| RAOF | apw: Can I get a linux-backport-modules-nouveau metapackage in your lbm ppa? The DDX requires the kernel module, so I need a Depends line on something that'll ensure a kernel module. | 00:22 |
|---|---|---|
| RAOF | apw: linux-backport-modules-nouveau is slightly broken; It creates a /sys/bus/pci/devices/0000:00:02.0/lbm-drm directory rather than /sys/bus/pci/devices/0000:00:02.0/drm, which breaks libdrm's KMS detection. | 04:02 |
| === bjf is now known as bjf-afk | ||
| Ng | the apport/whatever thing that asks questions about kernel oopses really needs "I don't know" answers to the questions | 09:24 |
| Ng | I don't know if the oops I'm currently reporting happened before, or what caused it. I didn't even know my kernel had oopsed until the little crash icon appeared in the panel ;) | 09:25 |
| smb | Ng, Heh, yeah maybe. But in a nutshell I'd use "no" then. Otherwise you would know. :-P | 09:28 |
| Ng | smb: indeed, I had to say no, but how do you distinguish between "I don't know, so I said no" and "I definitely know the answer is no"? :) | 09:30 |
| apw | heh ... not an unreasonable position | 09:33 |
| smb | Ng, I fear only from users adding comments to the created bug. apw Yeah, we might need to think on that | 09:33 |
| RAOF | apw: Good... morning? :) I've tried your linux-backports-modules-nouveau PPA; it works nicely, modulo the issues that are in backscroll. | 09:35 |
| apw | scrollback has no other messages from you here | 09:36 |
| RAOF | It doesn't? Ok. | 09:36 |
| RAOF | So, the first one is just a general packaging wishlist: the DDX package will need to depend on something to ensure there's a nouveau kernel module. A linux-backports-modules-nouveau{,-generic,etc} metapackage would be good. | 09:37 |
| RAOF | The second one is that I think renaming drm -> lbm_drm breaks libdrm's KMS detection; libdrm walks /sys/bus/pci/devices/$DRMDEVICE/, looking for either a drm:controlD* file, or a drm/ directory containing a controlD* file. | 09:39 |
| RAOF | The backports module creates a lbm_drm directory instead, so libdrm says that KMS is unavailable - and since the nouveau X driver no longer has userspace modesetting code, it bails when it detects that KMS is unavailable. | 09:40 |
| RAOF | Patching out the DDX's check for KMS makes everything work swimmingly, though. Compiz is back on my nouveau laptop :) | 09:41 |
| apw | RAOF, ahh ... hrm ... ok the meta package is in the pipe once LBM goes into the main archive a meta package will come with it, i'll try and remember to get that uploaded to the PPA | 10:34 |
| apw | the lbm_ prefix is more problematic | 10:35 |
| apw | as i thkn the /sys files are based on the module name, but i can investigate | 10:36 |
| apw | i block renamed them all to avoid the checks in the twin drm drivers for 'mkdir must work' in sysfs | 10:36 |
| abogani | Hi All, Someone could recall me how "updateconfigs" works? | 13:23 |
| abogani | If it works obviously... Executing "fakeroot debia/rules updateconfigs" return a "dh_testdir: cannot read debian/control: No such file or directory". | 13:25 |
| smb | abogani, Do a frd clean before that | 13:25 |
| abogani | smb: frd? | 13:26 |
| smb | abogani, My bad shortcut for fakeroot debian/rules | 13:26 |
| abogani | smb: Yeah you are right: it works now. | 13:27 |
| abogani | smb: Could you recall me how use updateconfigs? I just added a new flavour... | 13:28 |
| smb | abogani, updateconfigs just goes through the flavours and sorts options that are common to common levels global, arch, flavour | 13:29 |
| smb | debian.master/config.common.ubuntu defines options used by everything, i386/config.common.i386 those common for all i386 flavours and then i386/config.flavour.generic those specific to that flavour | 13:31 |
