From: Justin M. Forbes on gitlab.com
Merge Request: https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2263
For many years, rawhide kernels have forced users to run debug builds on
git snapshots by not building a non debug kernel as an option. While
this has served us well, in finding
From: Justin M. Forbes
Turn off forced debug builds
For many years, rawhide kernels have forced users to run debug builds on
git snapshots by not building a non debug kernel as an option. While
this has served us well, in finding occasional bugs that are less likely
to surface otherwise, the
From: Justin M. Forbes
Disable frame pointers
Fedora is defaulting to add -fno-omit-frame-pointer by default. This is
not necessary for kernel.
Signed-off-by: Justin M. Forbes
diff --git a/redhat/kernel.spec.template b/redhat/kernel.spec.template
index blahblah..blahblah 100755
---
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2240#note_1245811704
I will approve it, though honestly Makefile.variables was currently arranged
from most likely to be changed, meaning the further you go down the file, the
less likely it is that
From: Justin M. Forbes
Turn off CONFIG_MTK_T7XX for S390x
A build issue crept in over the weekend with the MTK_T7XX driver and
S390x. As I started digging, I realized this is likely hardware that
would never be used with S390x, and might be better off just disabled
for that arch. It is a
From: Justin M. Forbes
Fix up configs with SND_SOC_NAU8315 mismatch
Upstream commit ba7523bb0f494fc440d3a9bb0b665cfcaa192d0c now selects
SND_SOC_NAU8315 if SND_SOC_INTEL_SOF_NAU8825_MACH is set. This forces us
to either turn on SND_SOC_NAU8315 or turn off
SND_SOC_INTEL_SOF_NAU8825_MACH. I
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2179#note_1236077821
Happy to ack this once the unrelated commits are dropped.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2232#note_1236075645
Closing this MR as the upstream commit does indeed fix the issue.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2232#note_1234616148
Taking include in release off of this one as it is supposed to be fixed by
commit 99cb0d917ffa1ab628bb67364ca9b162c07699b1
___
kernel
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2240#note_1230424301
We are already building with `-Werror=format-security` Though we typically do
not run across many errors until a new toolchain is introduced. That is where
this gets problematic to
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2175#note_1226265381
https://lists.fedoraproject.org/archives/list/kernel@lists.fedoraproject.org/t
hread/HDANRFTC3JIHMSSJYL5Z5H4RM2ULZYZY/#E42HQLKEODQBZZP5LDNVBCDH5N32XHX5 That
was the reason, and it
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2175#note_1226168874
This shouldn't come out of draft until it has been approved from FESCo.
I still do not approve of renaming kernel-modules to kernel-modules-standard
for no actual gain or reason
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2225#note_1216465167
That is the general guidance so that we do not end up with uncommon items in
common, though this change is correct use of common, and where it would end up
anyway.
From: Justin M. Forbes
Revert "arm64: remove special treatment for the link order of head.o"
This reverts commit 994b7ac1697b4581b7726d2ac64321e3c840229b.
This commit caused a missing build-id for vmlinux on aarch64. I expect
it will be a simple revert upstream as this commit was not trying
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2173#note_1212551213
But is it in Fedora 36 - Rawhide built binaries?
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email
From: Justin M. Forbes
Add siphash_kunit and strscpy_kunit to mod-internal.list
These are 2 more kunit tests that came in during the 6.2 merge window
and need to be added to mod-internal.list for depmod to succeed.
Signed-off-by: Justin M. Forbes
diff --git
From: Justin M. Forbes
Add drm_kunit_helpers to mod-internal.list
Another kunit mod that came in during the 6.2 merge window, and needs to
be added to mod-internal.list for depmod to succeed.
Signed-off-by: Justin M. Forbes
diff --git a/redhat/scripts/mod/mod-internal.list
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2125#note_1203675945
```
error: File not found: /builddir/build/BUILDROOT/kernel-6.1.0-0.rc8.20221209gi
t0d1409e4ff08.62.eln124.aarch64/usr/lib/systemd/system/cpupower.service
error: File not found:
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2165#note_1203309926
So it seems this will now fail if we have no warnings.
