bug#28111: qemu testsuite fails

[Maxim Cournoyer]

ludo@gnu.org (Ludovic Courtès) writes:

> Mathieu Lirzin <mthl@gnu.org> skribis:
>
>> Mathieu Lirzin <mthl@gnu.org> writes:
>>
>>> ludo@gnu.org (Ludovic Courtès) writes:
>>>
>>>> Mathieu Lirzin <mthl@gnu.org> skribis:
>>>>
>>>>> My previous snippet didn't include the actual error
>>>>>
>>>>> TEST: tests/test-hmp... (pid=27743)
>>>>>   /arm/hmp/integratorcp:                                               OK
>>>>>   /arm/hmp/nuri: qemu-system-arm: cannot set up guest memory
>>>>> 'exynos4210.dram0': Cannot allocate memory
>>>>> Broken pipe
>>>>> FAIL
>>>>> GTester: last random seed: R02S53adf2b44f1ff46cb48bd55ebce8854c
>>>>> (pid=27751)
>>>>>   /arm/hmp/mps2-an511:                                                 OK
>>>>>   /arm/hmp/verdex:                                                     OK
>>>>>   /arm/hmp/ast2500-evb:                                                OK
>>>>>   /arm/hmp/smdkc210: qemu-system-arm: cannot set up guest memory
>>>>> 'exynos4210.dram0': Cannot allocate memory
>>>>> Broken pipe
>>>>> FAIL
>>>>>
>>>>> I have tried to build QEMU again with the same guix version and succeed.
>>>>> So Like previously reported this is underterministic.
>>>>
>>>> Could the initial error be the result of insufficient memory?
>>>
>>> Maybe.  I don't recall if I was doing something memory intensive while
>>> compiling.  Given that I have 8Gb of RAM and that, QEMU test suite seems
>>> quite greedy.
>>
>> Ohh I think the memory issue might be related to the fact I running the
>> commands from Emacs shell-mode.  I have encoutered a memory issue with
>> Emacs when running ‘guix system build’.  After the compilation process
>> stopped Emacs was still using more than 5Gb of RAM.
>
> Oh yes, that could definitely be the culprit.  :-/
>
> I guess we can close this bug and reopen it if this diagnosis turned out
> to be wrong.
>
> Ludo’.

Closing this stale bug report against QEMU 2.  We have fresher ones
against QEMU 5 :-).

Maxim