Dear stable kernel maintainers, Please consider the attached mbox file, which contains 9 patches which cherry pick cleanly onto 5.4:
1. commit fcf1b6a35c16 ("Documentation/llvm: add documentation on building w/ Clang/LLVM") 2. commit 0f44fbc162b7 ("Documentation/llvm: fix the name of llvm-size") 3. commit 63b903dfebde ("net: wan: wanxl: use allow to pass CROSS_COMPILE_M68k for rebuilding firmware") 4. commit 734f3719d343 ("net: wan: wanxl: use $(M68KCC) instead of $(M68KAS) for rebuilding firmware") 5. commit eefb8c124fd9 ("x86/boot: kbuild: allow readelf executable to be specified") 6. commit 94f7345b7124 ("kbuild: remove PYTHON2 variable") 7. commit aa824e0c962b ("kbuild: remove AS variable") 8. commit 7e20e47c70f8 ("kbuild: replace AS=clang with LLVM_IAS=1") 9. commit a0d1c951ef08 ("kbuild: support LLVM=1 to switch the default tools to Clang/LLVM")
This series improves/simplifies building kernels with Clang and LLVM utilities; it will help the various CI systems testing kernels built with Clang+LLVM utilities (in fact I will be pointing to this, if accepted, next week at plumbers with those CI system maintainers), and we will make immediate use of it in Android (see also: https://android-review.googlesource.com/c/platform/prebuilts/clang/host/linu...). We can always carry it out of tree in Android, but I think the series is fairly tame, and would prefer not to.
I only particularly care about 5+8+9 (eefb8c124fd9, 7e20e47c70f8, and a0d1c951ef08), but the rest are required for them to cherry-pick cleanly. I don't mind separating those three out, though they won't be clean cherry-picks at that point. It might be good to have Masahiro review the series. If accepted, I plan to wire up test coverage of these immediately in https://github.com/ClangBuiltLinux/continuous-integration/issues/300.
Most of the above landed in v5.7-rc1, with 94f7345b7124 landing in v5.6-rc1 and eefb8c124fd9 landing in v5.5-rc3.
On Thu, Aug 20, 2020 at 1:14 PM Nick Desaulniers ndesaulniers@google.com wrote:
Oh, I almost forgot, here's a picture of my cat who helped me by trying to lay on my keyboard mid-interactive rebase of this series. https://photos.app.goo.gl/J7CtBJtHmiuzhSfq7 I asked him nicely to move; he proceeded to bite me, and not comply.
Dear stable kernel maintainers, Please consider the attached mbox file, which contains 9 patches which cherry pick cleanly onto 5.4:
- commit fcf1b6a35c16 ("Documentation/llvm: add documentation on
building w/ Clang/LLVM") 2. commit 0f44fbc162b7 ("Documentation/llvm: fix the name of llvm-size") 3. commit 63b903dfebde ("net: wan: wanxl: use allow to pass CROSS_COMPILE_M68k for rebuilding firmware") 4. commit 734f3719d343 ("net: wan: wanxl: use $(M68KCC) instead of $(M68KAS) for rebuilding firmware") 5. commit eefb8c124fd9 ("x86/boot: kbuild: allow readelf executable to be specified") 6. commit 94f7345b7124 ("kbuild: remove PYTHON2 variable") 7. commit aa824e0c962b ("kbuild: remove AS variable") 8. commit 7e20e47c70f8 ("kbuild: replace AS=clang with LLVM_IAS=1") 9. commit a0d1c951ef08 ("kbuild: support LLVM=1 to switch the default tools to Clang/LLVM")
This series improves/simplifies building kernels with Clang and LLVM utilities; it will help the various CI systems testing kernels built with Clang+LLVM utilities (in fact I will be pointing to this, if accepted, next week at plumbers with those CI system maintainers), and we will make immediate use of it in Android (see also: https://android-review.googlesource.com/c/platform/prebuilts/clang/host/linu...). We can always carry it out of tree in Android, but I think the series is fairly tame, and would prefer not to.
I only particularly care about 5+8+9 (eefb8c124fd9, 7e20e47c70f8, and a0d1c951ef08), but the rest are required for them to cherry-pick cleanly. I don't mind separating those three out, though they won't be clean cherry-picks at that point. It might be good to have Masahiro review the series. If accepted, I plan to wire up test coverage of these immediately in https://github.com/ClangBuiltLinux/continuous-integration/issues/300.
