=== Kaloz_ is now known as Kaloz === ikonia_ is now known as ikonia [09:03] cking: thanks for the fixes, do you have a branch where I can pull from? I am not close to a patch-friendly email setup currently [09:05] amitk, i don't but I can do that in a mo [09:07] amitk, git://kernel.ubuntu.com/cking/idlestat [09:08] cking: great, thanks [09:08] btw, there are a bunch of big leaks that valgrind is finding, I've not yet had time to look at them all [09:09] amitk, do you intend to throw the code at tools like coverityscan? if you want I can set that up? [09:09] cking: we really should run valgrind/sparse/ etc. on this but we're racing against time to get it feature complete for kernel summit in Aug [09:10] cking: please go ahead, will you be able to share the reports? [09:10] i can share the reports and add whoever you want to be included on the project [09:11] btw, do you require this to be packed in debian? 'cos I am willing to do that to as I'd like this tool in ubuntu ;-) [09:12] cking: yes please! As you know, packaging has been a particular weakness of mine [09:12] i'll try and get it done when I have some "spare" moments [09:12] much appreciated [09:12] np [09:13] amitk, BTW, nice write-up in LWN [09:27] cking: thanks :) [11:29] hans109h, i have started the conversation upstream [14:54] mlankhorst, so ... do we still use RADEON_UMS under any circumstances or has that been excised from Xorg now [14:56] not in X.org at least since precise, maybe it was used on other archs but the xorg driver from quantal didn't support ums any more [14:57] and even in precise it was probably some old ppc's only that actually ended up using it [15:04] BenC, I've disabled powerpc64-emb in Utopic for FTBS in some arch code. How about having a look ? ('cause I think it is as a result of your SAUCE patches) [15:05] rtg: Ok, I’ll have it working by Monday [15:06] wfm === brendand_ is now known as brendand [16:18] rtg: on bug #1332137 I don't think he's saying that -8 shouldn't have been reverted for 7260, I think he wants it reverted for 3160 as well [16:18] bug 1332137 in linux-firmware (Ubuntu) "linux-firmware 1.127.3 has -8 firmware for Intel 3160, but not Intel 7260" [Undecided,Invalid] https://launchpad.net/bugs/1332137 [16:19] rtg: that's based off of the activity on bug #1293569 [16:19] bug 1293569 in linux-firmware (Ubuntu) "iwlwifi-7260-8.ucode causes network disconnections problems" [Medium,Confirmed] https://launchpad.net/bugs/1293569 [16:20] sforshee, oh. Maybe Trusty will be out soon. [16:20] rtg: yeah not sure that it makes sense to revert it at this point [16:59] bjf: hey for autotests; if a test might crash the kernel; how will autotest detect the failure? is it timeout based? [17:02] arges, yeah, it's likely to timeout though it also may just hang the jenkins job until someone goes and pokes it [17:02] rtg: thanks for applying tegra drm support [17:02] would it also be possible to increase the CMA size to 64M? [17:03] arges, it's fairly straightforward to add timeouts to the tests if desired (i've done it) [17:03] marvin24, this one ? CONFIG_CMA_SIZE_MBYTES=16 [17:04] yes [17:04] rtg: needs 64M [17:04] marok [17:04] marvin24, ok [17:04] rtg: thanks! [17:06] marvin24, done [17:07] rtg: ok, so thanks again for high speed :-) [17:07] np === kgunn is now known as Guest13280 [18:35] bjf: is there a way to have autotest install ubuntu package dependencies [18:37] arges, i do that via my own infrastructure. look at kernel-testing git repo; lib/testsprops.py [18:38] bjf: so that's done outside of autotool client test? [18:39] arges, yes [18:40] arges, though there's no reason the test itself can't do it [18:40] bjf: utils.system('apt-get install blah') ? [18:40] arges, yeah [18:42] arges, i install pkgs externally using my method and then reboot the system one final time before running the actual tests [18:51] bjf: http://autotest.readthedocs.org/en/latest/main/local/AddingTest.html#setup fyi [18:54] arges, yes i use that [18:54] arges, but not for installing packages [18:54] for the tarball extraction etc.the last comment shows how to use the 'client's system software manager' which uses apt [18:55] arges, that may well be but i want a reboot after installing packages and before running tests [18:55] bjf: ok