Test failures on macOS 12

[Christophe de Dinechin]

> On 9 Aug 2022, at 11:50, Andrea Bolognani <abologna at redhat.com> wrote:
> 
> On Tue, Aug 09, 2022 at 11:34:20AM +0200, Christophe de Dinechin wrote:
>> On 9 Aug 2022, at 11:28, Andrea Bolognani <abologna at redhat.com> wrote:
>>> Yeah, this seems to help and the change makes sense to me.
>>> 
>>> I wonder why we didn't run into this much earlier though? As I
>>> mentioned, the test runs successfully as-is on macOS 11. Plus, many
>>> other tests rely on library injection and yet work okay even without
>>> this change.

I must admit that this puzzled me a bit too. I spent a bit of time checking
for dyld warnings or anything else.

One explanation could be if in other cases, the symbols are marked as weak?
I did not check that. And I don’t have a macOS 11 machine to compare anymore.

>>> 
>>> Anyway, I'm happy to add my
>>> 
>>> Reviewed-by: Andrea Bolognani <abologna at redhat.com>
>>> 
>>> to this patch and push it. The authorship information looks a bit
>>> funky though, with the two S-o-bs...
>> 
>> I did not know which one you’d prefer (in case there is a policy).
>> If I get to choose, assign that to Red Hat (and change the author accordingly).
>> 
>> (and I’ll change my libvirt gitconfig accordingly in the future)
> 
> Done. I'll push once CI has passed.
> 
> It would be great if you could use git-publish for future code
> submissions: that way patches can be applied locally more
> conveniently by the reviewer. I was able to make it work regardless,
> it just took a bit more effort :)

Ack.

https://gitlab.com/c3d/libvirt/-/pipelines/608168172

> 
> -- 
> Andrea Bolognani / Red Hat / Virtualization
>