| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
| |
These betas are usually good enough for ~arch, but let's be
conservative and wait until next non-beta version to keyword
(users should feel free to opt-in now if they want to).
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
| |
Those using legacy branches are advised to stay on 6.1 LTS or older.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
| |
Early'ish but 525.85 seems uneventful, so might as well do this
now since skipped 525.78 while the stable 525.60 is getting old.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Seem to be mostly bugfixes so this will likely be the next
stable candidate perhaps in shortened 10-15 days if there's
no obvious issues.
Also removes the need for the clang15 patch, nvidia now
seems to explicitly pass -Wstrict-protypes on all these
packages (including e.g. nvidia-settings) and fixed these.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
| |
First stable for 0/525 branch, production branch with no known major
issues except at most backlight issues on some thinkpads (but don't
want to hold back use of newer vulkan extensions over this, please
use older branches if affected for now).
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
| |
linux-mod_src_compile normally does not use *FLAGS, but the
kernel blob is a different story and it fails quite easily.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Should "hopefully" cover most use cases without regressions.
Makes some degree of sense to do this on (at least) nvidia-drivers,
given it's widely used and compiles a lot of non-modules things and
users may not want to use the same toolchain for these.
e.g. can now do
clang+lld+thinlto kernel + gcc+bfd nvidia
or
gcc+bfd kernel + clang+lld+thinlto nvidia
Not that going to support every custom mixing possible, and please
consider clang kernel an unsupported configuration despite this
(hardly tested).
Could be interesting to have linux-mod handle this, but that may
be better to consider for EAPI=9 so it doesn't unexpectedly break
anything (meanwhile it's simple to use a package.env for non-nvidia).
Closes: https://bugs.gentoo.org/888587
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Seen a user miss the warning and wondering why it didn't just
die for attention. Was formerly worried it may die when it
shouldn't, but the exact string check makes that unlikely.
Fairly rare that this is an issue nowadays though, GCC_PLUGINS
are typically not actually in-use and so errors don't happen
(check won't trigger then either).
Test isn't perfect, doesn't consider some configurations
(just informational, so not very important), but still fix
with -native-symlinks while here.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When using lto, prevents revdeps from getting e.g.
ld.lld: error: undefined symbol: XNVCTRLQueryTargetAttribute
Having these may also help other issues like mismatches.
Skipping 0/390 given doesn't support XNVCTRL_CFLAGS, and 390 is
hardly supported at this point (EOL).
Not worth a revbump, will propagate as people upgrade their kernels.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
| |
Attempting to strip doesn't mangle them, but this is to avoid:
x86_64-pc-linux-gnu-strip: Unable to recognise the format ...
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
| |
New major branch for vulkan, and as usual little sense in keeping
the old for 0/vulkan as it won't be getting new extensions anymore.
If can't use, revert to non-0/vulkan.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
| |
All branches build fine with it, tested runtime a bit with 525
and "seems" fine at least for it too (not that I really use new
kernel branches until .5+ and gentoo-kernel bump).
For the now-unsupported 390 and if unpatched, that means planned
LTS 6.1 will likely be the kernel to stay on (forever) when 5.15
reaches end of life.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
| |
No real impact, just to have less unneeded differences.
wrt 510 there's technically dbus.conf, but realistically nobody
uses nvidia-powerd with older drivers and does not need a revbump
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
| |
Oops, but doesn't impact anything.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Same as nvidia-settings-390.144-raw-ldflags.patch, but applied
to kernel-module-source. Not a perfect fix given profiles don't
define LDFLAGS_arm64, but that's not a high priority and can
wait for nvidia's fix.
Note that modules are still generally fragile and untested with
llvm toolchain.
Closes: https://bugs.gentoo.org/885483
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
| |
Also note that it's ok to ignore if it works, hard to say how
everything behaves on different hardware and this is not hard
requirements (may start working again with new kernel/driver too).
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
|
| |
Unsure about exact status for this (was reportedly fixed in newer
drivers?), but seen a user have issues with it on 470.161.03.
Was still warning about this in 0/390, but also add 0/470 at least
(albeit this only check if set by default).
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
| |
... 510.85.02, 515.65.01
Bug: https://bugs.gentoo.org/884045
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Note that not giving runtime testing to 0/390 anymore given my
old hardware has gotten unreliable, please report if any notable
regressions.
But likely the last time have to look at a new 0/390 release given
NVIDIA support is ending this month. Won't be removed from ::gentoo
anytime soon but will get masked as a unsupported opt-in whenever
potential new security issues hit.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
| |
Obsolete new feature branch, please use production 515 or >=525 intead.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
| |
Masked beta version replaced by 525.60.11, no reason to keep.
Bug: https://bugs.gentoo.org/881025
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
| |
Non-beta new production branch and potential future stable
candidate (albeit 515.86.01 will be stable before it).
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
| |
New is stable for x86 now.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
| |
And drop now unnecessary conditional patches for kernel 6.0 support.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
| |
someday it will be stable for x86.. someday...
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/880879
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
| |
Remove the IBT warning, should now work in-theory with this
version (untested, please report if still have issues).
Also get to replace the no-gtk2 patch by BUILD_GTK2LIB=
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
| |
Reword a bit and add nvidia open issue link.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
|
|
|
|
|
|
| |
Kernel upstream is talking about enabling this by default (possibly
in 6.2), and users may be curious as to why they can't enable a
security feature.
Hoping nvidia will come up with something at least before 6.2.
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|
|
|
|
| |
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
|