Re: [OE-core] [PATCH 1/1] ltp: Update to 20240129

[Petr Vorel <petr.vorel@gmail.com>]

> On Mon, 2024-02-05 at 11:43 +0100, Petr Vorel wrote:
> > > Hi Petr,

> > > On Sun, 2024-02-04 at 00:52 +0100, Petr Vorel wrote:
> > > > Hi Alexandre, all,

> > > > > On 31/01/2024 19:32:43+0100, Petr Vorel wrote:
> > > > > > Removed patch accepted upstream as
> > > > > > 8c53cc947 ("testcases/kernel/syscalls/fcntl: define _LARGEFILE64_SOURCE")

> > > > > > Rebase runtest/mm patch.

> > > > > > Signed-off-by: Petr Vorel <petr.vorel@gmail.com>
> > > > > > ---
> > > > > > Hi Richard, Daniel, Khem,

> > > > > > could anybody please test this update if build (it should).

> > > > > > Also, it'd be good nice to rerun proc01 -m 128 and report me with the
> > > > > > details, maybe we can fix it. If still problematic, it'd be nice to turn
> > > > > > disable_hanging_tests.patch into proper git patch (makes rebasing patch
> > > > > > on new LTP version easier).

> > > > > This fails on the autobuilders:

> > > > > https://autobuilder.yoctoproject.org/typhoon/#/builders/95/builds/5957/steps/12/logs/stdio
> > > > > https://autobuilder.yoctoproject.org/typhoon/#/builders/96/builds/5917/steps/12/logs/stdio


> > > > Thanks for testing this, you probably mean the warning before python stack trace:

> > > > scp: /opt/ltp/results/connectors: No such file or directory
> > > > WARNING: core-image-sato-1.0-r0 do_testimage: File could not be copied. Output: scp: /opt/ltp/results/connectors: No such file or directory

> > > > The problem is not due packaging, but simply runtest/connectors was removed [1],
> > > > thus your tooling has to be fixed not to expect it (not trying to run it).
> > > > I'm sorry, we sometimes do cleanup which brings incompatibility issue
> > > > (here it did not make sense to start SUT just for single test, thus we moved the
> > > > test to runtest/kernel_misc).

> > > I did try again with that removed:

> > > https://autobuilder.yoctoproject.org/typhoon/#/builders/96/builds/5929/steps/12/logs/stdio
> > > https://autobuilder.yoctoproject.org/typhoon/#/builders/95/builds/5968/steps/12/logs/stdio

> > > which comes down to:

> > > scp: /opt/ltp/results/fsx: No such file or directory

> > > so I'll remove that and try again.

> > > Are there any additions we should be considering too? This all feels a
> > > little fragile so I'm wondering if we should be reworking the way we
> > > handle this?

> > I forget that also this one was removed in fb2b6a0b3 ("fs: Remove fsx runtest
> > suite") [1], therefore not a real error.

> Thanks, with that removed we had successful runs so I've merged it.

Thank you for testing and fixing these groups in
meta/lib/oeqa/runtime/cases/ltp.py, I completely overlooked this file.
BTW there are move groups which could be added to ltp_groups:

cve, crypto, ima, net.nfs, net_stress.ipsec_icmp, net.ipv6, numa, uevent,
ltp-aiodio.part1, ltp-aiodio.part2, ltp-aiodio.part3, ltp-aiodio.part4

(There are more, thus list mostly the most important ones)

> I appreciate we need to try and re-enable that test that was hanging
> but I'm a bit stretched with other issues at the moment to dive into
> that one unfortunately.

Understand. Test has been rewritten to new API in 2017, I would expect they
would work properly. But the patchset removal was done in 2020.

https://bugzilla.yoctoproject.org/show_bug.cgi?id=13802

But looking at
https://git.yoctoproject.org/poky/commit/?id=529bb922c786f1acd041ff9a8d2b675d067e2ae3
linked via:
https://bugzilla.yoctoproject.org/show_bug.cgi?id=13802#c68

The tests found a real bug
https://lore.kernel.org/lkml/20210616125157.438837-1-paul.gortmaker@windriver.com/

While I understand the pain when test breaks whole testing, removing valid tests
means the functionality is not tested and we blindly believe kernel works.

Kind regards,
Petr

> Cheers,

> Richard