Re: [OE-core] [PATCH 3/3] m4: Fix ptest on musl

["Mathieu Dubois-Briand" <mathieu.dubois-briand@bootlin.com>]

On Wed Aug 20, 2025 at 5:24 PM CEST, Khem Raj via lists.openembedded.org wrote:
> On Wed, Aug 20, 2025 at 7:40 AM Marko, Peter <Peter.Marko@siemens.com> wrote:
>>
>>
>>
>> > -----Original Message-----
>> > From: Khem Raj <raj.khem@gmail.com>
>> > Sent: Wednesday, August 20, 2025 16:33
>> > To: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>; Marko, Peter
>> > (FT D EU SK BFS1) <Peter.Marko@siemens.com>; Ross Burton
>> > <ross.burton@arm.com>
>> > Cc: openembedded-core@lists.openembedded.org
>> > Subject: Re: [OE-core] [PATCH 3/3] m4: Fix ptest on musl
>> >
>> > On Wed, Aug 20, 2025 at 4:10 AM Mathieu Dubois-Briand
>> > <mathieu.dubois-briand@bootlin.com> wrote:
>> > >
>> > > On Wed Aug 20, 2025 at 9:15 AM CEST, Khem Raj wrote:
>> > > > On Tue, Aug 19, 2025 at 11:23 PM Mathieu Dubois-Briand
>> > > > <mathieu.dubois-briand@bootlin.com> wrote:
>> > > >>
>> > > >> On Tue Aug 19, 2025 at 3:01 AM CEST, Khem Raj via
>> > lists.openembedded.org wrote:
>> > > >> > Fixes
>> > > >> > ../../sources/m4-1.4.20/tests/test-c32ispunct.c:261: assertion 'is == 0' failed
>> > > >> > ./test-c32ispunct.sh: line 36:   402 Aborted
>> > > >> > (core dumped) LC_ALL="$testlocale" ${CHECKER} ./test-
>> > c32ispunct${EXEEXT} 3
>> > > >> >
>> > > >> > FAIL: test-c32ispunct.sh
>> > > >> >
>> > > >> > Signed-off-by: Khem Raj <raj.khem@gmail.com>
>> > > >> > ---
>> > > >>
>> > > >> Hi Khem,
>> > > >>
>> > > >> Thanks for your patch.
>> > > >>
>> > > >> It looks like this is breaking tests, specifically on qemuriscv64:
>> > > >>
>> > > >> Traceback (most recent call last):
>> > > >>   File "/srv/pokybuild/yocto-
>> > worker/qemuriscv64/build/meta/lib/oeqa/sdk/cases/maturin.py", line 25, in
>> > test_maturin_list_python
>> > > >>     output = self._run("maturin list-python")
>> > > >>              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> > > >>   File "/srv/pokybuild/yocto-
>> > worker/qemuriscv64/build/meta/lib/oeqa/sdk/case.py", line 17, in _run
>> > > >>     return subprocess.check_output(". %s > /dev/null; %s;" % \
>> > > >>            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> > > >>   File "/usr/lib/python3.12/subprocess.py", line 466, in check_output
>> > > >>     return run(*popenargs, stdout=PIPE, timeout=timeout, check=True,
>> > > >>            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> > > >>   File "/usr/lib/python3.12/subprocess.py", line 571, in run
>> > > >>     raise CalledProcessError(retcode, process.args,
>> > > >> oeqa.utils.subprocesstweak.OETestCalledProcessError: Command '.
>> > /srv/pokybuild/yocto-worker/qemuriscv64/build/build/tmp/work/qemuriscv64-poky-
>> > linux/core-image-minimal/1.0/testsdkext/environment-setup-riscv64imafdc-poky-
>> > linux > /dev/null; maturin list-python;' died with <Signals.SIGSEGV: 11>.
>> > > >>
>> > > >
>> > > > seems quite unrelated, I wonder if its finding some other bug
>> > >
>> > > Yes, I can't understand what happens here :(
>> > > Yet this commit always triggers the failure.
>> > >
>> >
>> > Looking at logs there is some more info  immediately after what you
>> > posted above that is interesting
>> >
>> > Stdout:
>> > Trying to install python3-maturin-native...
>> >
>> > its running lib/oeqa/sdk/cases/maturin.py and
>> > it seems it needs python3-maturin-native in sstate cache and I wonder
>> > if it's missing because m4/m4-native dependency will rebuild
>> > python3-maturin-native or something is racing with it.
>> >
>> > There was a similar problem in recent past and a related fix was committed
>> > https://git.yoctoproject.org/poky/commit/?id=2239c2c91ab35c67ccd95afe0573500
>> > 37f7d0e59
>> >
>> > I now wonder if the test is exited when it does not find
>> > python3-maturin-native correctly or not.
>>
>> The fact that maturin test is executed means that it was installed successfully, so code in mentioned commit was not active.
>> Also segmentation fault suggests that the binary is installed, it just does not work correctly.
>
> Seems so, I would have expected SkipTest to execute if the install
> failed. so it seems to be installed fine, but segfaulting on run
> seems unrelated to above change which is modifying a testcase source
> file in m4, its not changing anything even in m4,  there
> was a possibility that modified m4-native is indirectly injecting some
> bug into maturin but thats unlikely.
>
> maybe run a build with just this single patch applied on top of the
> master branch.
>

https://autobuilder.yoctoproject.org/valkyrie/#/builders/45/builds/368

Also failing :(


-- 
Mathieu Dubois-Briand, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com