[04:56] apw, getting ready to look at the kernel now - I had a talk tonight 6 hour drive round trip and the meeting lasted 2.5 hours :-( [04:57] ouch [05:59] hello, is there a way to build kernels and only have make build modules that have changed/been patched? [06:03] I am currently using these instructions [06:03] https://wiki.ubuntu.com/KernelTeam/GitKernelBuild [06:03] is there a better way? === amitk-afk is now known as amitk [09:27] apw, ping [09:29] akgraner, hi [09:30] so I installed the Kernel and haven't had any heat issues [09:31] is there anything you need me to run to stress the CPU more - I have noticed though I don't hear the fans as much but that could just be me [09:33] if your tem stays arround 45 then all is good [09:33] though that patch still breaks mac books, so we still don't have it in the lucid kernel ... grrr [09:37] apw, alrighty - I'll keep my eyes open and watch that - but I haven't gotten above 48 in the last few hours [09:38] akgraner, perfect thanks ... i'll let you know when i have something more to test, likely after freeze now [09:38] and we'll have to ram the fix through the freeze if we are allowed [09:40] apw cool - just let me know what if anything you will need from me :-) I help where I can! Thanks again!! [10:35] JFo: I have confirmed bug 535315, which seems to have a stack of duplicates [10:35] Malone bug 535315 in linux "WARNING: at /build/buildd/linux-2.6.32/net/sched/sch_generic.c:261 dev_watchdog+0x262/0x270()" [High,Confirmed] https://launchpad.net/bugs/535315 === jussi01 is now known as jussi [13:38] mdz, thanks :) [13:39] I have seen several similar ones with a lot of duplicates too [13:39] one in particular from the last 48 hours is an ath9k issue [13:41] it is now on my list for review [14:29] JFo: thanks [14:29] my pleasure mdz === ricotz is now known as ricotz_ === ricotz_ is now known as ricotz [15:00] so.. um.. where are the headers-*_all -packages from latest mainline builds? [15:06] refic, I don't know what you mean. I have just looked and they are all there [15:06] in fact, I am looking now [15:08] I must be blind then [15:08] ah wait [15:08] I am misreading your statement [15:08] you are looking for the _all.deb I was thinking header...deb [15:08] sorry [15:08] yeah :) [15:09] :) [15:09] np, good to know I'm not blind after all [15:09] no, they are gone [15:10] apw, any reason we no longer have _all.deb in the Kernel PPA? [15:10] I vaguely recall a conversation about it, but I can't recall now the reason [15:12] bug 562742 your bug refic? [15:12] Malone bug 562742 in linux "r8169 ethernet MAC address changes in 2.6.32 kernel" [Undecided,New] https://launchpad.net/bugs/562742 [15:12] grrr, wrong bug number [15:13] that is what I get for trying to type it in [15:13] oh, err looking at the wrong bug too. [15:13] bug 562899 [15:13] Malone bug 562899 in linux "linux-headers-*_all.deb is missing on latest mainline builds" [Undecided,New] https://launchpad.net/bugs/562899 [15:13] that one ^ [15:14] nope, just found it [15:15] iirc there was some issue building them or some such. I just can't remember [15:15] we will find out refic [15:16] great, thanks [15:17] np [15:20] why is it saying that "linux is not a genuine package" when i try to report a bug? [15:21] Lucid [15:23] gnomefreak: I believe there are cases where that actually means "you aren't running the latest kernel" [15:23] ah yes because i cant [15:23] I don't think apport can tell the difference between an obsolete package and a non-Ubuntu package [15:24] 2.6.32-20-generic is broken [15:27] gnomefreak: I believe you can override it by editing the problem report in /var/crash [15:27] try deleting the UnreportableReason line [15:27] mdz: thanks [15:27] JFo: it might be worth looking into the issue gnomefreak mentioned, since it's quite common with the kernel (due to its changing package name) [15:27] gnomefreak, when did you update? [15:28] mdz, I agree [15:28] I think he may be suffering from the issue we worked yesterday [15:28] J a little while ago but that is first time in a week [15:28] gnomefreak, hmmm [15:28] are you able to boot at all gnomefreak? [15:29] JFo: i get plymouth but gdm never loads [15:29] gnomefreak, ok, thanks [15:29] dumped into TTY [15:29] hmm [15:32] JFo: the problem is: user tries to boot x.y.z-2, it fails, so they boot back to x.y.z-1, and then can't report the bug [15:32] (and they get a confusing error) [15:32] mdz: i dont have a crawsh report for this issue [15:38] gnomefreak: ah, of course [15:39] gnomefreak: in that case, maybe you could report the bug using a browser, and then run apport-collect to add the details [15:39] mdz: i will do it in a minute, thanks [15:41] mdz, I have added the kernel reporting issue to my notes for our next meeting [15:41] hmm, whats the kernel-img.conf option i need to make the kernel postinst create links in / ? [15:41] I'll bring it up before then, but we will discuss specifically at the meeting [15:42] JFo: thanks. if you can work out what the correct behavior should be, it should be possible to fix it in apport [15:42] gnomefreak, just let me know the bug number once you have it [15:43] mdz, I agree. I'll rely on ogasawara's experience with apport. i'm still new to the tool [15:44] JFo: will do. you want a bug for apport reporting and the gdm not loading? [15:45] gnomefreak, actually yes, if you don't mind [15:45] JFo: not at all [15:45] thanks [15:48] Bug #546743 [15:48] Malone bug 546743 in linux "Blank screen at first boot with ATI ES1000 and 10.04 server" [High,Confirmed] https://launchpad.net/bugs/546743 [15:49] sigh, is there any way to tell the kernel package to not create a backup link for the old kernel ? [15:50] ogra: man kernel-img.conf [15:50] mdz, which package is that in ? i dont seem to have it [15:50] (was indeed the first i tried before asking :) ) [15:50] ES1000 with KMS is generally miserable [15:50] ogra: kernel-package [15:50] ah [15:51] It's a horrible part with all kinds of random quirks in the DDX [15:51] * ogra somehow needs to convince the kernel to be installable in vfat /boot [15:52] brb, rebooting [15:54] hrm, no option that helps me to avoid the linking of backup kernels :/ [15:54] ogra, what stops it being so installable [15:54] dpkg: error processing /linux-image-2.6.33-500-omap_2.6.33-500.5_armel.deb (--install): [15:54] unable to make backup link of `./boot/vmlinuz-2.6.33-500-omap' before installing new version: Operation not permitted [15:55] apw, if it would mv instead of linking that would solve it [15:55] apw: vfat doesn't support symlinks [15:55] or hardlinks for that matter [15:55] ogra: I thought you were talking about something else [15:55] seems unusual to want to do that [15:55] mdz, no, i'm trying to get omap working [15:55] ogra: that's not the maintainer script; that's dpkg itself [15:56] oh [15:56] most systems which need to use FAT too contain their bootable kernels [15:56] install them into a normal /boot and then cp them into the fat at 'flash' time [15:56] yes, I think a vfat /boot is perhaps not the best solution [15:57] ogra, yeah thats dpkg ... i doubt that is ever going to work [15:57] hmm [15:57] on efi systems i used to work on they mounted the efi /boot thingy onto /boot/efi and did magic to get them into there at 'grub-install' type time [15:57] dpkg needs to do that in order to safely replace the file [15:58] well, sadly kernel-img.conf preinst_hook seems deprecated [15:59] else that could call something that mv's [15:59] I don't think it will help you anyway [15:59] effectively the vmlinuz file is never used anyway [15:59] mdz, if i manually mv it i can dpkg -i linux-image-... [15:59] ogra: then why do you need a vfat partition? [15:59] the only prob is that its a link [16:00] ogra: yes, but if you mv it, you are reintroducing the race condition that dpkg is trying to avoid [16:00] mdz, because u-boot on beagle doesnt support any other FS [16:00] it needs to read uImage from a vfat partition ... [16:00] ogra: surely it's using the vmlinuz file though? [16:00] and that also needs to essentially be the first one on the SD [16:01] currently its using nothing ... i havent written the flash-kernel code for that yet [16:01] ogra: it sounds to me like you just want to mount your vfat partition somewhere other than /boot [16:01] but by default it treis to run boot.scr which in turn loads uImage [16:01] ogra, so why can't we mount that on /boot/efi; cp uImage* /boot/efi; umount /boot/efi as part of the flash part [16:01] mdz, and copy uImage there ? [16:01] as apw says [16:01] well, we surely can [16:01] s/efi/something sane/ [16:01] uboot [16:01] or even leave it mounted, and just copy as needed [16:02] now i'm scared ... that requires a partman module ... [16:02] not what i was hoping for last minute before freeze :) [16:02] but yes, that sounds very sane [16:12] JFo: bug 563094 and bug 563102 [16:12] Malone bug 563094 in linux "unable to report a "linux" bug using ubuntu-bug" [Undecided,New] https://launchpad.net/bugs/563094 [16:12] Malone bug 563102 in linux "Unable to get any GUI other than Plymouth" [Undecided,New] https://launchpad.net/bugs/563102 [16:15] thanks gnomefreak [16:15] JFo: np [16:18] bug #542208 [16:18] Malone bug 542208 in linux "Please blacklist i830 from Kernel mode-setting" [Critical,In progress] https://launchpad.net/bugs/542208 [17:18] apw: around? [17:27] amitk, [17:30] apw: fsl-imx51 has control-scripts directory inside debian.fsl-imx51 with all the prerm/postrm, etc. Should it? mvl and ti don't. [17:31] amitk, yes, cause its 'old-school' 2.6.31 way of doing things [17:31] mvl and ti are both new-school and share one [17:31] amitk, thats on my radar for 'soon' as i want to pull back the new debian abstraction to karmic and thus to the .31 branches [17:31] aah, keep forgetting [17:32] yeah its a nightmare and one reason i want to roll it back ... so it _is_ the same and not confusing [18:11] jjohansen: still getting the oops on -21.31 [18:11] hggdh: :( [18:12] jjohansen: want me to run a mailine? [18:12] hggdh: sure [18:12] jjohansen: which one? Any prefs? [18:12] hggdh: not at the moment [18:13] jjohansen: will go to the latest 2.26.33, then. 2.26.34 would not even boot on my machine [18:13] hggdh: actually it would be good to try 32 and 33 [18:14] jjohansen: OK. latest .32 and latest .33 [18:14] yes please [18:14] will do [18:22] bug 563156 [18:22] Malone bug 563156 in linux "laptop runs hot, shorter battery life, fan always on - lucid beta2" [Undecided,New] https://launchpad.net/bugs/563156 [18:27] the mainline headers build still looks broken [18:27] linux-headers-2.6.34-020634rc4-generic depends on linux-headers-2.6.34-020634rc4 [18:27] and the latter doesn't exist [18:29] jjohansen: 2.6.32-11-lucid does not install the headers, missing a package [18:30] hggdh: hrrmm, okay I'm not sure what is up there will have to poke what of 2.6.33 [18:35] yeah some headers are missing from the mainline builds at the mo, not had a chance to look why [18:36] its on my list, but its freeze today ... so ... not today [19:12] firmware request: this should be easy enough to fix, the firmware is in gregkh's git repo: https://bugs.launchpad.net/bugs/508746 [19:12] Malone bug 508746 in linux "rtl819xE:request firmware fail!" [Low,Confirmed] [19:21] tumbleweed, its 3 hours to freeze its unlikely anything will change before then [19:30] apw: yeah, somebody just gave me a laptop to look at that, which was suffering from that bug 5 hours ago :0 [19:30] tumbleweed: Error: Could not parse data returned by Malone: list index out of range [19:30] :) [19:30] heh === yofel_ is now known as yofel === BalSak is now known as kern00b [22:54] hi guys. I'm not really a very big kernel expert, so I would rather ask the pro's. I have a box that's presently in production, and I/we had to manually roll in the NIC driver/module. If/when I dist-upgrade the kernel, would I have to manually re-roll the driver into the new kernel, or will it get ported in the upgrade? [23:01] kern00b: if the new kernel doesn't include the driver you will need to re-roll it [23:01] the new kernel may have the driver you need though [23:01] basically each kernel gets its own set of modules [23:02] jjohansen: thanks