[libvirt] Entering freeze for libvirt-1.1.0

Since Dan pushed the fine grained ACL control patches, this release really ought to be 1.1.0, I just tagged it in git and pushed tarball and rpms to the server: ftp://libvirt.org/libvirt/ I know that Laine has still a patch set to push, now it seems they all got ACK'ed so better get them in for rc2, which I will probably push on thursday. Then if everything looks fine we can have the final release on Monday. I gave it my usual user test, on Fedora, seems to work fine from my limited perspective. I just had to push the attached patch as a build breaker otherwise make rpm would wail because the polkit API definitions would not be packaged. Probably worth a double checking by Dan for correctness. Thanks in advance for feedback on the candidate release, Daniel -- Daniel Veillard | Open Source and Standards, Red Hat veillard@redhat.com | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | virtualization library http://libvirt.org/

I have just tagged the release candidate 2 in git and sent a tarball to the usual place (rpms are coming): ftp://libvirt.org/libvirt/ This includes the end of the patch set from Laine, and hopefully it won't require too many other patches. I tried it and it doesn't look obviously broken to me, please give it a try too, especially for portability :-) If all goes well the final release should be next Monday ! thanks ! Daniel -- Daniel Veillard | Open Source and Standards, Red Hat veillard@redhat.com | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | virtualization library http://libvirt.org/

On Thu, Jun 27, 2013 at 11:01 AM, Daniel Veillard <veillard@redhat.com>wrote:
I have just tagged the release candidate 2 in git and sent a tarball to the usual place (rpms are coming): ftp://libvirt.org/libvirt/
This includes the end of the patch set from Laine, and hopefully it won't require too many other patches. I tried it and it doesn't look obviously broken to me, please give it a try too, especially for portability :-) If all goes well the final release should be next Monday !
thanks !
Daniel
-- Daniel Veillard | Open Source and Standards, Red Hat veillard@redhat.com | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | virtualization library http://libvirt.org/
-- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Not sure if this is a regression yet, but migrations are failing for me on git (currently running both machines at the v1.1.0_rc2 tag) virsh migrate --live --p2p --persistent --copy-storage-all expo qemu+tcp://cocacola/system error: Failed to open file '/var/lib/libvirt/images/expo.img': No such file or directory The error message is from the remote that the file isn't there. I was under the impression that a NBD starts up and streams the data over. It gives me an error to hint that if I can include --tunnelled in the options. Using qemu 1.4.2 for reference (its Cole's patchset for 1.4.2 from Fedora's qemu.git) -- Doug Goldstein

On Thu, Jun 27, 2013 at 11:01 AM, Daniel Veillard <veillard@redhat.com>wrote:
I have just tagged the release candidate 2 in git and sent a tarball to the usual place (rpms are coming): ftp://libvirt.org/libvirt/
This includes the end of the patch set from Laine, and hopefully it won't require too many other patches. I tried it and it doesn't look obviously broken to me, please give it a try too, especially for portability :-) If all goes well the final release should be next Monday !
thanks !
Daniel
So far, I am getting linker errors for FreeBSD here: https://redports.org/buildarchive/20130628070800-42595/
-jgh -- Jason Helfman | FreeBSD Committer jgh@FreeBSD.org | http://people.freebsd.org/~jgh | The Power to Serve

On 06/28/2013 03:24 AM, Jason Helfman wrote:
On Thu, Jun 27, 2013 at 11:01 AM, Daniel Veillard <veillard@redhat.com <mailto:veillard@redhat.com>> wrote:
I have just tagged the release candidate 2 in git and sent a tarball to the usual place (rpms are coming): ftp://libvirt.org/libvirt/
This includes the end of the patch set from Laine, and hopefully it won't require too many other patches. I tried it and it doesn't look obviously broken to me, please give it a try too, especially for portability :-) If all goes well the final release should be next Monday !
thanks !
Daniel
So far, I am getting linker errors for FreeBSD here: https://redports.org/buildarchive/20130628070800-42595/
Sigh. I see the problem. Patch coming up...