```
No new configuration values exposed from merging master into os-build
232Pushing branch os-build to
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2174#note_1196429775
Curious as to why. We did have this in Fedora for years before we had kernel-
ark. But we also had every single patch broken out, and applied each with git
am. There was some
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2166#note_1195335799
I was saying it is fine as 32bit will be going away.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2166#note_1194864927
Fedora drops 32 bit arm when F36 goes EOL in roughly 7 months. At that point,
I expect we will do the same consolidation done for x86, and the aarch64
directory will go away,
From: Justin M. Forbes
Adjust path to compressed vmlinux kernel image for s390x
commit edd4a8667355 moved the location of the compressed vmlinux file
from arch/s390/boot/compressed/vmlinux to arch/s390/boot/vmlinux and
packaging needs to be updated accordingly.
Bugzilla:
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2161#note_1181584664
Yes, this makes sense. Fedora was using the common config before, but since it
is moved, we needed a new one.
___
kernel mailing list --
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2157#note_1177932177
Not so much, the bigger concern was your script in CS9 claiming mismatch with
ark and people dropping configs into common without respect to RHEL or Fedora.
As I understand it,
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2153#note_1175190946
I am not so worried about the common bit, we have a script that consolidates
matching ark and fedora entries into common which is run periodically.
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2137#note_1170782780
This is upstream now.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to
From: Johannes Weiner
fs: fix leaked psi pressure state
When psi annotations were added to to btrfs compression reads, the psi
state tracking over add_ra_bio_pages and btrfs_submit_compressed_read was
faulty. A pressure state, once entered, is never left. This results in
incorrectly elevated
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2124#note_1162434713
Not until upstream changes (I expect them to, but it hasn't happened yet):
```
config EFI_SECRET
tristate "EFI secret area securityfs support"
depends on EFI &&
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2124#note_1162366072
I am not quite sure the purpose of this MR. The config option is already
turned on in RHEL and in Fedora, though in different places.
```
cat
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2117#note_1151622605
`BuildRequires: python3-devel`
This is already present.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2108#note_1151085979
This is upstream now.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2112#note_1146671572
https://bugzilla.redhat.com/show_bug.cgi?id=1939095 Was the bug forcing this
to be turned on in RHEL, so it seems that this MR is wrong for both Fedora and
ELN/RHEL. Merging it
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1533#note_1146665556
The problem there, is you need =m for s390x and =y for s390x zfcpdump. I am
not sure how much testing goes into each build with s390x vs the zfcpdump
variant, but it seems for
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2112#note_1142157654
/block
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to kernel-le...@lists.fedoraproject.org
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2112#note_1142154016
Why? Not sure about RHEL, but Fedora and ELN are shipping linux-firmware with
xz compressed firmware files.
___
kernel mailing list --
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2110#note_1142110376
Correct, this MR will do nothing. We already have:
```
cat redhat/configs/common/generic/CONFIG_PTE_MARKER_UFFD_WP
CONFIG_PTE_MARKER_UFFD_WP=y
```
And:
```
config
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2111#note_1142106478
If you are going to do that, you should probably delete
redhat/configs/common/generic/arm/CONFIG_ACPI_VIDEO as well. The current
configs leave CONFIG_ACPI_VIDEO on for arm and x86.
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2083#note_1138719400
I think the real concern is that the actual driver (and Kconfig entry) for
CONFIG_DWMAC_TEGRA is not upstream yet, it is not even in linux-next just yet.
There was discussion on
From: Adrian Hunter
perf tools: Fix man page build wrt perf-arm-coresight.txt
perf build assumes documentation files starting with "perf-" are man
pages but perf-arm-coresight.txt is not a man page:
asciidoc: ERROR: perf-arm-coresight.txt: line 2: malformed manpage title
asciidoc: ERROR:
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2105#note_1137974600
Interestingly, if I switch into that directory, and try to git apply manually,
it doesn't complain at all, but doesn't apply. git apply --verbose shows that
it skips every patch.
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2105#note_1137946990
The issue I observed is that after a "fedpkg prep" there is no error around
git-apply, but the patch itself is not applied. This is actually quite
visible in the fact that we have
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2105#note_1137086820
Thanks for finding this, I was just trying to figure out why an issue was
popping up in the 6.0 test week.
