On Thu, Nov 23, 2017 at 10:40 AM, Linaro QA qa-reports@linaro.org wrote:
Summary
kernel: 4.14.0 git repo: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git git branch: master git commit: 5a787756b809888e8925d722862167f1229b58f7 git describe: v4.14-13151-g5a787756b809 Test details: https://qa-reports.linaro.org/lkft/linux-mainline-oe/build/v4.14-13151-g5a78...
Regressions (compared to build v4.14-12871-ged30b147e1f6)
hi6220-hikey - arm64: boot: * hi6220-hikey
* test src: not informed
This still suffers from https://bugs.linaro.org/show_bug.cgi?id=3345, the patch is on its way into mainline now, but it's not there yet.
An interesting observation is that we list "boot" as failed here, but also show other test cases as "success". For those, we clearly run into a variation of the same problem, and get a kernel Oops in the log, but it doesn't stop the test case from running.
I suppose the Oops can happen in different places (we only crash when reusing freed memory that has been allocated by some other module, and it's unpredictable which module gets this memory), and only some of them lead to a panic.
In general, I think we should treat a test run that causes an 'Oops' as failed, regardless of what the test output shows.
Arnd