| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
| |
This replaces inline code with functions from crossdev.eclass. Now also
supports LLVM/Clang crossdev because the eclass handles that
automatically.
Thanks wikky for help!
Signed-off-by: Alfred Persson Forsberg <cat@catcream.org>
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
| |
As strip-flags no longer removes -flto then a filter-lto is now
required to prevent bug 877343 from reopening, like has already
been applied to sys-libs/glibc.
Closes: https://bugs.gentoo.org/877343
Closes: https://github.com/gentoo/gentoo/pull/30942
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
| |
Signed-off-by: Mike Gilbert <floppym@gentoo.org>
|
|
|
|
| |
Signed-off-by: Mike Gilbert <floppym@gentoo.org>
|
|
|
|
|
|
|
|
|
| |
To create a symlink relative to /lib/ld-musl.so, we must know if ${ROOT}
will have /lib symlinked to /usr/lib. We rely on the split-usr USE flag to
determine this.
Closes: https://bugs.gentoo.org/903703
Signed-off-by: Mike Gilbert <floppym@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Given that musl has long added an `include` directive for this directory to
ld.so.conf, we believe not creating the directory was an accidental omission.
gcc-config has special code to handle the case where it doesn't exist, but
eselect modules like blas do not, and subsequently break.
I have tested this, and both gcc-config and eselect-blas now behave as expected.
There is no migration to do. The tools simply start using the directory once it
is available.
Signed-off-by: James Le Cuirot <chewi@gentoo.org>
|
|
|
|
|
|
|
|
| |
Needed to avoid SSP hacks for cross (if we don't install it, we have to
deal with conditionally patching GCC).
Bug: https://bugs.gentoo.org/862831
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
| |
Found reports of systemwide issues when Musl is compiled with -Ofast so this is
intended to at least save some users from one issue using Ofast systemwide.
Thanks-to: Krusin
Signed-off-by: Ian Jordan <immoloism@gmail.com>
Closes: https://github.com/gentoo/gentoo/pull/28884
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
| |
Anything actually linking against libcrypt should depend
on the virtual anyway, and this means we can have
libxcrypt properly depend on virtual/libc (which is needed
to ensure it can be installed first).
Bug: https://bugs.gentoo.org/877567
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
After discussing my bug with dalias on #musl I was informed that LTO has a negative
impact on Musl so fixing the issue I highlight rather than filtering wouldn't be the
desired outcome for the user.
This will also help crossdev users as a bonus as -flto applies to the
crosstoolchain and can be hard to spot at first.
Chat log to comfirm:
immolo: I ran into the -flto issue while testing musl built with lto and the only thing I can find on it was a workaround in 2015.
immolo: My question though is it even worth having lto enabled on musl in the first place as every distro just seems to filter it so
there must be a reason
dalias: not only is it not worth it. it's negative value
dalias: makes musl larger and slower
dalias: if you wanted libc.a to get LTO'd into static programs, that might have some value
dalias: but LTO on libc.so is worse than worthless
Closes: https://bugs.gentoo.org/877343
Thanks-to: dalias
Signed-off-by: Ian Jordan <immoloism@gmail.com>
Closes: https://github.com/gentoo/gentoo/pull/27824
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
| |
While musl doesn't split out libcrypt.so, it does
apparently for libcrypt.a.
Closes: https://bugs.gentoo.org/871726
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
vapier's comment is really it:
""'
> crt files are often hand written assembly, and for the init/fini ones,
> barely any code at all. so they tend not to have anything useful to add
> debugging wise, if the tools even support it in the first place.
>
> not really worth trying to add debugging info to really any of the crt
> files. we should figure out how to suppress the warning for them. or
> ignore it.
"""
Closes: https://bugs.gentoo.org/830213
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
| |
We want crypt on by default for this as sys-libs/libxcrypt isn't (yet?)
built as part as crossdev. Also, elide the blockers when in cross-*,
as it doesn't make sense to block the normal CBUILD libxcrypt at all
there when we're installing into /usr/${CHOST} anyway.
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
| |
We only need libxcrypt when we DON'T install crypt.h
Bug: https://bugs.gentoo.org/867991
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add USE=crypt to allow enabling/disabling crypt.h installation
(to allow sys-libs/libxcrypt[system] usage). Many things are starting
to want functions from libxcrypt itself (additional functions and
algorithms).
musl isn't removing crypt.h (and the relevant functions from its libc),
unlike glibc, but we need to allow disabling the installation of crypt.h
to allow libxcrypt[system] usage (which provides crypt.h & libcrypt.so
instead, with more functionality).
This brings musl in line with the changes we made for glibc a while
ago. The situation with glibc is a bit different because the migration
is mandatory there, while we're just strongly recommending it for musl
because sys-libs/libxcrypt[-system] causes headaches (see linked
PAM bug for an example, but I've also hit a similar issue with Python
yesterday).
Bug: https://bugs.gentoo.org/867991
Bug: https://bugs.gentoo.org/699422
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
| |
With EAPI 7+, ROOT and others will be blank (not /).
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
| |
I get an IPC error on verification in cross.
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
| |
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
| |
Tested with crossdev and a qemu-user lp64d chroot
Package-Manager: Portage-3.0.30, Repoman-3.0.3
Signed-off-by: Andreas K. Hüttel <dilfridge@gentoo.org>
|
|
|
|
|
|
|
|
|
| |
Added necessary framework to get this libc to run properly under a prefix.