| abogani | smb: Ok but where I should place the my-new-flavours-kernel-config-file (in way that will be considered by updateconfigs)? | 13:33 |
| smb | abogani, note (not sure how much you followed the changes since hardy) there are no generated files in git anymore. You create those with fakeroot debian/rules clean and hopefully should be able to remove with a distclean | 13:34 |
| * abogani really remain to Hardy times :-) | 13:34 | |
| smb | place it into debian.master/<arch>/config.flavour.<yourflavour>, then calling updateconfig should remove all stuff that is common to generic configs. Correct apw? | 13:35 |
| apw | correct | 13:35 |
| abogani | smb, apw: Thanks! | 13:35 |
| LLStarks | ogasawara, you know alot about blanking, right? | 16:22 |
| ogasawara | LLStarks: Uhh, I'm not sure what you mean by that. Can you be more specific. | 16:23 |
| ogasawara | LLStarks: better yet, point me to a bug # :) | 16:23 |
| LLStarks | https://bugs.launchpad.net/ubuntu/+source/gnome-power-manager/+bug/496842 | 16:24 |
| ubot3 | Malone bug 496842 in gnome-power-manager "Can't unblank my Inspiron 640m after a lid close" [Undecided,Confirmed] | 16:24 |
| === bjf-afk is now known as _bjf | ||
| LLStarks | ogasawara, that apport command does nothing after i answer the questions | 17:05 |
| LLStarks | nvm | 17:05 |
| LLStarks | it took a while | 17:05 |
| ogasawara | JFo: did you say you put bug 502219 on the list? | 18:07 |
| ubot3 | Malone bug 502219 in linux "Kernel errors triggered with SSD storage" [Medium,Triaged] https://launchpad.net/bugs/502219 | 18:07 |
| JFo | let me check | 18:07 |
| JFo | no, I don't have it | 18:08 |
| JFo | will investigate and add | 18:08 |
| ogasawara | JFo: thanks. It was brought up in the ubuntu dev week talk, so figured might be good to look at | 18:09 |
| JFo | certainly | 18:09 |
| JFo | btw, I have copied off your talk today for review later. | 18:11 |
| JFo | Maybe it will save you from some questions from me :) | 18:11 |
| ogasawara | JFo: cool, they usually archive it to some link in a few hours | 18:12 |
| JFo | yeah, I copied it and chopped out all the join/part messages | 18:12 |
| JFo | so it is easier for me to read | 18:13 |
| === mdomsch is now known as mdomschbb | ||
| crimsun | _bjf: what do you think of enabling verbose debugging for the c-o-d alsa-driver builds? | 18:36 |
| crimsun | _bjf: pro: ability to trace HDA codec init. con: ring buffer spam | 18:37 |
| crimsun | _bjf: for the pro, it's easier to detect when an incorrect quirk is being used or if an existing quirk needs to be modified. (I already fixed an issue with my new laptop this way.) | 18:38 |
| === mcasadevall is now known as NCommander | ||
| bdmurray | ogasawara_: I think bugs 414444, 413139 and 351119 might be dups ;-) | 21:05 |
| ubot3 | Malone bug 414444 in linux "[Sony Corporation PCG-FXA63(UC)] hibernate/resume failure" [Undecided,Incomplete] https://launchpad.net/bugs/414444 | 21:05 |
| ubot3 | Malone bug 413139 in linux "[Sony Corporation PCG-FXA63(UC)] hibernate/resume failure" [Undecided,Incomplete] https://launchpad.net/bugs/413139 | 21:05 |
| ubot3 | Malone bug 351119 in linux "[Sony Corporation PCG-FXA36(UC)] hibernate/resume failure" [Undecided,New] https://launchpad.net/bugs/351119 | 21:05 |
| moore | quick question: what's the difference between the kernels? specifically, who should be using the drm-intel-next kernels? | 21:19 |
| RAOF | apw: If it's hard to change the lbm_ prefix it would be trivial to patch libdrm to also match in lbm_drm/controlD*. | 21:41 |
| bryceh | heya RAOF | 21:48 |
| RAOF | bryceh: Howdie. | 21:54 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!