On Thu, Apr 09, 2020 at 07:52:27PM +0800, Zhenyu Zheng wrote:
> Hi,
>
> Thanks alot for all the reviews, I'm updating the codes, as for those
> headers,
> hwcap.h is here:
> https://github.com/torvalds/linux/blob/v5.6/arch/arm/include/asm/hwcap.h and
> should be available for 5.6,
> auxv.h is from glibc:
> https://sourceware.org/git/?p=glibc.git;a=blob;f=misc/sys/auxv.h;h=1a563e1337e64e4218c1613b68dc2551a762ba00;hb=HEAD

The key requirement from libvirt is that code must compile against our
declared set of supported platforms

    https://libvirt.org/platforms.html

Usually RHEL-7 is the oldest platform that causes trouble.

If code can't be made to compile on old platforms, then it is acceptable
to use conditional compilation to disable the code.

If you want to validate your patches build on all our required platforms,
you can take advantage of our recent switch to GitLab

Fork the libvirt repo:

  https://gitlab.com/libvirt/libvirt

and push your patches to a branch in your personal fork. This will trigger
our automated build CI jobs across all important platforms.

As an example here's a recent CI job run:

  https://gitlab.com/libvirt/libvirt/pipelines/134381772

Note, however, that most of the jobs are using x86, so if you have any
conditionally compiled arm code, we don't have good coverage for arm
on the old distros, so that may benefit from manual testing.

> > virCPUarmGetHost and the helpers it calls are architecture specific and
> > should not be even compiled in on non-ARM.
>
>  yeah, understandable, I just don't see other archs like ppc and s390 doing
> this so I didn't do it.

Regards,
Daniel
--
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|