Planning/2.7: Difference between revisions
Line 41: | Line 41: | ||
* doesn't build on Windows with older glib versions | * doesn't build on Windows with older glib versions | ||
* 9pfs filename sanitization | * 9pfs filename sanitization | ||
* PPC regression on prep ? | |||
=== Will be fixed in rc5 === | === Will be fixed in rc5 === |
Revision as of 15:44, 25 August 2016
Release Schedule
2016-05-12 | Beginning of development phase |
2016-06-28 | Soft feature freeze. All features
should have patches on the list by this date; major features should have initial code committed. |
Hard feature freeze. Tag -rc0,
only bug fixes committed after this point | |
Tag v2.7.0-rc1 | |
Tag v2.7.0-rc2 | |
Tag v2.7.0-rc3 | |
2016-08-22 | Tag v2.7.0-rc4 |
Tag v2.7.0 |
Known issues
Please list all patch series or bugs which need to be fixed for the release here. That way I can be sure I don't miss any when I'm rolling release candidates.
Also see Planning/2.7/Testing.
Not yet fixed in master
- virtio balloon bug of some description??
- doesn't build on Windows with older glib versions
- 9pfs filename sanitization
- PPC regression on prep ?
Will be fixed in rc5
- Compile failure in bsd-user
- virtio bugs with vq->inuse counts
Fixed in rc4
- commit 9c37146782e785 accidentally disabled a lot of the linux-user netlink support code, because the headers don't define IFLA_BR_* and IFLA_BRPORT_* as C preprocessor constants, so #ifdef IFLA_BRPORT_ID etc don't do what you might hope
- test-logging testcase leaves a lot of logfiles in /tmp/
Fixed in rc3
- Doesn't build on x86 OpenBSD or NetBSD. (Patch on list)
- Slirp code doesn't build on OpenBSD (see https://bugs.launchpad.net/qemu/+bug/1613133)
- Broken build of virtio-gpu if 'log' trace backend is not enabled (https://lists.gnu.org/archive/html/qemu-devel/2016-08/msg01252.html)
- Regression broken TLS support with chardevs (https://lists.gnu.org/archive/html/qemu-devel/2016-08/msg02824.html)
Fixed in rc2
wxx: Truncate files used for character devices
Will not be fixed for 2.7
- trying to attach a vhost-user device causes a segfault -- this is not a regression (was present in 2.6), so we'll fix it in 2.8
Targeted Features
See the ChangeLog/2.7 for full details.