+++ This bug was initially created as a clone of Bug #1206852 +++ Latest upstream release: 7.10.1 Current version/release in rawhide: 7.8.4-42.2.fc23 URL: http://haskell.org/ghc/ Please consult the package updates policy before you issue an update to a stable branch: https://fedoraproject.org/wiki/Updates_Policy More information about the service that created this bug can be found at: https://fedoraproject.org/wiki/Upstream_release_monitoring Please keep in mind that with any upstream change, there may also be packaging changes that need to be made. Specifically, please remember that it is your responsibility to review the new version to ensure that the licensing is still correct and that no non-free or legally problematic items have been added upstream. --- Additional comment from Upstream Release Monitoring on 2015-03-29 09:34:08 JST --- Scratch build failed http://koji.fedoraproject.org/koji/taskinfo?taskID=9358309 --- Additional comment from Jens Petersen on 2015-04-03 15:43:13 JST --- See https://github.com/fedora-haskell/ghc for packaging. It should reach copr soon... --- Additional comment from Jens Petersen on 2015-05-15 16:08:55 JST --- There is a copr repo now: https://copr.fedoraproject.org/coprs/petersen/ghc-7.10.1 (Also 7.10.2 snapshot build: https://copr.fedoraproject.org/coprs/petersen/ghc-7.10.2 ) --- Additional comment from Jens Petersen on 2015-05-20 16:19:19 JST --- 7.10 needs 3.5.0 < llvm < 3.6 to build (3.5.0 is buggy) on armv7. --- Additional comment from Jens Petersen on 2015-05-20 16:20:46 JST --- (tested on armv7hl by adding + --with-llc=%{_bindir}/llc --with-opt=%{_bindir}/opt \ to configure options in ghc.spec) --- Additional comment from Peter Robinson on 2015-05-20 18:35:16 JST --- (In reply to Jens Petersen from comment #4) > 7.10 needs 3.5.0 < llvm < 3.6 to build (3.5.0 is buggy) on armv7. So there's 3.4 built is there not? Let me know when you want a side tag to build it in rawhide. --- Additional comment from Jens Petersen on 2015-05-21 09:51:16 JST --- 7.10.1.20150511 quick build seems to succeed on F22 ppc64, ppc64le, s390, and s390x. --- Additional comment from Jens Petersen on 2015-05-21 09:52:38 JST --- (In reply to Peter Robinson from comment #6) > (In reply to Jens Petersen from comment #4) > > 7.10 needs 3.5.0 < llvm < 3.6 to build (3.5.0 is buggy) on armv7. > > So there's 3.4 built is there not? Yes, but llvm-3.4 is too old for ghc-7.10. I am planning to do a llvm35 package for F23 for this. --- Additional comment from Jens Petersen on 2015-05-21 14:57:12 JST --- So to summarize the current problems are on ARM: armv7: needs llvm-3.5.[12] aarch64: ld.gold makes configure fail (bug 1215546) (perhaps can patch to use ld.bfd - though that has other problems) - there may be more problems lurking behind that... Though 7.10.1 seems to build ok on aarch64 (like 7.8.4). --- Additional comment from Jens Petersen on 2015-05-22 10:27:51 JST --- For the record: # 7.10.1 and 7.10.1.20150511 fail on f22 armv7 with llvm34 like this: "inplace/bin/ghc-stage1" -optc-fno-stack-protector -optc-Wall -optc-Wextra -optc-Wstrict-prototypes -optc-Wmissing-prototypes -optc-Wmissing-declarations -optc-Winline -optc-Waggregate-return -optc-Wpointer-arith -optc-Wmissing-noreturn -optc-Wnested-externs -optc-Wredundant-decls -optc-Iincludes -optc-Iincludes/dist -optc-Iincludes/dist-derivedconstants/header -optc-Iincludes/dist-ghcconstants/header -optc-Irts -optc-Irts/dist/build -optc-DCOMPILING_RTS -optc-DUSE_LIBFFI_FOR_ADJUSTORS -optc-fno-strict-aliasing -optc-fno-common -optc-Irts/dist/build/autogen -optc-O2 -optc-fomit-frame-pointer -optc-g -optc-DDYNAMIC -optc-DRtsWay=\"rts_dyn\" -fPIC -dynamic -H32m -O -Iincludes -Iincludes/dist -Iincludes/dist-derivedconstants/header -Iincludes/dist-ghcconstants/header -Irts -Irts/dist/build -DCOMPILING_RTS -this-package-key rts -dcmm-lint -i -irts -irts/dist/build -irts/dist/build/autogen -Irts/dist/build -Irts/dist/build/autogen -O2 -c rts/ProfHeap.c -o rts/dist/build/ProfHeap.dyn_o /tmp/ghc575_0/ghc575_6.s: Assembler messages: /tmp/ghc575_0/ghc575_6.s:1226:0: Error: can't resolve `.rodata' {.rodata section} - `stg_DEAD_WEAK_info$def' {.text section} /tmp/ghc575_0/ghc575_6.s:1235:0: Error: can't resolve `.rodata' {.rodata section} - `stg_C_FINALIZER_LIST_info$def' {.text section} /tmp/ghc575_0/ghc575_6.s:1244:0: Error: can't resolve `.rodata' {.rodata section} - `stg_NO_FINALIZER_info$def' {.text section} /tmp/ghc575_0/ghc575_6.s:1333:0: Error: can't resolve `.rodata' {.rodata section} - `stg_END_STM_WATCH_QUEUE_info$def' {.text section} # 7.10.1.20150511 fails on f23 armv7 with llvm-3.6 like this: "inplace/bin/ghc-stage1" -static -H32m -O -Iincludes -Iincludes/dist -Iincludes/dist-derivedconstants/header -Iincludes/dist-ghcconstants/header -Irts -Irts/dist/build -DCOMPILING_RTS -this-package-key rts -dcmm-lint -i -irts -irts/dist/build -irts/dist/build/autogen -Irts/dist/build -Irts/dist/build/autogen -O2 -c rts/Updates.cmm -o rts/dist/build/Updates.o You are using a new version of LLVM that hasn't been tested yet! We will try though... /usr/bin/opt: /tmp/ghc30791_0/ghc30791_2.ll:7:6: error: unexpected type in metadata definition !0 = metadata !{metadata !"top", i8* null} ^ rts/ghc.mk:236: recipe for target 'rts/dist/build/StgStartup.o' failed # 7.10.1.20150511 fails on f22 armv7 with llvm-3.5.0 like this: checking /usr/bin/opt is version 3.5... yes checking whether bootstrap compiler is affected by bug 9439... no checking if llvm version is affected by bug 9920... yes configure: error: in `/builddir/build/BUILD/ghc-7.10.1.20150511': configure: error: Cannot compile for ARM with /usr/bin/llc. See GHC trac ticket #9920. See `config.log' for more details --- Additional comment from Upstream Release Monitoring on 2015-07-30 09:20:52 JST --- Latest upstream release: 7.10.2 Current version/release in rawhide: 7.8.4-46.fc23 URL: http://haskell.org/ghc/ Please consult the package updates policy before you issue an update to a stable branch: https://fedoraproject.org/wiki/Updates_Policy More information about the service that created this bug can be found at: https://fedoraproject.org/wiki/Upstream_release_monitoring Please keep in mind that with any upstream change, there may also be packaging changes that need to be made. Specifically, please remember that it is your responsibility to review the new version to ensure that the licensing is still correct and that no non-free or legally problematic items have been added upstream. --- Additional comment from Upstream Release Monitoring on 2015-07-30 09:22:42 JST --- Failed to kick off scratch build. cmd: sha256sum /var/tmp/thn-i_AEDc/100.0% return code: 1 stdout: stderr: sha256sum: /var/tmp/thn-i_AEDc/100.0%: No such file or directory --- Additional comment from Peter Robinson on 2015-09-17 20:58:19 JST --- Let me know when you want a side