Hi all,
As I'm not going to be able to make the meeting today (scheduling conflict with another thing here in Seattle), here's my initial critique of the existing reports:
On Thu, Aug 10, 2017 at 08:59:51PM +0000, Linaro QA wrote:
Tests: 1117 Failed: 23 Passed: 1092 Skipped: 2 Build: v4.4.80-568085-gc40a9a7
That means nothing to me. Especially as this really was a 4.4.81-rc build, right? So here I am thinking that 4.4.80 might have problems?
Why are you not picking up the -rc and other Makefile changes, they are part of the git tree, are you pulling directly from the quilt series instead?
Details: https://qa-reports.linaro.org/lkft/android-hikey-linaro-4.4-oe/build/v4.4.80...
series: lkft
What does "series" mean? What does "lkft" mean? Where can I go to look up these things?
What does "oe" mean? And again, where can I look those things up?
kernel-describe: v4.4.80-568085-gc40a9a7
Same as above? Why repeat it? How will it ever be different?
kernel-branch: android-hikey-linaro-4.4
Shouldn't this be reflected in the kernel version?
Where can I find these branches? What do the different branches mean?
build-url: https://ci.linaro.org/job/openembedded-lkft-aosp-4.4/DISTRO=rpb,MACHINE=hike... kernel-repo: https://android.googlesource.com/kernel/hikey-linaro kernel-config: https://snapshots.linaro.org/openembedded/lkft/morty/hikey/rpb/4.4/62/defcon... kernel-commit: c40a9a77daf770cd8ca75572235b015ef6ff6589
That's already listed above in the describe and version, right? Why list it again?
job_status: Complete
As others said, can this ever not be this? If so, really? How?
Regressions (compared to build v4.4.80-568084-g7f6b98c)
Your compared-to build means nothing to me. Again, the kernel version is off, and still, what does it mean? What can I do with this?
(none)
Nice.
Failures
hi6220-hikey:
- kselftest/pstore_tests -- never passed
- kselftest/ftracetest -- never passed
- kselftest/fw_fallback.sh -- never passed
- kselftest/pidns -- never passed
- kselftest/breakpoint_test_arm64 -- never passed
- kselftest/test_tag -- never passed
- kselftest/test_bpf.sh -- never passed
- kselftest/run_fuse_test.sh -- never passed
- ltp-syscalls-tests/pselect01_64 -- failing since build v4.4.79-568979-g43c93f7, from Aug. 1, 2017, 10:12 p.m.
- kselftest/run_vmtests -- failing since build v4.4.77-568946-g4d4632a, from July 21, 2017, 1:28 p.m.
- kselftest/test_kmod.sh -- never passed
- kselftest/bitmap.sh -- never passed
- kselftest/test_maps -- never passed
- ltp-syscalls-tests/utimensat01 -- never passed
- kselftest/test_lpm_map -- never passed
- kselftest/test_align -- never passed
- kselftest/gpio-mockup.sh -- never passed
- kselftest/test_lru_map -- never passed
- kselftest/seccomp_bpf -- never passed
- kselftest/run.sh -- never passed
- ltp-syscalls-tests/runltp_syscalls -- never passed
- kselftest/owner -- never passed
- kselftest/inconsistency-check -- never passed
You all said you were working on this. What I really want to know is if I broke something, and about these things that are continuing to fail, what the status is for them. Are they build system issues? Upstream kernel issues? Merge issues with your SoC tree? Something else?
All changes (compared to build v4.4.80-568084-g7f6b98c)
+--- v4.4.80-568084-g7f6b98c, hi6220-hikey | +--- v4.4.80-568085-gc40a9a7, hi6220-hikey
Again, build version problems :(
| | pass n/a kselftest/execveat fail n/a kselftest/fw_filesystem.sh fail n/a kselftest/test_verifier
I don't understand this, 2 new things are failing and one new thing is passing? Shouldn't the 2 new failures be listed above as regressions?
And should we have a separate "this is now fixed" category?
thanks,
greg k-h