[06:24] woops, nexus7 panic on nfs4 still exists [06:25] mlankhorst: so there's a way to fix the race once and for all [06:25] \o/ [06:25] Bring out the low orbital satellite death ray! [06:26] the short-term fix is to add a job that emits an event on startup or started plymouth-splash, and use that in *dm [06:26] *dm.conf [06:26] that should cover the case where plymouth is in initramfs, for one reason or another [06:26] can it be trivially backported to precise/quantal? [06:27] I hacked it locally to work nicely with 'and (started cryptdisks-enable or started plymouth-splash)' but it works only in this one case :) [06:27] should be [06:27] this is where the bikeshedding starts :) [06:27] like, what to call the job [06:27] bug99something [06:27] 982889 :) [06:28] .conf [06:28] just call it something ugly and then let the sru admin decide the final name [06:30] yeah, not the time for that now [06:33] it's what I end up doing, anyway [06:48] reviewer always gets to decide the name ;P [07:04] ugh building on nexus7 is taking ages [07:04] I have a good feeling about the fix though, your backtrace from bug 56578 seemed like a dupe of the corruption I was hitting [07:04] bug 56578 in mlview (Ubuntu) "Please sync mlview (universe) from unstable" [Undecided,Fix released] https://launchpad.net/bugs/56578 [07:05] * mlankhorst slaps ubottu around a bit iwth a large trout [07:05] I tried the patch and it didn't fix it for me [07:06] yeah that's why I'm compiling myself, see what's still wrong [07:06] it's the same one whot is seeing, and what I had with 1.14 too [07:06] ah [07:06] good [07:06] hmm should I push my tree somewhere? [07:07] 1.14 is pain, the backport is easier to handle [07:07] 1.14 was working for me though with the patch I wrote [07:07] with synaptics [07:08] for nexus7 you'll need to do tricks to be able to run unity [07:08] or run the backport [07:09] I'm just running nexus7 + 1.13 [07:09] default image [07:09] I'm trying with all of the patches linked from that bug [07:09] yeah [07:10] that's what I meant to push somewhere [07:10] it's building that's a pain [07:10] build it parallel=3 [07:10] nfs4 mount causes the nexus to reboot, so I had to force nfsvers=3 [07:10] and wireless is not that good here :( [07:11] copy the source locally [07:11] yeah next time :P [07:11] faster that way [07:11] I'll try connecting it through usb [07:11] actually should be doable on my panda, it already bridges the network to a wireless access point anyway [07:12] I don't run the access point atm since it seems to cause memory corruption [07:34] added 'status plymouth-splash | grep -q "start/running" && exit 0' before exec lightdm in the job, without adding any preconditions [07:34] seems to work, but there's a strange timeout [07:34] and this would only fix lightdm.. [08:04] yeah getting same crash on nexus [08:05] http://paste.ubuntu.com/5697733/ [08:12] we're missing c0a752d2864872023216005375a6a1973fadeffe but probably harmless [08:20] fix for #57301 too, dno about that, probably needs the commit before that as well [08:22] doesn't hurt to try [08:30] I feel like just doing a raw diff on exevents.c and grabs.c, see what differs.. [08:32] hm same, except pointer barriers [10:09] ogra_: I have no luck so far with digging through the pointer barrier stuff :/ [10:09] :/ [10:09] sad [10:19] I think that x1.14 is crashing in the same way tbh [10:29] hard to tell since you dont have awroking driver for it [10:29] *a working [10:31] I built 1.14 without the video abi changing bits [10:34] the only thing I can see that's different is pointer barriers though, which shouldn't get in the way anyway [10:44] tjaalton: yeah but compiz isn't the same either [10:46] what does that change? [10:48] they're handled in different places with the old patch from what I can tell [10:51] you mean unity? [10:52] hm I'll try the same 1.13 on macbook, see if it crashes [10:52] i used the staging version which was ported to 1.14 [11:00] as I expected, 1.13.3 works on the macbook [11:03] hm the only difference seems to be that I don't get the window drag thingy with synaptics [11:03] oh right, 3 buttons is middle click [11:05] hmm, so we just need to ship macbooks with the nexus7 images ? [11:05] nah [11:05] macbook uses synaptics [11:08] tjaalton: I'm going to assume for now that the reason I can't trigger it on my macbook is that it's a different bug, I don't get the 3 fingers move window thing that I did get on nexus7 [11:25] ogra_: can I enable the onscreen keyboard from the nexus on desktop? [11:28] ah got it [11:28] BOOM [11:29] now lets see if the same happens on 1.14 [11:35] oops looks like it was the original bug, forgot to reapply the patches to 1.13 :/ [11:58] mlankhorst: it's not the same driver either [12:09] mlankhorst: i told you ;) [12:10] tjaalton: true but I was hoping I could find a way to reproduce it on the macbook [12:10] sure, would've been easier [12:10] any touchscreen device should do === PsychoCandy is now known as RadiumCat [12:37] oh btw also fun, I noticed that in plymouth card0 is used blindly, instead of looking for boot vga [12:42] i pinged whot the other day btw, said he's been busy with other bugs lately and hasn't had time to look into this again [12:43] :( [12:43] kinda promised to look soon I think [12:43] lets see if I can crank up the debugging and figure it out myself then [12:46] or at least make some xorg-integration-tests thing out of it [13:13] hi everyone [13:15] I'm building a package for an upstream project that can only be compiled to 32bits. This project depends on the presence of libGL.so.1 and libX11.so.6 at execution time. What are the best packages to depend on? [13:15] is it libx11-6 and libgl1? [13:16] libgl1-mesa-glx [13:17] the best package to depend on is ${shlibs:Depends}, and let dpkg-shlibs figure it out from there. [13:17] yeah, true [13:17] *dpkg-shlibdeps [13:17] tjaalton: thank you. It looks like there are other providers of this binary like nvidia*. Will that be ok? [13:17] should be [13:18] jcristau: thanks. I already have this one but these packages didn't get installed [13:18] you need to fix the packaging then [13:18] tjaalton: do you some minutes to mentor me? [13:19] DamienCassou: sorry, no. try #ubuntu-motu [13:19] ok, thank you [13:46] why does xinput2 have to be some complicated [13:53] * mlankhorst guesses it's related to TouchRemovePointerGrab being a stub [14:12] I think I'll give up, there's too much obfuscation between dix and Xi to make sense of anything.. the whole emulation, touch listeners and active vs passive grabs is not helping much either [14:15] cnd: ^ looks like you managed to stay interested in xi2 longer than me! [14:27] is nvidia-319 coming to ubuntu linux 13.04? [14:29] public repo? [14:29] Multi/Universe? [14:30] tseliot knows when [14:30] tjaalton: but you think it comes? [14:31] what do you need from 319 that you care so much? [14:33] mlankhorst: do you even know what it offers...?? [14:33] mlankhorst: you are able to render stuff on nvidia graphics card, and display it via intel drivers [14:34] .. on a precise xserver? [14:34] you know more than me then [14:39] or precise kernel for that matter [14:39] mlankhorst: yeah, xinput2.2 is a hairy hairy beast... [14:40] 2.3 now, pointer barriers [14:40] but I don't believe it's the cause of the bug, so xi2.2 is fine with me too [14:42] heh [14:46] gotwig was wrong [14:46] can't use -intel [14:46] well perhaps if you hack xserver you could, by pretending nvidia is the primary device [14:47] it supports prime but definitely won't work in Precise [14:48] you need the drm-trololol-gpl patches first [14:48] i doubt they were in 3.5 kernel, so unless the backported hsw has them it won't work even with the 3.5 kernel\ [14:49] didn't they work around the gpl symbols? [14:49] yeah by using the drm helper stuff they wrote [14:49] it's all in our 3.8 kernel [14:49] yep [14:49] but 3.8 is not in precise yet, I think? [14:49] ohey, it even might be [14:51] oh well I'll have to file a bug to get the dependencies for xserver-xorg-lts-raring ready then [14:53] right [15:05] will we still get 9.1? [15:11] probably [15:12] http://packages.qa.ubuntu.com/qatracker/milestones/266/builds/41730/testcases/1515/results [15:12] mostly intel testers though [15:13] anyone with radeon & nouveau please test ^ :) [15:15] worksforme? what exactly do you want from the tests? [15:18] hey [15:19] gotwig: realized you can't use -intel driver with nvidia? :) [15:33] he