tag for building --- Additional comment from Upstream Release Monitoring on 2016-01-05 09:18:19 JST --- Latest upstream release: 7.10.3 Current version/release in rawhide: 7.8.4-46.fc23 URL: http://haskell.org/ghc/ Please consult the package updates policy before you issue an update to a stable branch: https://fedoraproject.org/wiki/Updates_Policy More information about the service that created this bug can be found at: https://fedoraproject.org/wiki/Upstream_release_monitoring Please keep in mind that with any upstream change, there may also be packaging changes that need to be made. Specifically, please remember that it is your responsibility to review the new version to ensure that the licensing is still correct and that no non-free or legally problematic items have been added upstream. --- Additional comment from Upstream Release Monitoring on 2016-01-05 09:19:42 JST --- --- Additional comment from Upstream Release Monitoring on 2016-01-05 09:21:39 JST --- Scratch build failed http://koji.fedoraproject.org/koji/taskinfo?taskID=12409763
The Copr repo is https://copr.fedoraproject.org/coprs/petersen/ghc-7.10.3
https://petersen.fedorapeople.org/copr/ghc-7.10.3-51.fc23.src.rpm built okay in Koji F25 modulo ghc-rpm-macros for x86_64. http://koji.fedoraproject.org/koji/taskinfo?taskID=14257799
armv7 build: http://koji.fedoraproject.org/koji/taskinfo?taskID=14257858
aarch64 scratch f25 ghc-7.10.3-51.fc23.src.rpm Task info: http://arm.koji.fedoraproject.org/koji/taskinfo?taskID=3566918 ppc-koji build --scratch f25 ~/copr/ghc/ghc-7.10.3-51.fc23.src.rpm Task info: http://ppc.koji.fedoraproject.org/koji/taskinfo?taskID=3411104 failed to start ppc-koji build --scratch f24 ~/copr/ghc/ghc-7.10.3-51.fc23.src.rpm Task info: http://ppc.koji.fedoraproject.org/koji/taskinfo?taskID=3411107 s390-koji build --scratch f24 ~/copr/ghc/ghc-7.10.3-51.fc23.src.rpm Task info: http://s390.koji.fedoraproject.org/koji/taskinfo?taskID=2244312
(In reply to Jens Petersen from comment #3) > armv7 build: http://koji.fedoraproject.org/koji/taskinfo?taskID=14257858 This failed :-( with: "inplace/bin/ghc-stage1" -optc-marm -optc-fno-stack-protector -optc-Wall -optc-Wextra -optc-Wstrict-prototypes -optc-Wmissing-prototypes -optc-Wmissing-declarations -optc-Winline -optc-Waggregate-return -optc-Wpointer-arith -optc-Wmissing-noreturn -optc-Wnested-externs -optc-Wredundant-decls -optc-Iincludes -optc-Iincludes/dist -optc-Iincludes/dist-derivedconstants/header -optc-Iincludes/dist-ghcconstants/header -optc-Irts -optc-Irts/dist/build -optc-DCOMPILING_RTS -optc-DUSE_LIBFFI_FOR_ADJUSTORS -optc-fno-strict-aliasing -optc-fno-common -optc-Irts/dist/build/autogen -optc-O2 -optc-fomit-frame-pointer -optc-g -optc-DDYNAMIC -optc-DRtsWay=\"rts_dyn\" -fPIC -dynamic -H32m -O -Iincludes -Iincludes/dist -Iincludes/dist-derivedconstants/header -Iincludes/dist-ghcconstants/header -Irts -Irts/dist/build -DCOMPILING_RTS -this-package-key rts -dcmm-lint -i -irts -irts/dist/build -irts/dist/build/autogen -Irts/dist/build -Irts/dist/build/autogen -O2 -c rts/Arena.c -o rts/dist/build/Arena.dyn_o /tmp/ghc31776_0/ghc_7.s: Assembler messages: /tmp/ghc31776_0/ghc_7.s:1226:0: Error: can't resolve `.rodata' {.rodata section} - `stg_DEAD_WEAK_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1235:0: Error: can't resolve `.rodata' {.rodata section} - `stg_C_FINALIZER_LIST_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1244:0: Error: can't resolve `.rodata' {.rodata section} - `stg_NO_FINALIZER_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1333:0: Error: can't resolve `.rodata' {.rodata section} - `stg_END_STM_WATCH_QUEUE_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1342:0: Error: can't resolve `.rodata' {.rodata section} - `stg_END_INVARIANT_CHECK_QUEUE_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1351:0: Error: can't resolve `.rodata' {.rodata section} - `stg_END_STM_CHUNK_LIST_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1360:0: Error: can't resolve `.rodata' {.rodata section} - `stg_NO_TREC_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1369:0: Error: can't resolve `.rodata' {.rodata section} - `stg_MSG_TRY_WAKEUP_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1378:0: Error: can't resolve `.rodata' {.rodata section} - `stg_MSG_THROWTO_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1387:0: Error: can't resolve `.rodata' {.rodata section} - `stg_MSG_BLACKHOLE_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1396:0: Error: can't resolve `.rodata' {.rodata section} - `stg_MSG_NULL_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1405:0: Error: can't resolve `.rodata' {.rodata section} - `stg_END_TSO_QUEUE_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1414:0: Error: can't resolve `.rodata' {.rodata section} - `stg_GCD_CAF_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1423:0: Error: can't resolve `.rodata' {.rodata section} - `stg_STM_AWOKEN_info$def' {.text section} /tmp/ghc31776_0/ghc_7.s:1528:0: Error: can't resolve `.rodata' {.rodata section} - `stg_MVAR_TSO_QUEUE_info$def' {.text section} rts/ghc.mk:236: recipe for target 'rts/dist/build/StgMiscClosures.o' failed make[1]: *** [rts/dist/build/StgMiscClosures.o] Error 1 make[1]: *** Waiting for unfinished jobs.... Makefile:71: recipe for target 'all' failed make: *** [all] Error 2 error: Bad exit status from /var/tmp/rpm-tmp.6sDPUf (%build) and i have no idea...
(In reply to Jens Petersen from comment #4) > aarch64 scratch f25 ghc-7.10.3-51.fc23.src.rpm > Task info: http://arm.koji.fedoraproject.org/koji/taskinfo?taskID=3566918 aarch64 finished > ppc-koji build --scratch f24 ~/copr/ghc/ghc-7.10.3-51.fc23.src.rpm > Task info: http://ppc.koji.fedoraproject.org/koji/taskinfo?taskID=3411107 ppc64 and ppc64le finished > s390-koji build --scratch f24 ~/copr/ghc/ghc-7.10.3-51.fc23.src.rpm > Task info: http://s390.koji.fedoraproject.org/koji/taskinfo?taskID=2244312 s390 finished (but apparently it is gone for f25) Here is a s390x build: s390-koji build --scratch f25 ~/copr/ghc/ghc-7.10.3-51.fc23.src.rpm Task info: http://s390.koji.fedoraproject.org/koji/taskinfo?taskID=2244451
s390x failed though on f25: "/usr/bin/ghc" -o ghc/stage1/build/tmp/ghc-stage1 -hisuf hi -osuf o -hcsuf hc -static -H32m -O -package-db libraries/bootstrapping.conf -hide-all-packages -i -ighc/. -ighc/stage1/build -ighc/stage1/build/autogen -Ighc/stage1/build -Ighc/stage1/build/autogen -optP-include -optPghc/stage1/build/autogen/cabal_macros.h -package array-0.5.0.0 -package base-4.7.0.2 -package bytestring-0.10.4.0 -package directory-1.2.1.0 -package filepath-1.3.0.2 -package ghc-7.10.3 -package process-1.2.0.0 -package unix-2.7.0.1 -Wall -XHaskell2010 -no-user-package-db -rtsopts -odir ghc/stage1/build -hidir ghc/stage1/build -stubdir ghc/stage1/build -optl-L'/builddir/build/BUILD/ghc-7.10.3/compiler/stage1/build' -optl-L'/builddir/build/BUILD/ghc-7.10.3/libraries/transformers/dist-boot/build' -optl-L'/usr/lib64/ghc-7.8.4/process-1.2.0.0' -optl-L'/builddir/build/BUILD/ghc-7.10.3/libraries/hpc/dist-boot/build' -optl-L'/builddir/build/BUILD/ghc-7.10.3/libraries/hoopl/dist-boot/build' -optl-L'/builddir/build/BUILD/ghc-7.10.3/libraries/bin-package-db/dist-boot/build' -optl-L'/usr/lib64/ghc-7.8.4/directory-1.2.1.0' -optl-L'/usr/lib64/ghc-7.8.4/unix-2.7.0.1' -optl-L'/usr/lib64/ghc-7.8.4/time-1.4.2' -optl-L'/usr/lib64/ghc-7.8.4/old-locale-1.0.0.6' -optl-L'/usr/lib64/ghc-7.8.4/filepath-1.3.0.2' -optl-L'/builddir/build/BUILD/ghc-7.10.3/libraries/binary/dist-boot/build' -optl-L'/usr/lib64/ghc-7.8.4/containers-0.5.5.1' -optl-L'/usr/lib64/ghc-7.8.4/bytestring-0.10.4.0' -optl-L'/usr/lib64/ghc-7.8.4/deepseq-1.3.0.2' -optl-L'/usr/lib64/ghc-7.8.4/array-0.5.0.0' -optl-L'/usr/lib64/ghc-7.8.4/base-4.7.0.2' -optl-L'/usr/lib64/ghc-7.8.4/integer-gmp-0.5.1.0' -optl-L'/usr/lib64/ghc-7.8.4/ghc-prim-0.3.1.0' -optl-L'/usr/lib64/ghc-7.8.4/rts-1.0' -optl-lrt -optl-lutil -optl-ldl -optl-lpthread -optl-lgmp -optl-lm -optl-lrt -optl-ldl -optl-lffi -static -H32m -O -package-db libraries/bootstrapping.conf -hide-all-packages -i -ighc/. -ighc/stage1/build -ighc/stage1/build/autogen -Ighc/stage1/build -Ighc/stage1/build/autogen -optP-include -optPghc/stage1/build/autogen/cabal_macros.h -package array-0.5.0.0 -package base-4.7.0.2 -package bytestring-0.10.4.0 -package directory-1.2.1.0 -package filepath-1.3.0.2 -package ghc-7.10.3 -package process-1.2.0.0 -package unix-2.7.0.1 -Wall -XHaskell2010 -no-user-package-db -rtsopts ghc/stage1/build/Main.o ghc/stage1/build/hschooks.o collect2: error: ld returned 1 exit status ghc/ghc.mk:114: recipe for target 'ghc/stage1/build/tmp/ghc-stage1' failed make[1]: *** [ghc/stage1/build/tmp/ghc-stage1] Error 1 Makefile:71: recipe for target 'all' failed make: *** [all] Error 2 error: Bad exit status from /var/tmp/rpm-tmp.lVjWG5 (%build)
$ s390-koji build --scratch f24 ~/copr/ghc/ghc-7.10.3-51.fc23.src.rpm --arch-override=s390x Task info: http://s390.koji.fedoraproject.org/koji/taskinfo?taskID=2244461
(In reply to Jens Petersen from comment #8) > $ s390-koji build --scratch f24 ~/copr/ghc/ghc-7.10.3-51.fc23.src.rpm > --arch-override=s390x > Task info: http://s390.koji.fedoraproject.org/koji/taskinfo?taskID=2244461 this completed So right now the only problem is really armv7hl. I'll see if any of Debian's patches help...
(In reply to Jens Petersen from comment #5) > (In reply to Jens Petersen from comment #3) > > armv7 build: http://koji.fedoraproject.org/koji/taskinfo?taskID=14257858 > > This failed :-( with: : Debian's armel-revert-ghci-fixes didn't help: same error on F25 again. http://koji.fedoraproject.org/koji/taskinfo?taskID=14309475 I will try F24 to see if this makes a difference: http://koji.fedoraproject.org/koji/taskinfo?taskID=14310652
I realized the earlier 7.10.3 builds didn't have llvm35 setup correctly... Once that is done it does indeed build okay on F24 armv7hl too. http://koji.fedoraproject.org/koji/taskinfo?taskID=14323168 I will request a f25-ghc sidetag soon in Koji from Releng.
confirmed f25 armv7hl also okay (and debian armel-ghci-revert-fix patch not needed). http://koji.fedoraproject.org/koji/taskinfo?taskID=14327369 The f25-ghc buildroot is ready. Next will merge the ghc-rpm-macros and ghc into Fedora pkg git master.
ghc-7.10.3 is in f25-ghc now.