Re: [PATCH v2] meson: Propagate gnutls dependency to migration

[Daniel P . Berrangé]

On Tue, Mar 23, 2021 at 04:04:48PM +0100, Juan Quintela wrote:
> Jessica Clarke <jrtc27@jrtc27.com> wrote:
> > Commit 3eacf70bb5a83e4775ad8003cbca63a40f70c8c2 neglected to fix this
> > for softmmu configs, which pull in migration's use of gnutls.
> >
> > This fixes the following compilation failure on Arm-based Macs:
> >
> >   In file included from migration/multifd.c:23:
> >   In file included from migration/tls.h:25:
> >   In file included from include/io/channel-tls.h:26:
> >   In file included from include/crypto/tlssession.h:24:
> >   include/crypto/tlscreds.h:28:10: fatal error: 'gnutls/gnutls.h' file not 
> > found
> >   #include <gnutls/gnutls.h>
> >            ^~~~~~~~~~~~~~~~~
> >   1 error generated.
> >
> > (as well as for channel.c and tls.c)
> >
> > Signed-off-by: Jessica Clarke <jrtc27@jrtc27.com>
> 
> Reviewed-by: Juan Quintela <quintela@redhat.com>
> 
> I would preffer to be the other way around, i.e. modularize tls out of
> migration, but a fast look at it shows it as a big job.  As said on the
> commint message, channel.c & tls.c both use tls not conditionally.
> Once there, multifd.c also requires tls compiled in.
> Can we realistically build qemu without tls nowadays?

The intention is that use of TLS is conditionalized *only* in the
crypto/tls*.c and include/crypto/tls*.h files.

The rest of QEMU is supposed to use TLS unconditionally, and will
simply get an error message at runtime if it was unavailable. This
is precisely to avoid littering the whole codebase in QEMU with
conditionals on GNUTLS.


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 :|