___
kernel mailing list --
From: Justin M. Forbes
Add kasan_test to mod-internal.list
Another new kunit test that fails depmod until it is added to
mod-internal.list.
Signed-off-by: Justin M. Forbes
diff --git a/redhat/mod-internal.list b/redhat/mod-internal.list
index blahblah..blahblah 100644
---
From: Justin M. Forbes
Rename rename FORCE_MAX_ZONEORDER to ARCH_FORCE_MAX_ORDER in configs
Upstream commit 0192445cb2f7e renamed FORCE_MAX_ZONEORDER to
ARCH_FORCE_MAX_ORDER which made it appear as a new config item for us in
pending. While I have fixed up pending, this MR just fixes up the
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2041#note_1128275734
I might argue when working a bug report, the uname tells you exactly which
config it is, and you don't need their /boot to get it. It is in our rpm, if
not, they wouldn't have a
From: Justin M. Forbes
Add new drm kunit tests to mod-internal.list
The drm subsystem added a number of new kunit tests in the 6.1 cycle.
These need to be added to mod-internal.list for depmod to succeed.
Signed-off-by: Justin M. Forbes
diff --git a/redhat/mod-internal.list
From: Justin M. Forbes
Add acpi video to the filter_modules.sh for rhel
The ACPI_VIDEO module now has a dep on the symbol wmi_evaluate_method
and as a result it needs to be added to the module filter or depmod will
fail.
Signed-off-by: Justin M. Forbes
diff --git
From: Justin M. Forbes
Change acpi_bus_get_acpi_device to acpi_get_acpi_dev
Upstream commit 45e9aa1fdbb2e renamed acpi_bus_get_acpi_device to
acpi_get_acpi_dev. As we are carrying an out of tree patch [1] which calls
acpi_bus_get_acpi_device, we need to make the corresponding change
ourselves
From: Justin M. Forbes
Add new drm kunit tests to mod-internal.list
The drm subsystem added a number of new kunit tests in the 6.1 cycle.
These need to be added to mod-internal.list for depmod to succeed.
Signed-off-by: Justin M. Forbes
diff --git a/redhat/mod-internal.list
From: Justin M. Forbes
Add fortify_kunit to mod-internal.list
Another kunit test which has been added and needs to end up in
mod-internal so that depmod succeeds.
Signed-off-by: Justin M. Forbes
diff --git a/redhat/mod-internal.list b/redhat/mod-internal.list
index blahblah..blahblah 100644
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2035#note_1119734952
Yes, there are a lot of them, and it is an issue with how the scripts create
an MR for a new config item. They just aren't smart enough to know if
something is arch specific, so
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2030#note_1116538902
I reset the approval rules, this is a no-op re-org. It does not change the
most config files at all. The only change for RHEL is debug x86_64 kernels
now set
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1923#note_1116349115
There is a negligible performance impact, rhbz 2115997 has some testing
results.
___
kernel mailing list --
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2022#note_1105911452
Do you have a need to flip CONFIG_ARM64_VA_BITS_47 for some reason? We build
with 48, and there is no dep on this config item with current kernels in this
repository.
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2019#note_1100691692
This still seems much more complex than simply adding a buildrequres on glibc-
langpack-en and forcing LC_ALL=en_US.UTF-8.
___
kernel
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2011#note_1098884245
Why would it be a problem to set LC_ALL=en_US.utf8 instead of LC_ALL=C? That
still sets a specific locale, and leaves the result in the order that people
expect.
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2011#note_1098852942
If you remove the sort, it is again in a rather illogical order.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2011#note_1098747753
I do understand the problem that you are trying to solve, and I also
understand that the generated config files do build the kernels in the same
way. The problem is in how the
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2011#note_1098637550
I am reverting this, we can perhaps find another solution that will work, but
I am not willing to commit the config files that this generated to dist-git
for a build today.
