13:11:18 #startmeeting pci_meeting 13:11:19 Meeting started Tue Sep 9 13:11:18 2014 UTC and is due to finish in 60 minutes. The chair is irenab. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:11:20 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:11:22 The meeting name has been set to 'pci_meeting' 13:11:24 Hi 13:11:45 baoli: I just started the meeting 13:11:46 hi 13:11:51 sorry, working on patches. forgot about the meeting 13:12:00 i late also 13:12:04 networking isuure 13:12:35 lets make just short update if you do not have any more issue for today, ok? 13:13:07 i notice the request object patch give exception while no pci request in the instance, i think this is might overkill, i leave comment on that 13:13:13 I started integrating the patches with the refactoring, hopefully I can have all the patches up end of the day 13:13:45 baoli: I'll follow, once you upload, will review 13:13:50 heyongli: I addressed that in the wrong patch, and I just corrected a moment ago 13:14:03 irenab, that would be great, thanks! 13:14:36 heyongli: I noticed that jenkins failure was caused by the exception 13:15:19 heyongli, irenab, have you guys tested resize with PCI devices? 13:15:50 baoli: no, just usual create/erminate flows 13:15:50 resize have been test ever 13:16:02 but not test on sriov 13:16:16 baoli: do you expect some problem there? 13:16:34 heyongli: I'm not sure it's due to my testbed, but resize doesn't work with the generic pci passthrough 13:16:53 baoli: taking the patches today, should it work? 13:16:57 ooh, maybe 13:17:43 long time ago it been tested, it seems broken if not working now 13:17:57 irenab, given all the changes coming in, I think the sr-iov patches may not work any more. So wait until I have all the updates pushed up. 13:18:28 baoli: can you please send some email once all patches are up? 13:18:46 irenab: sure 13:19:16 even if there is a problem with resize, we can follow with bug fix anter FFE 13:19:29 s/anter/after 13:19:56 heyongli: if a VM doesn't have pci devices, and resized with PCI devices, it's no longer accessible after resizing. If a VM does have pci devices, after resizing with more PCI devices, the resulting xml would have wrong number of pci devices. 13:20:08 irenab, we need to put this in the kilo list 13:20:25 i check the resize on common pci passthrough and dig out what's wrong 13:20:28 i will 13:20:41 heyongli: that sounds great! 13:21:00 baoli: let's verify if it works and if not, will add it to Kilo list 13:21:13 irenab, sounds good 13:21:52 pci lack of gate testing , maybe a CI, so it might broken while more patches get merged 13:22:40 anything else? Otherwise, I will go back to work 13:22:56 let's hope that we can get everything in by the end of Friday 13:23:32 baoli: if there is anything I can help, just ping. Even if means to test something, feel free 13:23:35 to ask 13:23:47 irenab, sure. 13:23:57 ok, so good luck! 13:24:01 see you next week 13:24:12 see you 13:24:14 see you. 13:24:17 #endmeeting