On Tue, 7 Jan 2025 at 03:57, Orion Poplawski <orion@xxxxxxxx> wrote:
On 1/6/25 08:48, Kevin Fenzi wrote:
> On Mon, Jan 06, 2025 at 12:17:09PM +0000, Richard W.M. Jones wrote:
>>
>> valgrind works by interpreting each instruction, so yes the overhead
>> is rather large. On the other hand I really appreciate its ability to
>> find bugs.
>>
>> >From the errors below it seems to be a problem with Neoverse code
>> (https://en.wikipedia.org/wiki/ARM_Neoverse) which from my
>> understanding is not related to the low end hardware I have at home
>> (Raspberry Pis and similar).
>>
>> I think you'd need access to an Ampere or AWS Graviton system :-(
>>
>> qemu is able to emulate Neoverse N1, N2 & V1, so running a software
>> emulated qemu-system-aarch64 virtual machine may be the way to go.
>> (https://fedoraproject.org/wiki/Architectures/AArch64/Install_with_QEMU)
>
> We do have some maintainer test machines if that would help:
>
> https://fedoraproject.org/wiki/Test_Machine_Resources_For_Package_Maintainers
The problem as I noted in my original email is that it does not crash on
the test machine. So far it only crashes in koji.
FWIW, I've reproduced this on the test machine by setting OMP_NUM_THREADS=12.
Iñaki Úcar
-- _______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