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2011#note_1098634600
For example, before the change this is all one chunk beginning at line 1138:
```
CONFIG_CLK_RV110X=y
CONFIG_CLK_SP810=y
# CONFIG_CLKSRC_STM32_LP is not set
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2011#note_1098582094
The consequence of this is it changed the sorting of generated config files,
and now every option turned off is listed first in alphabetical order (
because it starts with '#' )
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2009#note_1098499621
That will work as well. Prarit and I were just discussing getting rid of the
x86_64 subdirectory all together, and then using the clean up script to catch
all of the duplicates.
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2018#note_1098363402
This seems like unnecessary complexity to me. In the end, we are adding spec
complexity and making packagers try to divine the correct place to put a new
module, when it doesn't
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2009#note_1098347268
The `redhat/configs/generic/aarch64/` does not exist in this tree or in RHEL
9, and putting a config entry there is ignored. Please move this to
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/2016#note_1096312041
I am not sure why you are failing, or seeing what you are, the cpumask_kunit
entry was added with MR https://gitlab.com/cki-project/kernel-
ark/-/merge_requests/2000 the day that
From: Justin M. Forbes
Add cpumask_kunit to mod-internal.list
Another new kunit module that came in with 6.0.
Signed-off-by: Justin M. Forbes
diff --git a/redhat/mod-internal.list b/redhat/mod-internal.list
index blahblah..blahblah 100644
--- a/redhat/mod-internal.list
+++
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1992#note_1074292303
This is upstream now.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to
From: Yang Jihong
perf tools: Fix compile error for x86
Commit a0a12c3ed057 ("asm goto: eradicate CC_HAS_ASM_GOTO") eradicates
CC_HAS_ASM_GOTO, perf on x86 call asm_volatile_goto when compiling __GEN_RMWcc.
However, asm_volatile_goto is not declared, which causes compilation error:
In file
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980#note_1067979814
That's perfect. The change is in the upstream versioning and now it is
reflected clearly. I am happy with that.
___
kernel mailing list
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1980#note_1067939935
I actually prefer something along the lines of:
`- Linux v6.0.0-0.rc1.3cc40a443a04`
It technically isn't a rebase, it is a merge, but for as long as I can recall,
prior to ark,
From: Justin M. Forbes
Add python3-setuptools buildreq for perf
With the 6.0 updates for perf, the build fails with a missing
ModuleNotFoundError: No module named 'setuptools' if we don't have
python3-setuptools as a buildreq.
Signed-off-by: Justin M. Forbes
diff --git
From: Justin M. Forbes on gitlab.com
Merge Request: https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1976
New test modules added during the 6.0 merge window need to be added to
mod-internal.list for depmod to succeed.
Signed-off-by: Justin M. Forbes
---
redhat/mod-internal.list | 4
From: Justin M. Forbes
Add cros_kunit to mod-internal.list
This one is Fedora only at the moment, and aarch64
Signed-off-by: Justin M. Forbes
diff --git a/redhat/mod-internal.list b/redhat/mod-internal.list
index blahblah..blahblah 100644
--- a/redhat/mod-internal.list
+++
From: Justin M. Forbes
Add new tests to mod-internal.list
New test modules added during the 6.0 merge window need to be added to
mod-internal.list for depmod to succeed.
Signed-off-by: Justin M. Forbes
diff --git a/redhat/mod-internal.list b/redhat/mod-internal.list
index blahblah..blahblah
From: Justin M. Forbes
Add new tests to mod-internal.list
New test modules added during the 6.0 merge window need to be added to
mod-internal.list for depmod to succeed.
Signed-off-by: Justin M. Forbes
diff --git a/redhat/mod-internal.list b/redhat/mod-internal.list
index blahblah..blahblah
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1422#note_1051481242
While this started out as a good idea, it may be coming back to bite us. So
far during the 6.0 merge window, 2 new commits have come in which cause major
issues with this setup:
From: Justin M. Forbes
Change CRYPTO_BLAKE2S_X86 from m to y
Upstream commit 2d16803c562ec changed the Kconfig entry for
CRYPTO_BLAKE2S_X86 from tristate to bool, meaning if we wish to keep it
enabled, it must be moved to built-in.
Signed-off-by: Justin M. Forbes
diff --git
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1923#note_1049409477
That's the rebase, we do it every .0 release. Sorry I have not commented on
this yet, as I was on vacation last week, and we are now going into the merge
window/test week madness.
