_MMap.c,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Hunk #0-2, #6: | NetBSD-specific way of switching pages from rw- to r-x. Even when the | RTS linker is not used, libraries/ghci/GHCi/InfoTable.hsc | (fillExecBuffer) calls rts/ExecPage.c (allocateExecPage) to store some | executable code, then calls freezeExecPage() and fails. If we were to | disable PaX MPROTECT entirely then we could omit this change, but of | course that's not the right thing to do. | | Merge request: | https://gitlab.haskell.org/ghc/ghc/-/merge_requests/9769 | |Hunk #3-5: | NetBSD mmap(2) doesn't work the same way as that of Linux. We | essentially can't use the hint address to map a new page based on the | address of the previously mapped one. | | Merge request: | https://gitlab.haskell.org/ghc/ghc/-/merge_requests/9769 | | Minimal test case: | https://gist.github.com/depressed-pho/a629247b48b3e6178e35a14c62e9d44f | | tech-kern thread: | https://mail-index.netbsd.org/tech-kern/2022/02/17/msg027969.html | |--- rts/linker/MMap.c.orig 2022-12-23 16:19:02.000000000 +0000 |+++ rts/linker/MMap.c -------------------------- Patching file rts/linker/MMap.c using Plan A... Hunk #1 succeeded at 46. Hunk #2 succeeded at 198. Hunk #3 succeeded at 262. Hunk #4 succeeded at 316. Hunk #5 succeeded at 390. Hunk #6 succeeded at 425. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-rts_linker_MMap.h => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-rts_linker_MMap.h Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-rts_linker_MMap.h,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Needed for supporting NetBSD PaX MPROTECT. See patch-rts_linker_MMap.c for |details. | |--- rts/linker/MMap.h.orig 2023-01-10 12:41:09.417481027 +0000 |+++ rts/linker/MMap.h -------------------------- Patching file rts/linker/MMap.h using Plan A... Hunk #1 succeeded at 54. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-utils_genprimopcode_Main.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-utils_genprimopcode_Main.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-utils_genprimopcode_Main.hs,v 1.2 2023/11/02 14:32:04 pho Exp $ | |primops.txt contains Unicode characters. Don't assume the current locale is |"*.UTF-8". | |> LC_ALL=C ./genprimopcode --data-decl < ./primops.txt |> genprimopcode: : hGetContents: invalid argument (cannot decode byte sequence starting from 226) | |https://gitlab.haskell.org/ghc/ghc/-/merge_requests/11544 | |--- utils/genprimopcode/Main.hs.orig 2023-10-23 03:34:51.628735818 +0000 |+++ utils/genprimopcode/Main.hs -------------------------- Patching file utils/genprimopcode/Main.hs using Plan A... Hunk #1 succeeded at 13. Hunk #2 succeeded at 117. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-utils_haddock_driver_Main.hs => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-utils_haddock_driver_Main.hs Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-utils_haddock_driver_Main.hs,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Workaround for haddock dying when the locale isn't "*.UTF-8". Of course |this isn't a real fix. Haddock is very fragile when it comes to locale |setting, and fixing it properly would be seriously a lot of work. | |--- utils/haddock/driver/Main.hs.orig 2020-07-28 17:05:24.000000000 +0000 |+++ utils/haddock/driver/Main.hs -------------------------- Patching file utils/haddock/driver/Main.hs using Plan A... Hunk #1 succeeded at 1. done => Verifying /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-utils_haddock_haddock.cabal => Applying pkgsrc patch /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96/patches/patch-utils_haddock_haddock.cabal Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |$NetBSD: patch-utils_haddock_haddock.cabal,v 1.1 2023/10/23 08:06:49 pho Exp $ | |Don't assume we always build the threaded RTS. | |Pull request: |https://github.com/haskell/haddock/pull/1575 | |--- utils/haddock/haddock.cabal.orig 2023-01-11 07:37:47.265320221 +0000 |+++ utils/haddock/haddock.cabal -------------------------- Patching file utils/haddock/haddock.cabal using Plan A... Hunk #1 succeeded at 59. done ===> Regenerating configuration scripts for ghc-9.6.3nb3 configure.ac:90: warning: The macro `AC_CONFIG_HEADER' is obsolete. configure.ac:90: You should run autoupdate. ./lib/autoconf/status.m4:719: AC_CONFIG_HEADER is expanded from... configure.ac:90: the top level configure.ac:92: warning: The macro `AC_CONFIG_HEADER' is obsolete. configure.ac:92: You should run autoupdate. ./lib/autoconf/status.m4:719: AC_CONFIG_HEADER is expanded from... configure.ac:92: the top level configure.ac:488: warning: The macro `AC_PROG_CC_C99' is obsolete. configure.ac:488: You should run autoupdate. ./lib/autoconf/c.m4:1659: AC_PROG_CC_C99 is expanded from... m4/fp_set_cflags_c99.m4:7: FP_SET_CFLAGS_C99 is expanded from... configure.ac:488: the top level configure.ac:489: warning: The macro `AC_PROG_CC_C99' is obsolete. configure.ac:489: You should run autoupdate. ./lib/autoconf/c.m4:1659: AC_PROG_CC_C99 is expanded from... m4/fp_set_cflags_c99.m4:7: FP_SET_CFLAGS_C99 is expanded from... configure.ac:489: the top level configure.ac:490: warning: The macro `AC_PROG_CC_C99' is obsolete. configure.ac:490: You should run autoupdate. ./lib/autoconf/c.m4:1659: AC_PROG_CC_C99 is expanded from... m4/fp_set_cflags_c99.m4:7: FP_SET_CFLAGS_C99 is expanded from... configure.ac:490: the top level configure.ac:491: warning: The macro `AC_PROG_CC_C99' is obsolete. configure.ac:491: You should run autoupdate. ./lib/autoconf/c.m4:1659: AC_PROG_CC_C99 is expanded from... m4/fp_set_cflags_c99.m4:7: FP_SET_CFLAGS_C99 is expanded from... configure.ac:491: the top level configure.ac:1013: warning: The macro `AC_TRY_RUN' is obsolete. configure.ac:1013: You should run autoupdate. ./lib/autoconf/general.m4:2997: AC_TRY_RUN is expanded from... lib/m4sugar/m4sh.m4:692: _AS_IF_ELSE is expanded from... lib/m4sugar/m4sh.m4:699: AS_IF is expanded from... ./lib/autoconf/general.m4:2249: AC_CACHE_VAL is expanded from... ./lib/autoconf/general.m4:2270: AC_CACHE_CHECK is expanded from... m4/fp_check_timer_create.m4:5: FP_CHECK_TIMER_CREATE is expanded from... configure.ac:1013: the top level configure.ac:9: warning: The macro `AC_HELP_STRING' is obsolete. configure.ac:9: You should run autoupdate. ./lib/autoconf/general.m4:204: AC_HELP_STRING is expanded from... ./lib/autoconf/general.m4:1553: AC_ARG_WITH is expanded from... configure.ac:9: the top level ===> Creating toolchain wrappers for ghc-9.6.3nb3 ===> Configuring for ghc-9.6.3nb3 ===> Extracting bootstrapping compiler for ghc-9.6.3nb3 ===> Preparing bootstrapping compiler for ghc-9.6.3nb3 /usr/bin/bash: line 0: cd: /home/pbulk/build-disk/lang/ghc96/work/bootkit-dist/ghc-9.2.1-boot-x86_64-unknown-solaris2: No such file or directory *** Error code 1 Stop. bmake[1]: stopped in /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96 *** Error code 1 Stop. bmake: stopped in /data/jenkins/workspace/pkgsrc-trunk-x86_64/lang/ghc96