On 06/28/2013 03:44 AM, Laine Stump wrote:
On 06/28/2013 03:24 AM, Jason Helfman wrote:
On Thu, Jun 27, 2013 at 11:01 AM, Daniel Veillard <veillard@redhat.com <mailto:veillard@redhat.com>> wrote:
I have just tagged the release candidate 2 in git and sent a tarball to the usual place (rpms are coming): ftp://libvirt.org/libvirt/
This includes the end of the patch set from Laine, and hopefully it won't require too many other patches. I tried it and it doesn't look obviously broken to me, please give it a try too, especially for portability :-) If all goes well the final release should be next Monday !
thanks !
Daniel
So far, I am getting linker errors for FreeBSD here: https://redports.org/buildarchive/20130628070800-42595/
Sigh. I see the problem. Patch coming up...
Okay, I pushed the following patch: https://www.redhat.com/archives/libvir-list/2013-June/msg01171.html Can you apply this patch locally to the source tar and re-run your test build to make sure nothing else is broken (since there won't be another rc before release)? If there are still problems, you can find us in #virt on irc.oftc.net. commit a757822233f707c4ed75986f5903e26e40f3cdfa Author: Laine Stump <laine@laine.org> Date: Fri Jun 28 04:00:54 2013 -0400 util: fix build error on non-Linux systems Building on FreeBSD had this linker error: /work/a/ports/devel/libvirt/work/libvirt-1.1.0/src/.libs/libvirt.so: undefined reference to `virPCIDeviceAddressParse' This was caused by the new use of virPCIDeviceAddressParse in a portion of virpci.c that wasn't linux-only (in commit 72c029d8). The problem was that virPCIDeviceAddressParse had originally been defined inside #ifdef _linux (because it was only used by another function that was inside the same ifdef). The solution is to move it out to the part of virpci.c that is compiled on all platforms. (Because the portion that was "moved" was 40-50 lines, but only moved up by 15 lines, the diff for the patch is less than non-informative - rather than showing that part that I moved, it shows the bit that was previously before the moved part, and now sits *after* it.)

On Thu, Jun 27, 2013 at 9:46 PM, Doug Goldstein <cardoe@gentoo.org> wrote:
On Thu, Jun 27, 2013 at 11:01 AM, Daniel Veillard <veillard@redhat.com>wrote:
I have just tagged the release candidate 2 in git and sent a tarball to the usual place (rpms are coming): ftp://libvirt.org/libvirt/
This includes the end of the patch set from Laine, and hopefully it won't require too many other patches. I tried it and it doesn't look obviously broken to me, please give it a try too, especially for portability :-) If all goes well the final release should be next Monday !
thanks !
Daniel
-- Daniel Veillard | Open Source and Standards, Red Hat veillard@redhat.com | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | virtualization library http://libvirt.org/
-- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Not sure if this is a regression yet, but migrations are failing for me on git (currently running both machines at the v1.1.0_rc2 tag)
virsh migrate --live --p2p --persistent --copy-storage-all expo qemu+tcp://cocacola/system error: Failed to open file '/var/lib/libvirt/images/expo.img': No such file or directory
The error message is from the remote that the file isn't there. I was under the impression that a NBD starts up and streams the data over. It gives me an error to hint that if I can include --tunnelled in the options.
Using qemu 1.4.2 for reference (its Cole's patchset for 1.4.2 from Fedora's qemu.git)
-- Doug Goldstein
Sorry for the noise. I apparently have an instance still running Michal's patchset that provided that functionality on one of my machines without realizing it. -- Doug Goldstein

On 27/06/2013, at 5:01 PM, Daniel Veillard wrote:
I have just tagged the release candidate 2 in git and sent a tarball to the usual place (rpms are coming): ftp://libvirt.org/libvirt/
Sorry for the delay, only just now remembered to test this. :) Compilation is failing on OSX 10.7 x64 with: ************************************************************** CC libvirt_util_la-virnetdevveth.lo CC libvirt_util_la-virnetdevvlan.lo CC libvirt_util_la-virnetdevvportprofile.lo CC libvirt_util_la-virnetlink.lo CC libvirt_util_la-virnodesuspend.lo CC libvirt_util_la-virnuma.lo util/virnetdevtap.c:315:26: error: use of undeclared identifier 'IF_MAXUNIT' for (i = 0; i <= IF_MAXUNIT; i++) { ^ 1 error generated. make[3]: *** [libvirt_util_la-virnetdevtap.lo] Error 1 make[3]: *** Waiting for unfinished jobs.... [all] Error 2 make[1]: *** [all-recursive] Error 1 make: *** [all] Error 2 ************************************************************** There doesn't seem to be a definition for IF_MAXUNIT on OSX (10.7 anyway): $ grep -ri IF_MAXUNIT /usr/include/ $ Interestingly, it's also not on a RHEL 6.4 box locally here either: $ grep -ri "IF_MAXUNIT" /usr/include/ $ Though, it's possible that box is missing required devel packages for libvirt. (unsure) Googling shows IF_MAXUNIT in context of FreeBSD stuff. Not sure of the right approach here. eg exclude compiling this on OSX, or alternatively make-it-work on OSX. Hope that helps. :) + Justin -- Open Source and Standards @ Red Hat twitter.com/realjustinclift

