=== doko__ is now known as doko [09:05] could someone take a look at signon in the unapproved queue please? it’s holding off other landings [09:36] please could a release team member take a look at bug 1377218 [09:36] bug 1377218 in ceilometer (Ubuntu) "[FFe] ceilometer 2014.2.rc1" [High,New] https://launchpad.net/bugs/1377218 [09:36] I'd like to get that into the review queue today if possible please === Pici` is now known as Guest66735 === rbasak_ is now known as rbasak === arges` is now known as arges [12:34] hi; can you inspect the signon package which is currently in the unapproved queue? it's coming from a phone ci landing; thanks [12:48] can someone accept mesa? contains some fixes for mir === barry` is now known as barry === jhodapp_ is now known as jhodapp === Ursinha-afk is now known as Ursinha [14:31] arges: hey, just a note, the fix for bug 1374617 is required for the test cases for bugs 1374622 and 1374612 [14:31] bug 1374617 in Ubuntu Trusty "[FFE] New package: ipxe-precise" [High,Confirmed] https://launchpad.net/bugs/1374617 [14:31] bug 1374622 in libvirt (Ubuntu Trusty) "Support migration from 12.04" [High,Fix committed] https://launchpad.net/bugs/1374622 [14:31] bug 1374612 in qemu (Ubuntu Precise) "[FFE] add pc-1.0-precise machine type" [High,Confirmed] https://launchpad.net/bugs/1374612 [14:48] hallyn: did you upload the ipxe fix yet? [14:50] arges: I thought i did. [14:50] hallyn: ah its in the NEW queeu [14:50] queue [14:51] yup. if that needs more time i can do the verificatoins with the pkgs out of ppa, but that feels like cheating :) [14:51] hallyn: you should ping an AA about approving that package. cjwatson / infinity ^^^ ipxe-precise in trusty/new queue [14:54] arges: oh, ok, thanks. infinity: the ipxe-precise is the last bit for the p->t qemu vm migration drama [16:02] hi again; this is a quick nudge to ask for attention on the 'signon' package in the unapproved queue [16:31] dbarth: any bug number or other pointer to why it'll fix more problems than it'll add? [16:36] sorry === georgelorch2 is now known as georgelorch [16:46] Riddell: not really; the change is here and is meant to secure access from client apps better [16:46] Riddell: https://code.launchpad.net/~online-accounts/signon/packaging/+merge/237211 [16:47] i guess i can ask mardy to comment on the original merge prop. to make a stronger case for it not creating compatibility issues for example === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [18:55] hallyn: That s/ppc/ppc64/ thinko in your qemu changelog scared me. [18:55] infinity: well it was overly conservative, not overly liberal :) [18:56] was going to refuse running ppc64el-user under ppc64 [18:56] hallyn: No, no. The typo. Where you said you removed "Support-ppc64le.patch" [18:56] hallyn: You meant ppcle. [18:56] oh [18:56] feh, yeah [18:57] those machine names are getting almost as nnoying as arm64/amd64 [18:57] *smirk* [18:57] i keep d/ling the wrong rootfs's now, thanks to arm64 [18:57] Anyhow, I suspect IBM (and others) will push back hard against people doing 32-bit LE or, at least, make sure they do it right, so dropping the premature support seems sane. [19:04] hallyn: For your peace of mind, though, I can confirm that the be/le changes you just made are also correct. [19:04] hallyn: ppc kernels are endian-specific, so running le on be or be on le isn't native. [19:04] infinity: ok, thx. (slangasek had also confirmed it on #debian-qemu) [19:05] hallyn: Check. [19:05] as for the 32-bit le, i figured better to drop it before release [19:05] hallyn: I've praised IBM repeatedly for doing the right thing there. They debated (for a long time) doing endian-neutral kernels with duplicated syscall tables, and I'm so glad they didn't. It's a maintenance nightmare. [19:05] (See x86/x86_64/x32, or the horror that is mips) [19:16] the thing that goes mips in the night [19:17] the x86 model is wrong, the syscalls should be associated with different kernel personalities [19:20] wouldn't that remove our compat syscall exploit fun? [19:23] slangasek: Don't even get me started on personalities, and their completely inconsistent application across architectures. [19:23] ok then I won't [19:23] slangasek: I feel strongly that's something someone should fix, but I also suspect it would require inventing a time machine. [19:23] also: ICBS [19:24] (which I'm pretty sure stood for intercontinental bowel syndrome?) [19:24] Inter-Continental Ballistic Suck? [19:24] http://www.linuxmisc.com/1-linux-setup/f2cd5ed0c51dd5d6.htm [19:26] dat stoppen. Hieronder leest u hoe u dat kunt doen. [19:26] Hoe herken ik geadresseerde reclame? [19:27] oops [19:27] I'm not going to bother translating that, and just assume it was erotic. [19:29] Oh dear. I went and translated it anyway, and the first line is so much better after my comment. [19:29] Could someone on the release team please manually approve Wine1.6 for promotion for me? Britney gets confused because the amd64 version has a cross-arch dependency. [19:29] YokoZar: Oh, let me have a look. [19:31] infinity: Thank you :) [19:31] YokoZar: Fix committed, should migrate in the next run, if nothing else is broken. [19:31] infinity: Excellent, thanks. Will I need to do this every upload? [19:32] YokoZar: Until we come up with a better way to figure this out, yeah. A ping wouldn't hurt. === bjf is now known as bjf[afk] [20:07] arges: infinity: so I went ahead and did the upgrade tests for qemu/qemu-kvm (and about to for libvirt) using the kvm-ipxe-precise from ppa. still do need the ipxe-precise package in trusty to really call it fixed though. (though the situation without that is no worse than the situation without these current qemu+libvirt packages, so technically they don't have to be blocked by it) [20:07] * hallyn waiting on two parallel long-asse qa-regression-tests on his laptop [20:08] hallyn: cool === Ursinha is now known as Ursinha-afk === Ursinha_ is now known as Ursinha