Re: [PATCH 6.6 000/129] 6.6.90-rc2 review

[Ron Economos]

On 5/8/25 23:19, Greg Kroah-Hartman wrote:
> On Thu, May 08, 2025 at 06:16:12PM +0530, Naresh Kamboju wrote:
> > On Thu, 8 May 2025 at 17:00, Greg Kroah-Hartman
> > <gregkh@xxxxxxxxxxxxxxxxxxx> wrote:
> > > This is the start of the stable review cycle for the 6.6.90 release.
> > > There are 129 patches in this series, all will be posted as a response
> > > to this one.  If anyone has any issues with these being applied, please
> > > let me know.
> > >
> > > Responses should be made by Sat, 10 May 2025 11:25:47 +0000.
> > > Anything received after that time might be too late.
> > >
> > > The whole patch series can be found in one patch at:
> > >          https://www.kernel.org/pub/linux/kernel/v6.x/stable-review/patch-6.6.90-rc2.gz
> > > or in the git tree and branch at:
> > >          git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-6.6.y
> > > and the diffstat can be found below.
> > >
> > > thanks,
> > >
> > > greg k-h
> > Regressions on riscv with allyesconfig and allmodconfig builds failed with
> > clang-20 and gcc-13 toolchain on the stable-rc 6.6.90-rc1 and 6.6.90-rc2
> >
> > * riscv, build
> >    - clang-20-allmodconfig
> >    - gcc-13-allmodconfig
> >    - gcc-13-allyesconfig
> >
> > Regression Analysis:
> >   - New regression? Yes
> >   - Reproducibility? Yes
> >
> > Build regression: riscv uprobes.c unused variable 'start'
> >
> > Reported-by: Linux Kernel Functional Testing <lkft@xxxxxxxxxx>
> >
> > ## Build error riscv
> > arch/riscv/kernel/probes/uprobes.c: In function 'arch_uprobe_copy_ixol':
> > arch/riscv/kernel/probes/uprobes.c:170:23: error: unused variable
> > 'start' [-Werror=unused-variable]
> >    170 |         unsigned long start = (unsigned long)dst;
> >        |                       ^~~~~
> > cc1: all warnings being treated as errors
> Oh that's wierd.  riscv defines flush_icache_range() as "empty" so then
> this patch does nothing in these older kernels.  Ah, it's an inline
> function in newer kernel trees as well so that the build warning isn't
> there anymore.
>
> As this change feels odd for 6.6 and older kernels, AND it's causing
> build warnings, I'm just going to drop it and if the riscv maintainers
> really want it applied to these trees, will take a working backport from
> them.
>
> thanks,
>
> greg k-h

This is the same build failure from last week. There's a fixup patch 
"riscv: Replace function-like macro by static inline function" upstream 
commit 121f34341d396b666d8a90b24768b40e08ca0d61 that doesn't apply to 
6.6, 6.1 and 5.15.

You dropped the patch "riscv: uprobes: Add missing fence.i after 
building the XOL buffer" last week, but it snuck back in this week.