asked about raring, not precise [15:39] bjsnider: yeah still wont work with optimus in the way he described it.. [15:40] still, one must admire the strength of his convictions [15:43] meh it was probably on phoronix like that [15:56] any chance to see nvidia-319 on ubuntu 13.04 repo? [15:59] at some point it will be there [16:00] bjsnider: but for that we need more actual packages of Xorg and xrandr I guess? [16:00] i don't see why [16:01] e.g xrandr 1.4 [16:01] and normaly you need kernel 3.9, for the prime helpers [16:02] but kernel 3.8 is going to be the default kernel for ubuntu, right? [16:03] well, maybe, but i don't think there's any rush because some people have tried it the past couple of days and reported it doesn't work [16:05] bjsnider: for me it also doesnt work normaly [16:05] bjsnider: I use optimus [16:05] but when you look at the changelog.. well, there are not many improvements [16:05] that's what they've been testing [16:06] and if you install nvidia-319, bumblebee-nvidia gets removed. Why? [16:07] because the two are mutually exclusive [16:07] intel_do_flush_locked failed: Input/output error [16:07] hate you too intel [16:08] * seb128 reboots [16:09] bjsnider: they are what? [17:13] gotwig: trying to install nvidia blob ? [17:13] :) [17:14] he believes it will give him optimus :/ [17:14] héhé :P [17:14] dupondje_: what? [17:14] Let me believe :( [17:14] * gotwig cries [17:15] it isn't even capable to give me a gnome-shell here :( [17:15] dupondje_: I ask if nvidia-319 is going to come to ubuntu repos [17:15] you can try the xorg ppa :) [17:16] mlankhorst: which I could help getting nouveau optimus ready :) [17:16] dupondje_: When I try to use such PPA's, my multitouch does no longer work [17:16] when nvidia-319 is in ubuntu repos, I am sure that I am not going to break my system, and all dependencies are in place [17:17] dupondje_: nouveau is a good thing. The optimus support is a nice idea [17:17] I hope for you :) Tried to install it yesterday, and gnome-shell didn't even start :( [17:17] dupondje_: do you think, nouveau is going to be able to use prime helpers? [17:17] ._. [17:17] so *sad* [17:17] dupondje_: its not gnome shells fault [17:18] I know .. its the **** nvidia driver's fault ;) [17:18] hey there sry.. [17:18] like i said, its not GS fault [17:41] dupondje_, you're a fellow gnome-shell user? [18:30] bjsnider: yep :) === seb128_ is now known as seb128 [20:25] hey [20:25] I have an asus zenbookux32vd [20:25] I have optimus, nvidia and intel hd graphics [20:25] When i try to follow this guide to reset the DPI with xrandr https://help.ubuntu.com/community/AsusZenbook it doesnt work [20:26] However, if I e.g apply VGA1 when I use an external display it works [20:26] how can I find out the name of my notebook screen for xrandr? [20:27] xrandr -v [20:27] xrandr program version 1.3.5 Server reports RandR version 1.4 [20:28] it's probably called LVDS1 [20:28] sorry, xrandr -q [20:28] eDP1 [20:29] ah, one of those. [20:29] ok [20:29] I can now set the gama [20:29] but the DPI are still untouched [20:29] why [20:29] I set higher and lower [20:29] nothing changes [20:30] same issue [20:30] from ubuntu 12.04 - 13.04 [20:31] bryce: what can I do [20:35] gotwig, http://lmgtfy.com/?q=ubuntu+raring+how+to+change+the+font+dpi+settings [20:36] bryce: text scaling is not the same [20:37] bryce: I already use text scaling, and btw. these AskUbuntu answers are more than bad. They are noob friendly.. this is a good article: http://robswain.tumblr.com/post/14209540835/xorg-gnome-3-on-high-dpi-screens [20:45] bryce: maybe I am just stupid [20:45] I think it works [20:45] but I just cant notice it [20:57] mlankhorst: btw, I had uploaded 2.21.6-0u1 already [20:57] so you'd need to update the changelog if you merge from d-e [21:27] mlankhorst: too late :) [21:35] mlankhorst, I went ahead and included the change in my upload. If it should not go in let me know asap so I can reverse it. [21:36] and yeah +1 to changelog entries :-) [21:45] huh, we don't have xrandr 1.4 in x11-xserver-utils yet [21:45] didn't know that [21:50] ..but provider stuff is patched in [21:51] oh I did it, getting old it seems [22:07] heh [22:07] I've been running into that myself.