On 1/19/2022 09:13, Sebastian Huber wrote:
On 19/01/2022 15:42, Kinsey Moore wrote:
On 1/18/2022 10:23, Sebastian Huber wrote:
On 18/01/2022 17:19, Kinsey Moore wrote:
Hardware test results have recently been posted for AArch64.
Closes #4581
It would be nice if you could run also the new validation tests on
this branch:
https://git.rtems.org/sebh/rtems.git/log/?h=validation
It looks like when running with that branch, tar02 and psxftw01 are
consistently failing with a PC in the stack since sometime after
68b0db358c81a7f770a44ceb819197bdc320bc99 which is what the other test
run used. I'll have to track that down.
Thanks for testing. It would be good to do a git bisect to figure out
the problem.
I'm currently doing that now and should have it narrowed down soon.
ts-performance-no-clock-0 gets test-too-long, but passes with
modified tester timeouts and all the other validation tests pass just
fine.
Great, did the ts-validation-intr test run also? It fails on my Qemu
simulator.
Yes, that test was included in the validation tests and passed.
I think the issue with ts-performance-no-clock-0 is the size of the
data cache. However, the data cache is an important factor for the
performance data, so there is no way around this.
Good to know. We may want to bump up the default test-too-long limit if
this is going to be the case.
Kinsey
_______________________________________________
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel
