=== borschty_ is now known as borschty [03:28] hello all, I know I must be sounding like a broken record. I am growing more frustrated with trying to get suspend working [03:28] is there any way I can help debug an issue with the drm.ko module? Removing it fixes my suspend (at the cost of hardware accel) [04:03] can I get some help trying to compile a new version of the i915 driver with some patches I found? === thegodfather is now known as fabbione [12:25] amitk: hows aufs looking? [12:26] good, I am compiling the lpia kernel and upload that, then I apply your patches to ubuntu kernel, test with some live cds [12:27] laga: there won't be a kernel upload before beta it seems though. sorry. [12:28] unless we get a fix to the e1000e bug for which rtg has sent a patch [12:32] i think i can live with that :) [12:34] laga: atleast the mobie team will be using your aufs for their live-usb-sticks. They have a separate kernel where I can push these patches today. [12:34] yeah, i need it in generic for mythbuntu-diskless. [12:34] *your aufs patches, that is [12:51] amitk: what do you mean by "Signed-off:", I should use gpg-sign?-) [12:54] tjaalton, I guess he just meant that the patch should include a line "Signed-off-by: name " [12:54] smb_tp: oh, ok [12:54] tjaalton, It also would be interesting: where did that patch come from? It wasn't in upstream yesterday [12:55] smb_tp: from the linuxwacom tarball [12:55] linuxwacom.sourceforge.net [12:56] src/linux-2.6.26, diff the files against our kernel [12:57] unless I've f****d things [12:57] +up :) [12:58] tjaalton, :) If it works for you. Just have to see whether this also goes upstream or has to be promoted. [12:58] tjaalton: exactly what smb_tp said. And it is good to know the provenance of the patch (where it came from) [12:59] amitk: ok, should've included the url then. I'll update the patch [13:00] my tablet doesn't seem to need the module at all, so I haven't tested it myself [13:01] tjaalton: in that case, do you know if it _does_ fix it for somebody? No point in applying the patch otherwise. [13:01] amitk: of course, as I said "bug foo confirms that .." [13:02] there are a couple of users who got it working by compiling this version of the module === thegodfather is now known as fabbione [13:04] tjaalton: ok. read through the bug now. Please send Signed-off line with a URL [13:06] amitk: like this? http://users.tkk.fi/~tjaalton/dpkg/wacom-1.49.patch [13:06] uh [13:07] ok there [13:08] tjaalton: looks ok [13:13] amitk: great [13:59] hi. one questions. whats the idea for those 3g modems that also are storage devices. usually folks now write their own udev rules to get that working. is there anything we can do in general to prefer the modem driver vs. the storage thing? [14:02] asac: Arne showed me one that had two different USB ids. On insertion, it first acted as a mass storage device, then you had run some program to switch to the other device. If other devices are similar it seems like something that should be handled in userspace. [14:03] amitk: well. i talked to dan williams at some point and he said that the storage thing should be blacklisted or something [14:04] amitk: not really sure what he means by that. will try to ask him again [14:05] asac: i guess he means blacklisting a specific usb id from the storage driver. That would make the storage never work. Better ask him. [14:06] amitk: right. but the storage thing is really just a tiny storage used to ship the windows drivers [14:06] amitk: i had something similar for a wifi stick [14:06] fwlanusb [14:06] before i fixed the driver it was always storage [14:06] with the windows driver on it [14:07] but once fwlanusb driver was proper, things just worked (e.g. no storage device popping up) [14:07] amitk: i think dan williams didnt say "blacklist", but "give the modem driver higher prio" ... does that make sense to you? [14:08] asac: not sure what giving higher prio means. [14:09] amitk: maybe making udev first modprobe the modem driver before trying storage? [14:10] anyway. i will try to get a word from the master on this ;) [14:23] asac, there are 3g sticks that really need an userspace app to tell them "no, dont do storage, do 3g" [14:24] i think option makes those [14:24] Zhenech: ok. but there area also others that would do the right thing when the modem driver would be tried first? [14:26] asac, to be honest, I dont know, I only had "normal" modems and these "i need a app" sticks in my hands till now [14:26] asac: the ones Zhenech mentions are the only ones I have seen. If there are saner ones, then by all means we could blacklist them in the storage driver. [14:27] ok. so in the end NM probably has to do that :( [14:27] you can find a lot of information about these on pharscape.org [14:28] asac, perfectly udev would do "oh, its a storage, but the id I know to be a 3g, lets call to enable modem" [14:28] so it happens on plugin of the modem directly [14:28] Zhenech: yeah. but then how do we find all the ids of those devices? [14:28] as everyone uses udev, but not everyone uses NM [14:29] Zhenech: yeah. but people not using NM are usually more skilled ;) and even like to find out about udev rules :-P [14:29] but agreed [14:29] at best this would happen at the lowest level possible [14:30] see this post http://www.pharscape.org/component/option,com_forum/Itemid,68/page,viewtopic/t,545/ [14:30] it contains the utily I mentioned for option cards [14:30] and udev rules with all the shiny ids needed :) [14:32] here are probably more http://www.pharscape.org/component/option,com_forum/Itemid,68/page,viewtopic/t,425/ [14:32] asac, sponsor me such a card and I'll package it :> [14:33] Zhenech: show me where the source is ;) [14:33] i cant find it :) [14:33] (for ozercdoff) [14:34] the udev.tar.gz attached to the post has everything [14:34] the name is ... bad ;) [14:34] Zhenech: isnt that released fro ma more official place? [14:35] not that I would know off [14:35] afaik linux support for the option hardware is only semi-official [14:36] Zhenech: ok. but its hso-udev.tar.gz ... or is there a plain udev.tar.gz that i dont see? [14:37] http://www.pharscape.org/component/option,com_forum/Itemid,68/page,viewtopic/t,545/ lists only udev.tar.gz, 26kb big [14:43] (the other link has a hso-udev file) [14:50] ping abogani [15:01] DBO: pong [15:01] abogani, can you help me understand why your kernel is the only kernel with which my machine will suspend? [15:01] i am on intrepid, using the linux-rt kernel you created [15:02] it suspends wonderfully, every other kernel hard locks on resume when the drm module is loaded (they work fine if I dont load drm) [15:03] It is very hard to say. Generally rt incur in suspend/hibernate problems... [15:03] i tried recompiling the latest -generic kernel for intrepid using your config, and I even tried using your config against the stable 2.6.26 kernel [15:04] i can't seem to grab the source for your kernel however, apt-get sourceing it just gets me a debian/ [15:04] abogani, did you do anything special to the drm module? [15:04] rt config is very close to -generic one except for rt specific configurations. [15:04] No at all. [15:05] any clue why my console is screwed up in your -rt kernel... it blinks red and black off and on [15:05] What video driver are you using? [15:05] i915 [15:05] I am just trying to piece it all together, maybe with enough hints I can figure out how to get a reliable machine =/ [15:06] Sorry i don't have experience to this type of hardware. [15:07] abogani, I also experience some freezing with your kernel [15:08] something about bad page state [15:08] I had emailed you a backtrace to your ubuntu email right before contacting you here [15:18] DBO: I suspect that upstreams already have fix that bug. Anyway keep in account that -rt will be dropped in Intrepid. [15:19] intrepid is not going to have -rt at all? [15:19] Not provided by me at least. [15:20] Sorry [15:20] =/ sucky... its the only real kernel that works for me. [15:20] I can't be the only one out there with an intel x4500 integrated gpu [15:20] well I know I am not, there is at least one other guy with the same problem [16:33] smb_tp: any news on the ports build? [16:34] munckfish, I've prepared something that should build but I still have to get someone to upload it... [16:34] smb_tp: ok [16:35] smb_tp: seeing as it's not uploaded now - if I wanted to get another patch in now before the beta would that be a PITA or easy? [16:37] munckfish, Well closer to PITA since I pushed the things needed for the upload to the repo already [16:37] smb_tp: np I'll wait till after [16:58] Hello. I am looking for the netback.c file in the netback driver for the latest Hardy stable Kernel. I downloaded the source from the repo and cloned the hardy git repo and the file is missing. Where should I look for such file? [17:01] What is the kernel module name? [17:02] Isn't that in Xen? [17:13] IntuitiveNipple: Xen would be the Git Repo? [17:16] The only Ubuntu xen repo's I know of are Chuck Short's, for intrepid [17:17] Where do the sources for llinux Xen kernel would come from? The said file is not under Xen [17:18] Only Xenbus stuff is [17:18] elventear: I'm not sure [17:19] There is an Ugly bug in the Xen Kernel, and I am trying to debug it. I have been told that there is where the bug might reside [17:19] But I want the source that Ubuntu uses [17:20] elventear: For building it, last time I looked, it required the "custom-binary-xen" target in the Ubuntu kernel build process. [17:20] elventear: And "debian/rules custom-prepare-xen" should show the xen patches [17:21] Ok [17:21] Thanks [17:21] elventear: That info came from the kernel mailing-list in April, so it ought to be valid [17:21] elventear: Although, that was Hardy of course === thegodfather is now known as fabbione [18:04] BenC, , persia and i were looking at bluetooth support in intrepid and running into lots of issues with btusb. is there any particular reason we are tied to switching to it when the hci-usb stuff is still available? [18:05] (or anyone else whom is relatively informed on this) === calc_ is now known as calc [19:57] BenC: ping? [19:58] rtg: ping? [19:58] fabbione: yo [19:59] rtg: hey.. did any of you pulled from my gfs1/2 tree into intrepid (both normal and ports)? [19:59] i can't see anything in the last kernel releases [19:59] fabbione: I remember doing it once, but nothing recently. [20:00] BenC has been leading the charge on Intrepid. I've been busy on other things. [20:00] i did send another pull request after 2.6.27 was first uploaded to intrepid [20:00] lemme dredge through the list and find it. [20:01] no BenC didn't pull [20:02] you are shipping a broken gfs1 and gfs2 [20:02] s/you/BenC [20:02] eek [20:02] i said it correctly before :) [20:02] BenC didn't pull [20:03] Did you send your pull request in September? I don't see it, but my eyes might be getting tired. [20:03] probably august [20:04] https://lists.ubuntu.com/archives/kernel-team/2008-August/002854.html [20:04] there is one from 27th of Aug [20:04] in my sent folder [20:05] https://lists.ubuntu.com/archives/kernel-team/2008-August/003004.html [20:05] this one [20:05] so 002854.html still has the request for -ports that has not been done at all [20:05] the latter has extra for supported arches [20:05] that has not been done either [20:06] fabbione: I'll sic smb on the https://lists.ubuntu.com/archives/kernel-team/2008-August/002854.html request, and handle the Intrepid pull myself. [20:06] rtg, can you check if the current patchset for btusb suspend support is added into the current 2.6.27 tree we're shipping with, i ididnt think it was? http://www.kernel.org/pub/linux/kernel/people/holtmann/patch-2.6.26-mh5.gz [20:07] rtg: "sic smb"? [20:07] rtg, you do what? [20:07] superm1: uno momento [20:07] smb_tp_: maybe you should respin the -ports upload after pull fabbione's gfs2/gfs1 patches. [20:08] marcel posts these to http://www.bluez.org/, and this was just a few days old - so i'd think not generally [20:08] ah ok [20:08] got it [20:08] rtg: anyway it's _very_ important you get the changes for supported arches [20:08] rtg: what you have there for gfs1 is totally broken [20:08] gfs2 is exporting unrequired symbols and other stuff [20:08] fabbione: ok, I'm getting the message :) [20:09] ok.. [20:09] thanks [20:09] rtg, hm. ok. munckfish isn't around but then I can take his one more thing as well [20:19] fabbione, ports pulled [20:20] smb_tp_: thanks [20:29] fabbione: I pulled after rebasing your tree. Its been long enough that I'm doing a test build before pushing it back out. [20:45] superm1: looks like that pathc is already in Linus' tree, so we'll pick it up on the next rebase. [20:46] s/pathc/patch/ [20:46] rtg, any estimate on when that should be happening? [20:47] should be soon. -rc7 is close to week old now, isn't it? [20:48] fabbione: I pushed your bits [20:48] okay great, thanks rtg === calc is now known as calc_ === calc_ is now known as calc