As the comment says, the value is only used to get different distfile
names.
I will build a new bootstrap for amd64. For aarch64, it would be
nice if someone else could do it.
has been set to <empty> on py3 ports since Feb 2021, since when there
has been enough churn that existing PLIST entries etc referencing it
are not trustworthy. no change to python-module(5) needed as it wasn't
documented there anyway.
* Fix ghc and cabal-install version numbers in the cabal module.
* Remove i386 as it has long being gone.
* Remove unused since switching to hadrian MAKE_FLAGS.
The major effort bootstrapping the cross-compiler by Habib Alamin
ha.alamin at gmail. Thanks a lot!
OK jturner@ kili@
Tested by jturner@, phessler@, tb@
to check for working avx512 (the cbits patch uses xgetbv before
checking cpuid, so the test results in SIGILL on old machines)
reported/tested by James Cook
on machines with AVX512 on OS which don't support it yet. Problem seen by
Evan Silberman and ajacoutot, fix for libraries/text/cbits/measure_off.c
from Evan. ok gnezdo@
The massive change is due to upstream switching to a new build system
called hadrian. It needs to be bootrapped separately and its sources
are kept as a bootstrap bundle.
Documentation is no longer deemed worth packaging due to space and
time overhead compared to its utility.
Most of the work happened at h2k23, so this is a belated thank you
to University of Coimbra and the portuguese usual suspects.
OK kili@
fixing a build failure with sphinx >= 7 reported by tb@
It's a git binary patch which neither patch nor gpatch can handle, so sadly
a BDEP on git (for git-apply) is required, but at least it can be confined
to a single Makefile chunk, and the relevant files in ghc haven't been
touched in years, so it's easy to remove again when needed.
It's broken since OpenBSD-6.3, nothing in the ports tree is or was
using it.
When necessary, I could easily write a separate "library" interfacing
pledge(2) and unveil(2) that then should be hosted on hackage.haskell.org.
ok gnezdo@, kn@, sthen@
Some kind of info tables are embedded into text making the build
process fail in freshly built ghc-cabal. The generated code reads from
ZCMain_main_info.
OK tb, kili
Mostly mechanical, except for having to run ./boot per
https://gitlab.haskell.org/ghc/ghc/-/issues/21626
Removed the %n patches merged into the release. The ./configure patch
is no longer needed and wouldn't apply due to #21626 above.
OK kili@
Our recipe for rebuilding bootstrap binaries produces binaries prone
to SEGV. The platform has limited memory and was already failing to
build haddocks for ghc.
Consolidated PLIST back into a single file.
ok kili@
gnezdo