On Fri, Jun 28, 2013 at 03:01:58PM +0100, Justin Clift wrote:
On 27/06/2013, at 5:01 PM, Daniel Veillard wrote:
I have just tagged the release candidate 2 in git and sent a tarball to the usual place (rpms are coming): ftp://libvirt.org/libvirt/
Sorry for the delay, only just now remembered to test this. :)
Compilation is failing on OSX 10.7 x64 with:
************************************************************** CC libvirt_util_la-virnetdevveth.lo CC libvirt_util_la-virnetdevvlan.lo CC libvirt_util_la-virnetdevvportprofile.lo CC libvirt_util_la-virnetlink.lo CC libvirt_util_la-virnodesuspend.lo CC libvirt_util_la-virnuma.lo util/virnetdevtap.c:315:26: error: use of undeclared identifier 'IF_MAXUNIT' for (i = 0; i <= IF_MAXUNIT; i++) {
[snip]
Googling shows IF_MAXUNIT in context of FreeBSD stuff. Not sure of the right approach here. eg exclude compiling this on OSX, or alternatively make-it-work on OSX.
Yes, this is from code that is only designed to work on FreeBSD. I guess the OS-X BSD base isn't close enough to support this properly. Try this patch diff --git a/src/util/virnetdevtap.c b/src/util/virnetdevtap.c index 265676c..4e2f32a 100644 --- a/src/util/virnetdevtap.c +++ b/src/util/virnetdevtap.c @@ -275,7 +275,7 @@ cleanup: VIR_FORCE_CLOSE(fd); return ret; } -#elif defined(SIOCIFCREATE2) && defined(SIOCIFDESTROY) +#elif defined(SIOCIFCREATE2) && defined(SIOCIFDESTROY) && defined(IF_MAXUNIT) int virNetDevTapCreate(char **ifname, int *tapfd, int tapfdSize, Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|

On 28/06/2013, at 3:05 PM, Daniel P. Berrange wrote:
On Fri, Jun 28, 2013 at 03:01:58PM +0100, Justin Clift wrote: <snip>
Googling shows IF_MAXUNIT in context of FreeBSD stuff. Not sure of the right approach here. eg exclude compiling this on OSX, or alternatively make-it-work on OSX.
Yes, this is from code that is only designed to work on FreeBSD. I guess the OS-X BSD base isn't close enough to support this properly.
Try this patch
diff --git a/src/util/virnetdevtap.c b/src/util/virnetdevtap.c index 265676c..4e2f32a 100644 --- a/src/util/virnetdevtap.c +++ b/src/util/virnetdevtap.c @@ -275,7 +275,7 @@ cleanup: VIR_FORCE_CLOSE(fd); return ret; } -#elif defined(SIOCIFCREATE2) && defined(SIOCIFDESTROY) +#elif defined(SIOCIFCREATE2) && defined(SIOCIFDESTROY) && defined(IF_MAXUNIT) int virNetDevTapCreate(char **ifname, int *tapfd, int tapfdSize,
Thanks Dan, that works well. Libvirt is now compiling, and virsh seems to be working happily (with a minimal amount of testing anyway). :) Regards and best wishes, Justin Clift -- Open Source and Standards @ Red Hat twitter.com/realjustinclift
participants (6)
-
Daniel P. Berrange
-
Daniel Veillard
-
Doug Goldstein
-
Jason Helfman
-
Justin Clift
-
Laine Stump