Closes: https://bugs.gentoo.org/833192
Signed-off-by: Esteve Varela Colominas <esteve.varela@gmail.com>
Closes: https://github.com/gentoo/gentoo/pull/24169
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
| |
Signed-off-by: Sam James <sam@gentoo.org>
Closes: https://github.com/gentoo/gentoo/pull/22984
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
| |
Move around some declarations so we get sensible default
values but only perform the sanity checks _after_ we
parse command line arguments to take into account e.g.
-r (root).
Bug: https://bugs.gentoo.org/666246
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
| |
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
| |
Need to ensure the rest of the ebuild is Prefix ready anyway
but there's no reason to make it worse with stuff we
recently added, so let's use ${ED} there.
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The build system seems to create an absolute symlink to libc.so on
the host which may not exist.
If it doesn't (to avoid being disruptive, we could do this unconditionally),
create a new one relative within ${D} to facilitate SYSROOT installs.
I've hit this a few times when using crossdev but finally dug into it
a bit more.
Bug: https://bugs.gentoo.org/732482
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
| |
As per http://www.musl-libc.org/, "musl has moved to a new domain:
musl.libc.org"
Bug: https://bugs.gentoo.org/794484
Signed-off-by: Bertrand Jacquin <bertrand@jacquin.bzh>
Package-Manager: Portage-3.0.18, Repoman-3.0.2
Closes: https://github.com/gentoo/gentoo/pull/21138
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
|
|
|
|
| |
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.14, Repoman-3.0.2
Signed-off-by: Jory Pratt <anarchy@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.13, Repoman-3.0.2
Signed-off-by: Jory Pratt <anarchy@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.13, Repoman-3.0.2
Signed-off-by: Jory Pratt <anarchy@gentoo.org>
|
|
|
|
| |
Signed-off-by: Mike Gilbert <floppym@gentoo.org>
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/764755
Signed-off-by: Mike Gilbert <floppym@gentoo.org>
|
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/755071
Package-Manager: Portage-3.0.9, Repoman-3.0.2
Signed-off-by: Fabian Groffen <grobian@gentoo.org>
|
|
|
|
|
|
|
| |
Reported-by: Laine Gholson
Closes: https://bugs.gentoo.org/595950
Package-Manager: Portage-2.3.100, Repoman-2.3.22
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
|
|
|
|
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/642612
Closes: https://bugs.gentoo.org/645626
Package-Manager: Portage-2.3.99, Repoman-2.3.22
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
Signed-off-by: Jory Pratt <anarchy@gentoo.org>
|
|
|
|
|
|
| |
Bug: https://bugs.gentoo.org/642784
Package-Manager: Portage-2.3.87, Repoman-2.3.20
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Since all ebuilds in the tree are EAPI>=4,
`|| die` on builtin commands is redundant
and dead code.
Closes: https://github.com/gentoo/gentoo/pull/13940
Reviewed-by: Ulrich Müller <ulm@gentoo.org>
Reviewed-by: Michał Górny <mgorny@gentoo.org>
Reviewed-by: Andreas K. Hüttel <dilfridge@gentoo.org>
Signed-off-by: David Seifert <soap@gentoo.org>
|
|
|
|
|
|
| |
This reverts commit d3b662a5b4d652def877bc1ee55086182ee1c7c6.
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.69, Repoman-2.3.16
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
|
|
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/667126
Signed-off-by: Steffen Kuhn <nielson2@yandex.com>
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
|
|
|
|
|
| |
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
Package-Manager: Portage-2.3.51, Repoman-2.3.11
|
|
|
|
|
| |
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
Package-Manager: Portage-2.3.51, Repoman-2.3.11
|
|
|
|
| |
Package-Manager: Portage-2.3.40, Repoman-2.3.9
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
CROSSCOMPILE_OPTS is a USE_EXPAND of a single item: headers-only.
Convert it to a global USE flag instead.
Mechanical ebuild rename done as:
$ sed -e 's@crosscompile_opts_headers-only@headers-only@g' \
-i $(git grep -l headers-only)
'headers-only' flag is used by crossdev to bootstrap stage1 compiler
before libc is available.
crossdev switched to USE=headers-only in =sys-devel/crossdev-20171230.
Bug: https://bugs.gentoo.org/642712
Package-Manager: Portage-2.3.19, Repoman-2.3.6
|
|
|
|
|
|
| |
This reverts commit 0fcf2e327095c1329a4044950cadfda036db8245.
More testing is required before we commit to this.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently building a musl libc toolchain inside of gentoo prefix with
crossdev will fail, due to installing files outside of the prefix. Added
${EPREFIX} and ${ED} where apropriate fixed this issue.
Tested in a prefix with toolchain x86_64-gentoo-linux-musl, and tested on bare
gentoo with x86_64-gentoo-linux-musl.
Acked-by: blueness
Package-Manager: Portage-2.3.19, Repoman-2.3.6
Signed-off-by: Marty E. Plummer <hanetzer@protonmail.com>
Closes: https://bugs.gentoo.org/642612
Closes: https://github.com/gentoo/gentoo/pull/6678
|
|
|
|
| |
Package-Manager: Portage-2.3.8, Repoman-2.3.3
|
| |
|
|
|
|
| |
Signed-off-by: Robin H. Johnson <robbat2@gentoo.org>
|
|
|
|
| |
Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
|