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1905#note_1022360309
Fedora typically does enable new crypto devices as modules where available.
The exceptions are devices which have been shut off due to export controls, or
if someone has given a
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1904#note_1022326088
Closing as this is upstream now.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to
From: Thomas Zimmermann
drm/aperture: Run fbdev removal before internal helpers
Always run fbdev removal first to remove simpledrm via
sysfb_disable(). This clears the internal state. The later call
to drm_aperture_detach_drivers() then does nothing. Otherwise,
with
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1901#note_1013792512
Actually it seems that there is no rtla.debug showing up.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1900#note_1013293356
builddep can be passed a spec file or srpm as well, unfortunately it seems to
choke if passed the kernel.spec.template, so we would need a processed
template into a proper spec.
From: Justin M. Forbes
Add rtla subpackage for kernel-tools
As a package review request came up to create a new rtla package, it was
determined that it makes the most sense to just build it as a subpackage
of kernel-tools because that is where it is maintained upstream. While
the Fedora
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1900#note_1012576729
'sudo dnf builddep kernel' should install the list of required packages
without much hassle.
___
kernel mailing list --
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1861#note_988096375
Upstream now.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to
From: Jason Wang
virtio-rng: make device ready before making request
Current virtio-rng does a entropy request before DRIVER_OK, this
violates the spec:
virtio spec requires that all drivers set DRIVER_OK
before using devices.
Further, kernel will ignore the interrupt after commit
From: Justin M. Forbes
Fix typo in Makefile for Fedora Stable Versioning
Commit e9d807ca3424 got rid of the special rhel versioning for Fedora
but I accidentally moved the SNAPSHOT=0 into the not fedora section when
it was added for Fedora to begin with. This resulted in all builds for
stable
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1855#note_973966184
Thanks for doing this.
___
kernel mailing list -- kernel@lists.fedoraproject.org
To unsubscribe send an email to
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1853#note_972579252
This isn't a RHEL change, the only RHEL movement at all is moving a file from
ark to common so that Fedora inherits the RHEL config. The average commit
message for fedora configs
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1834#note_971070625
As we get further in this release cycle, possibly around the rc6 time frame, I
plan to do a rather large config cleanup provided I have the cycles to do so.
We have a couple of
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1801#note_969753502
This seems correct, looking at a dist-dump-variables.
```
BASEVERSION=5.19.0-0.rc0.d1dc87763f40.10 (perhaps misleadingly named?)
SPECVERSION=5.19.0
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1828#note_969355344
@omos It seems like it would, but in practice it does not. I typically sit
down and do a large number of fedora configs in a one sitting, one commit.
Particularly for obvious ones,
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1841#note_967953140
More likely, we will pick up riscv before RHEL does, though I do not know a
timeframe on that.
___
kernel mailing list --
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1841#note_967949964
An incorrect example, but an example. Fedora has not built i686 kernels for a
very long time. We do still build armv7 kernels until Fedora 36 is EOL, but
those are going away as
From: Justin M. Forbes
Explicitly turn off CONFIG_KASAN_INLINE for ppc
While upstream commit 41b7a347 made us turn on KASAN_OUTLINE for ppc
because the depends for KASAN_INLINE are no longer met, those config
options are part of a "one or the other" Choice menu. Having both
enabled seems to
From: Justin M. Forbes
Turn on KASAN_OUTLINE for ppc debug
Upstream commit 41b7a347bf14 forces outline-only KASAN support for
Book3S. As a result we need to move PPC to outline mode in the configs
or we have a mismatch.
Signed-off-by: Justin M. Forbes
diff --git
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1834#note_963572229
1. Right, the pipelines fail, I have to manually fix this stuff up every time
a new one appears. When it is fixed, there is no automated process to go and
re-run failed pipelines. I
From: Justin M. Forbes on gitlab.com
https://gitlab.com/cki-project/kernel-ark/-/merge_requests/1834#note_963551781
The pipelines were failing today due to something different (KASAN config for
ppc). I already cleaned all of this up in pending yesterday to make pipelines
work, and I have added
201 - 300 of 577 matches
Mail list logo