On 03/21/2018 09:03 AM, Andrea Bolognani wrote:
On Wed, 2018-03-21 at 07:34 -0400, John Ferlan wrote:
create mode 100644 tests/qemucapabilitiesdata/caps_2.12.0-gicv2.aarch64.replies create mode 100644 tests/qemucapabilitiesdata/caps_2.12.0-gicv3.aarch64.replies create mode 100644 tests/qemucapabilitiesdata/caps_2.12.0.ppc64.replies create mode 100644 tests/qemucapabilitiesdata/caps_2.12.0.x86_64.replies
What, not adventurous enough for s390 too?! :-)
I usually generate capabilities data on the native architecture, as that makes a diffence at least for aarch64. I've never bothered obtaining access to an s390x machine, but I guess it shouldn't be too difficult :)
While 2.12 isn't "complete" yet, we may as well update now and then try to remember later on when 2.12 finalizes to do the final update. I was off trying to recollect the process of generating the files ;-)
We have hit feature freeze for QEMU. A few pull requests are still pending to be merged, but if you wait until middle of next week, the capabilities should not change again before release unless bad bugs are found.
I see that 2.12 hit rc0... since there's a few series upstream that are essentially waiting for the real 2.12 replies/xml files - perhaps it'd be good to just update this patch and get it pushed...
I was not aware of other series being held up by this, but if there is indeed any I can definitely collect the data again using v2.12.0-rc0 and push.
Held up is perhaps a strong word, but there's a couple series where 2.12 is where a feature was added (sev - secure encrypted vms, chardev-fd-pass, query-cpus-fast, and of course 'this' series). Then there's the virtio-gpu-ccw that'll be impacted by changes or merges w/ qemu_capabilities. I know I have an unposted series that will have merge conflicts too... John