Libguestfs - Apr 2018 - Re: [nbdkit PATCH] tests: Skip guestfs tests on CentOS 6

On 04/06/2018 04:26 AM, Richard W.M. Jones wrote:
> On Thu, Apr 05, 2018 at 03:03:41PM -0500, Eric Blake wrote:
>> CentOS 6 has libguestfs-devel 1.20.11, which predates the support
>> in guestfs_add_drive_opts() for requesting an nbd drive instead
>> of a local file.  The guestfs plugin can still be built, so no
>> configure changes are needed; but skip the tests that fail to
>> compile so that 'make check' can at least complete.
>>
>> Signed-off-by: Eric Blake <eblake@redhat.com>
> 
> 
>> +#ifndef GUESTFS_ADD_DRIVE_OPTS_PROTOCOL
> 
> There's a symbol defined for this kind of test:
> 
> #ifndef GUESTFS_HAVE_ADD_DRIVE_OPTS
>   ...
> 
> (from <guestfs.h>).
The libguestfs on CentOS 6 does define GUESTFS_HAVE_ADD_DRIVE_OPTS (and
the function family guestfs_add_drive_opts*), however, it only supports
up through option 5 GUESTFS_ADD_DRIVE_OPTS_CACHEMODE.  The compile
failure was that GUESTFS_ADD_DRIVE_OPTS_PROTOCOL was undefined.  (The
documentation at http://libguestfs.org/guestfs.3.html#guestfs_add_drive
doesn't list which version various options were added in, only that the
function itself was added in 0.3, which predates 1.20.11).
> 
> However in this case I wonder if it would be better to test for this
> in the ./configure script and disable the HAVE_LIBGUESTFS automake
> conditional (but probably not error out -- we want to still compile on
> RHEL 6).  What do you think?
Might be easier; I'll play with it, and post a v2 patch along those
lines (certainly a smaller patch, but it does mean that we won't even
attempt to build plugins/guestfs - I guess that's still okay for an
older release).

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org

On Fri, Apr 06, 2018 at 09:02:01AM -0500, Eric Blake
wrote:
> On 04/06/2018 04:26 AM, Richard W.M. Jones wrote:
> > On Thu, Apr 05, 2018 at 03:03:41PM -0500, Eric Blake wrote:
> >> CentOS 6 has libguestfs-devel 1.20.11, which predates the support
> >> in guestfs_add_drive_opts() for requesting an nbd drive instead
> >> of a local file.  The guestfs plugin can still be built, so no
> >> configure changes are needed; but skip the tests that fail to
> >> compile so that 'make check' can at least complete.
> >>
> >> Signed-off-by: Eric Blake <eblake@redhat.com>
> > 
> > 
> >> +#ifndef GUESTFS_ADD_DRIVE_OPTS_PROTOCOL
> > 
> > There's a symbol defined for this kind of test:
> > 
> > #ifndef GUESTFS_HAVE_ADD_DRIVE_OPTS
> >   ...
> > 
> > (from <guestfs.h>).
> 
> The libguestfs on CentOS 6 does define GUESTFS_HAVE_ADD_DRIVE_OPTS (and
> the function family guestfs_add_drive_opts*), however, it only supports
> up through option 5 GUESTFS_ADD_DRIVE_OPTS_CACHEMODE.  The compile
> failure was that GUESTFS_ADD_DRIVE_OPTS_PROTOCOL was undefined.  (The
> documentation at http://libguestfs.org/guestfs.3.html#guestfs_add_drive
> doesn't list which version various options were added in, only that the
> function itself was added in 0.3, which predates 1.20.11).
Oh I see, yes that makes more sense.
> > However in this case I wonder if it would be better to test for this
> > in the ./configure script and disable the HAVE_LIBGUESTFS automake
> > conditional (but probably not error out -- we want to still compile on
> > RHEL 6).  What do you think?
> 
> Might be easier; I'll play with it, and post a v2 patch along those
> lines (certainly a smaller patch, but it does mean that we won't even
> attempt to build plugins/guestfs - I guess that's still okay for an
> older release).
Rich.

-- 
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming and virtualization blog: http://rwmj.wordpress.com
Fedora Windows cross-compiler. Compile Windows programs, test, and
build Windows installers. Over 100 libraries supported.
http://fedoraproject.org/wiki/MinGW

On 04/06/2018 09:05 AM, Richard W.M. Jones wrote:
>>> However in this case I wonder if it would be better to test for
this
>>> in the ./configure script and disable the HAVE_LIBGUESTFS automake
>>> conditional (but probably not error out -- we want to still compile
on
>>> RHEL 6).  What do you think?
>>
>> Might be easier; I'll play with it, and post a v2 patch along those
>> lines (certainly a smaller patch, but it does mean that we won't
even
>> attempt to build plugins/guestfs - I guess that's still okay for an
>> older release).
Okay, done in v2 (much smaller patch, once I figured out where in
configure.ac to add the compile check).

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org