Most of the above landed in v5.7-rc1, with 94f7345b7124 landing in v5.6-rc1 and eefb8c124fd9 landing in v5.5-rc3. -- Thanks, ~Nick Desaulniers
On Thu, Aug 20, 2020 at 01:27:32PM -0700, Nick Desaulniers wrote:
On Thu, Aug 20, 2020 at 1:14 PM Nick Desaulniers ndesaulniers@google.com wrote:
Oh, I almost forgot, here's a picture of my cat who helped me by trying to lay on my keyboard mid-interactive rebase of this series. https://photos.app.goo.gl/J7CtBJtHmiuzhSfq7 I asked him nicely to move; he proceeded to bite me, and not comply.
I think this needs to be an ongoing change to the stable kernel process. Since such things require documentation, I propose:
diff --git a/Documentation/process/stable-kernel-rules.rst b/Documentation/process/stable-kernel-rules.rst index 06f743b612c4..928f6f3ce6e2 100644 --- a/Documentation/process/stable-kernel-rules.rst +++ b/Documentation/process/stable-kernel-rules.rst @@ -35,6 +35,7 @@ Rules on what kind of patches are accepted, and which ones are not, into the Procedure for submitting patches to the -stable tree ----------------------------------------------------
+ - Preference may be given to submissions featuring pictures of cats. Also dogs. - If the patch covers files in net/ or drivers/net please follow netdev stable submission guidelines as described in :ref:`Documentation/networking/netdev-FAQ.rst <netdev-FAQ>`
Dear stable kernel maintainers, Please consider the attached mbox file, which contains 9 patches which cherry pick cleanly onto 5.4:
- commit fcf1b6a35c16 ("Documentation/llvm: add documentation on
building w/ Clang/LLVM") 2. commit 0f44fbc162b7 ("Documentation/llvm: fix the name of llvm-size") 3. commit 63b903dfebde ("net: wan: wanxl: use allow to pass CROSS_COMPILE_M68k for rebuilding firmware") 4. commit 734f3719d343 ("net: wan: wanxl: use $(M68KCC) instead of $(M68KAS) for rebuilding firmware") 5. commit eefb8c124fd9 ("x86/boot: kbuild: allow readelf executable to be specified") 6. commit 94f7345b7124 ("kbuild: remove PYTHON2 variable") 7. commit aa824e0c962b ("kbuild: remove AS variable") 8. commit 7e20e47c70f8 ("kbuild: replace AS=clang with LLVM_IAS=1") 9. commit a0d1c951ef08 ("kbuild: support LLVM=1 to switch the default tools to Clang/LLVM")
And FWIW, "yes please" from me as well. This makes things muuuch easier to test LTS with Clang.
On Thu, Aug 20, 2020 at 01:14:36PM -0700, Nick Desaulniers wrote:
Dear stable kernel maintainers, Please consider the attached mbox file, which contains 9 patches which cherry pick cleanly onto 5.4:
- commit fcf1b6a35c16 ("Documentation/llvm: add documentation on
building w/ Clang/LLVM") 2. commit 0f44fbc162b7 ("Documentation/llvm: fix the name of llvm-size") 3. commit 63b903dfebde ("net: wan: wanxl: use allow to pass CROSS_COMPILE_M68k for rebuilding firmware") 4. commit 734f3719d343 ("net: wan: wanxl: use $(M68KCC) instead of $(M68KAS) for rebuilding firmware") 5. commit eefb8c124fd9 ("x86/boot: kbuild: allow readelf executable to be specified") 6. commit 94f7345b7124 ("kbuild: remove PYTHON2 variable") 7. commit aa824e0c962b ("kbuild: remove AS variable") 8. commit 7e20e47c70f8 ("kbuild: replace AS=clang with LLVM_IAS=1") 9. commit a0d1c951ef08 ("kbuild: support LLVM=1 to switch the default tools to Clang/LLVM")
This series improves/simplifies building kernels with Clang and LLVM utilities; it will help the various CI systems testing kernels built with Clang+LLVM utilities (in fact I will be pointing to this, if accepted, next week at plumbers with those CI system maintainers), and we will make immediate use of it in Android (see also: https://android-review.googlesource.com/c/platform/prebuilts/clang/host/linu...). We can always carry it out of tree in Android, but I think the series is fairly tame, and would prefer not to.
I only particularly care about 5+8+9 (eefb8c124fd9, 7e20e47c70f8, and a0d1c951ef08), but the rest are required for them to cherry-pick cleanly. I don't mind separating those three out, though they won't be clean cherry-picks at that point. It might be good to have Masahiro review the series. If accepted, I plan to wire up test coverage of these immediately in https://github.com/ClangBuiltLinux/continuous-integration/issues/300.
Most of the above landed in v5.7-rc1, with 94f7345b7124 landing in v5.6-rc1 and eefb8c124fd9 landing in v5.5-rc3.
All now queued up, thanks.
greg k-h
linux-stable-mirror@lists.linaro.org
-
Greg KH -
Kees Cook -
Nick Desaulniers