[Xen-devel] 4.2 Release Plan / TODO

[Ian Campbell <Ian.Campbell@xxxxxxxxxx>]

Plan for a 4.2 release:
http://lists.xen.org/archives/html/xen-devel/2012-03/msg00793.html

The time line is as follows:

19 March        -- TODO list locked down
2 April         -- Feature Freeze
                                                << WE ARE HERE
Mid/Late April  -- First release candidate      << SEE BELOW
Weekly          -- RCN+1 until it is ready

My initial guesstimate for starting RCs appears to have been somewhat
optimistic. I think we need to have reduced at least the blockers lists
rather significantly before we start thinking of doing RCs.

A fairly large proportion of the list has been posted at least once, and
much of it seems ready (or almost ready) to go in. There are however
still some smaller items which are unclaimed. Based on feedback from
those who are working on the bigger outstanding items it seems like most
of them ought to be ready to land in the next few weeks.

On that basis I think doing a first release candidate in Mid/Late May is
more realistic. I'll reflect that in next weeks update

The updated TODO list follows. Per the release plan a strong case will
now have to be made as to why new items should be added to the list,
especially as a blocker, rather than deferred to 4.3.

hypervisor, blockers:
      * [BUG] Zombie driver domains.  There's a bug where PV guests with
        devices passed through with xl hang around as "zombies", with
        one outstanding page reference.  I think this should really be a
        blocker. I'm working on this at the moment (George Dunlap).
              * One of several recent reports of Zombie domains, which
                may or may not be all related.
              * List as hypervisor for now, but may turn out to be a
                tools issue.
        Fixed (DONE, Tim Deegan)
      * Performance regression due to contention on p2m lock. (Tim,
        Andres)
 
tools, blockers:
      * libxl stable API -- we would like 4.2 to define a stable API
        which downstream's can start to rely on not changing. Aspects of
        this are:
              * Safe fork vs. fd handling hooks. Involves API changes
                (Ian J, patches posted)
              * libxl_wait_for_free_memory/libxl_wait_for_memory_target.
                Interface needs an overhaul, related to
                locking/serialization over domain create (deferred to
                4.3). IanJ to add note about this interface being
                substandard but otherwise defer to 4.3.
              * libxl_{FOO}_exec. Should return a data structure
                declaring what to do, not fork and exec themselves.
                However we can defer this until 4.3 (therefore DONE with
                no work).
              * libxl_*_path. Majority made internal, only configdir and
                lockdir remain public (used by xl). Good enough?
              * Interfaces which may need to be async:
                      * libxl_domain_suspend. Probably need to move
                        xc_domain_save into a separate process, as per
                        <20366.40183.410388.447630@xxxxxxxxxxxxxxxxxxxxxxxx>. 
Likely need to do the same for xc_domain_restore too. I'm not sure if IanJ is 
working (or planning to work on) this.
                      * libxl_domain_create_{new,restore} -- IanJ has
                        patches as part of event series.
                      * libxl_domain_core_dump. Can take a dummy ao_how
                        and remain synchronous internally. (IanC, patch
                        posted)
                      * libxl_device_{disk,nic,vkb,add}_add (and
                        remove?). Roger Pau Monnà fixes disk as part of
                        hotplug script series and adds infrastructure
                        which should make the others trivial. (Roger
                        investigating)
                      * libxl_cdrom_insert. Should be easy once
                        disk_{add,remove} done, IanJ to look at (or
                        doing so?).
                      * libxl_device_disk_local_{attach,detach}. Become
                        internal as part of Stefano's domain 0 disk
                        attach series (patches posted)
                      * libxl_device_pci_{add,remove}. Roger
                        investigating along with other add,remove
                        functions.
                      * libxl_xen_tmem_*. All functions are "fast" and
                        therefore no async needed. Exception is
                        tmem_destroy which Dan Magenheimer says is
                        obsolete and can just be removed. (Ian C, patch
                        posted to remove tmem_destroy)
                      * libxl_fork -- IanJ's event series will remove
                        all users of this.
      * [BUG] Manually ballooning dom0.  xl mem-set 0 [foo] fails with
        "libxl: error: libxl.c:2569:libxl_set_memory_target: cannot get
        memory info from /local/domain/0/memory/static-max: No such file
        or directory". This might be suitable for an enthusiastic
        on-looker. (George Dunlap, in the absence of said onlooker)
      * xl compatibility with xm:
              * [BUG] cannot create an empty CD-ROM driver on HVM guest,
                reported by Fabio Fantoni in
                <4F9672DD.2080902@xxxxxxxxxx>
              * [BUG] does not honour scheduler weight params, reported
                by Dieter Bloms in <20120423193518.GA16134@xxxxxxxx>,
                Dieter has posted a patch.
      * More formally deprecate xm/xend. Manpage patches already in
        tree. Needs release noting and communication around -rc1 to
        remind people to test xl.
              * xl to refuse to run if xend is runnig, Roger Pau MonnÃ
                (patch posted)
      * Domain 0 block attach & general hotplug when using qdisk backend
        (need to start qemu as necessary etc) (Stefano S, patches
        posted)
      * file:// backend performance.
              * qemu-xen-traditional and upstream qemu-xen performance
                has been improved and is now satisfactory.
              * Xen 4.2 will prefer blktap2 if a kernel module is
                available (e.g. older dom0 kernel or DKMS provided
                kernel module) and will fallback to qemu qdisk if no
                blktap2 is available.
              * There will be no userspace "blktap3" for Xen 4.2
      * Improved Hotplug script support (Roger Pau MonnÃ, patches
        posted)
      * Block script support -- follows on from hotplug script (Roger
        Pau MonnÃ)

