On Fri, 30 Dec 2022 at 01:19, Philipp Kern wrote:
On 2022-12-29 14:22, Hugh McMaster wrote:
> Please retry idzebra 2.2.6-2 on s390x.
> Some timing tests failed on that arch but not on others.
> gb idzebra_2.2.6-2 . s390x
That's not entirely accurate. test/gils/test3.sh does not seem to be a
timing test. Unfortunately s390x is an architecture that is prone to a
"on that arch but not others" because it's the only big endian 64bit
architecture.
I gave it back just to rule out a fluke.
Thanks. The build worked this time. Unfortunately, hurd-i386 failed on two tests.
But please change the packaging
to actually output test logs on failure. It looks like the test in
question is writing to a $name.log file and just printing "FAIL: $name"
is not very useful.
Agreed. I’ll look into getting more useful output from the test logs.
Hugh