On Thu, Nov 05, 2015 at 17:33:52 +0000, Daniel Berrange wrote:
The patches for introducing virtlogd will be significantly simplified if we don't need to worry about parsing stderr during startup. This is required prior to QEMU 0.11 so that we can get the dyanamically allocated /dev/pty/NNN paths.
I'm so glad to see something like this. Not only to simplify adding virtlogd but also we will be able to remove some _very_ old cruft.
The QEMU 0.12.1 release was shipped in RHEL-6 vintage distros and is already quite old, so seems like a fair target version to aim for as the minimum required.
I'd like to add that I think this is still too vintage. It's now almost 6 years from that point: commit 6c412ddf1cc0c41a7c36064a4a9c428e99c52ff8 Author: Anthony Liguori <aliguori@us.ibm.com> Date: Sat Dec 19 08:23:00 2009 -0600 Update for 0.12.0 release and right after that: commit fe1b69708c72b163d3acdf2bb012e169d2d3dda0 Author: Anthony Liguori <aliguori@us.ibm.com> Date: Sat Dec 19 19:31:18 2009 -0600 Update version and changelog for 0.12.1 The initial qemu version in RHEL-6 was indeed 0.12.1, but if you look at the current state of it you'll notice that it was patched quite a lot so it rather stopped to resemble the 0.12 release and was really holding up with upstream for quite a while. Libvirt was even rebased during the current lifetime of rhel/centos 6 distros to 0.10.2 (and it has quite a few patches on top of that too). Libvirt 0.10.2 was released almost 3 years after qemu 0.12.0: commit f8fbeb50d52520a109d71c8566fed2ea600650ec Author: Daniel Veillard <veillard@redhat.com> Date: Mon Sep 24 12:06:05 2012 +0800 Release of libvirt-0.10.2 So while having rhel-6 as a support target might look cool, nobody would actually use such old code anyways, since they can get tested and patched packages with a ton of new features. By being a bit more aggresive we could also cut off the need to use QEMU_CAPS_DEVICE (as you've pointed out further down) and possibly even mandate the use of QMP which was initially added by qemu 0.12.0 but I remember that there were some problems with that. I'm realizing though that picking the line where we cut support is really hard, but I think it will pay off eventually. Even with the proposed state.
By dropping support for anything older than QEMU 0.12.0 we can remove the code for parsing stderr. The QEMU 0.12.0
I'm so happy to see all that code go away. Peter