hypervisor, nice to have:
      * solid implementation of sharing/paging/mem-events (using work
        queues) (Tim Deegan, Olaf Herring et al -- patches posted)
              * "The last patch to use a waitqueue in
                __get_gfn_type_access() from Tim works.  However, there
                are a few users who call __get_gfn_type_access with the
                domain_lock held. This part needs to be addressed in
                some way."
              * Deferred until 4.3
      * Sharing support for AMD (Tim, Andres), in, marked as
        experimental (so, DONE, as far as 4.2 is concerned).
      * PoD performance improvements (George Dunlap)

tools, nice to have:
      * Configure/control paging via xl/libxl (Olaf Herring, lots of
        discussion around interface, general consensus reached on what
        it should look like. Olaf has let me know that he is probably
        not going to have time to do this for 4.2, will likely slip to
        4.3 unless someone else want to pick up that work)
              * Will defer until 4.3.
      * Upstream qemu feature patches:
              * Upstream qemu PCI passthrough support (Anthony Perard,
                patches sent)
              * Upstream qemu save restore (Anthony Perard, Stefano
                Stabellini, qemu patches applied, waiting for libxl etc
                side which has been recently reposted)
      * Nested-virtualisation. Currently "experimental". Likely to
        release that way.
              * Nested SVM. Tested in a variety of configurations but
                still some issues with the most important use case (w7
                XP mode) [0]  (Christoph Egger)
              * Nested VMX. Needs nested EPT to be genuinely useful.
                Need more data on testedness etc (Intel)
        DONE, at least as far as 4.2 is concerned.
      * Initial xl support for Remus (memory checkpoint, blackholing)
        (Shriram, waiting on libxl side of qemu upstream save/restore)
      * xl compatibility with xm:
              * xl support for autospawning vncviewer (vncviewer=1 or
                otherwise) (Goncalo Gomes, waiting on new version of
                patches)
              * support for vif "rate" parameter (Mathieu GagnÃ, first
                part applied)

[0] http://lists.xen.org/archives/html/xen-devel/2012-03/msg00883.html


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel