=== infinity2 [n=adconrad@cerberus.0c3.net] has joined #ubuntu-kernel === madc [n=d@adsl-156-75-197.msy.bellsouth.net] has joined #ubuntu-kernel === anibal [n=anibal@debian/developer/anibal] has joined #ubuntu-kernel === marcin_ant [n=marcin@194.114.146.126] has joined #ubuntu-kernel === infinity [n=adconrad@cerberus.0c3.net] has joined #ubuntu-kernel === infinity2 [n=adconrad@cerberus.0c3.net] has joined #ubuntu-kernel === persia [n=ubuntu@p1033-ipbf37marunouchi.tokyo.ocn.ne.jp] has joined #ubuntu-kernel [03:07] linux-image-2.6.20-15-lowlatency 2.6.20-15.25 fails to boot on my hardware. Are there still known issues with drive detection (at which you are all hard at work), or should I file a bug? [03:08] Please file a bug, especially if you can reproduce it with the non-lowlatency kernel [03:08] mjg59: OK. Thanks. === persia [n=ubuntu@p1033-ipbf37marunouchi.tokyo.ocn.ne.jp] has left #ubuntu-kernel [] === infinity [n=adconrad@cerberus.0c3.net] has joined #ubuntu-kernel === TheMuso [n=luke@ubuntu/member/themuso] has joined #ubuntu-kernel === zul [n=chuck@ubuntu/member/zul] has joined #ubuntu-kernel === persia [n=ubuntu@p1033-ipbf37marunouchi.tokyo.ocn.ne.jp] has joined #ubuntu-kernel === infinity2 [n=adconrad@cerberus.0c3.net] has joined #ubuntu-kernel [04:10] publishing -15.25/i386 now; ETA half an hour or so [04:12] cjwatson: I hope you get to have some sleep soon? [04:12] it's a point of view [04:13] oops, meant to publish d-i source at the same time [04:13] oh well, more runs needed anyway === ..[topic/#ubuntu-kernel:cjwatson] : Ubuntu kernel development discussion ONLY | Kernel Wiki: https://wiki.ubuntu.com/KernelTeam | 2.6.20-15.25 Uploaded - If you have a BUG, it's too late to get it in for release. [04:19] I kind of need to test alternate/ps3 too which means waiting for d-i to be available [04:19] might do that tomorrow though [04:21] cjwatson: You got a work PS3? [04:21] Man. I don't get the cool hardware any more. [04:23] it's less cool when you don't have an HDTV [04:24] and have to hook it up to vlc in tiny-resolution mode ;) [04:24] hehe [04:27] You need a work HDTV [04:28] I do [04:28] me too? :P [04:28] though i'd rather a 30" dell kthnxbye [04:28] :) [04:43] -15.25/i386 up and on all archive.ubuntu.com mirrors [04:46] cjwatson, thanks so much. [04:47] hey, you guys did the real work ... [04:47] You're the one awake at 3AM [04:47] now it's just language packs and hotkey-setup and then I can build livefses [04:47] when I've kicked that off I'll almost certainly crash [04:48] won't take all that much longer with three buildds munching on it === infinity [n=adconrad@cerberus.0c3.net] has joined #ubuntu-kernel === defendguin [n=supertux@ip70-180-86-93.no.no.cox.net] has joined #ubuntu-kernel [05:12] mjg59: i think that last kernel update today fixed my issue [05:12] defendguin: Excellent [05:12] It /should/ fix all of them [05:13] i thought yesterday's update was the one with the patch? [05:16] mjg59: did something change in hal that made the original update not work properly? [05:18] No [05:18] But the first upload yesterday may have been missing that line [05:18] heh [05:20] i remember reading the changes in the update manager and seeing the patch was in. needless to say i was disappointed when the card didn't work [05:21] http://librarian.launchpad.net/7325843/buildlog_ubuntu-feisty-ia64.linux-source-2.6.20_2.6.20-15.25_FAILEDTOBUILD.txt.gz [05:21] argh! [05:23] -15.24 is there === cjwatson is kinda tempted to ignore this for release now [05:24] -14 kernels removed === BenC [n=bcollins@debian/developer/bcollins] has joined #ubuntu-kernel [05:40] i think you guys cleared up all the issues i had with this laptop being supported under linux this cycle === purpleposeidon [n=poseidon@adsl-71-141-90-16.dsl.snfc21.sbcglobal.net] has joined #ubuntu-kernel [05:42] BenC: 04:21 http://librarian.launchpad.net/7325843/buildlog_ubuntu-feisty-ia64.linux-source-2.6.20_2.6.20-15.25_FAILEDTOBUILD.txt.gz [05:42] BenC: I propose to ignore that for release, but FYI [05:42] looks like the ABI-ignore hack failed === infinity [n=adconrad@cerberus.0c3.net] has joined #ubuntu-kernel === purpleposeidon [n=poseidon@adsl-71-141-90-16.dsl.snfc21.sbcglobal.net] has left #ubuntu-kernel ["/dev/null'd!"] === infinity2 [n=adconrad@cerberus.0c3.net] has joined #ubuntu-kernel === jml [n=jml@59.167.203.115] has joined #ubuntu-kernel === jml [n=jml@59.167.203.115] has joined #ubuntu-kernel === alessio [n=alessio@host-84-222-111-33.cust-adsl.tiscali.it] has joined #ubuntu-kernel === lfittl [n=lfittl@213.129.230.12] has joined #ubuntu-kernel === jml [n=jml@59.167.203.115] has joined #ubuntu-kernel [06:27] cjwatson: Yeah, I mentioned earlier, it's fixed in git now that I have the ABIs from -15.24 [06:27] fix it first update [06:38] BenC, cjwatson it will still make ia64 uninstallable because of the ABI change in d-i [06:38] would be much easier to keep it in sync [06:39] oh actually ABI is the same [06:39] just version [06:39] hmm [06:39] yeah [06:39] can work === tuxmaniac [n=aanjhan@unaffiliated/tuxmaniac] has joined #ubuntu-kernel === orangey [n=orangey@bas3-london14-1096778599.dsl.bell.ca] has joined #ubuntu-kernel [07:02] hey all! [07:04] I'm trying to figure out a bug that appears to be a regression in the kernel. [07:05] the sound on this NC6230 used to work after suspend, but stopped working around the -13 mark? [07:06] However, I'm honestly not even sure a) if it was intentional that the bug was fixed; or b) how to track down which kernel module broke it. [07:11] orangey: boot into the kernel with working suspend; check out upstream hg alsa-{driver,kernel}; execute hgcompile from alsa-driver; install it; confirm that it's still working with current upstream alsa === defendguin [n=supertux@ip70-180-86-93.no.no.cox.net] has joined #ubuntu-kernel [07:12] crimsun: stupidly, I wiped the old kernel ; ) I'll remake it and go from there. Thank you. [07:12] the hotkey-setup update that got released does it include those bugs mjg59 fixed the other night? [07:12] crimsun: what do you mean by hg alsa-driver ? [07:13] i remember BenC said they could be included [07:16] orangey: do you know how i can view the changelog of the hotkey-setup update? [07:16] defendguin: https://lists.ubuntu.com/archives/feisty-changes/2007-April/008524.html [07:16] defendguin: also, https://lists.ubuntu.com/archives/feisty-changes/2007-April/008483.html [07:17] perfect [07:18] i just wanted to make sure i didn't screw it up with the update :-) [07:18] orangey: what's unclear? [07:18] orangey: also, this discussion is veering into support instead of development. Contact me in #ubuntu+1 if you have further questions. [07:18] thanks crimsun [07:22] crimsun: pretty much the whole workflow. upstream hg alsa-driver ? I'm actually not sure what that is, other than that possibly I should get the deb from debian unstable and install it or some such.. [07:23] crimsun: I'm on +1 now. === ThiefOfBaghdad [n=aanjhan@unaffiliated/tuxmaniac] has joined #ubuntu-kernel === gnomefre1k [n=gnomefre@adsl-221-54-53.rmo.bellsouth.net] has joined #ubuntu-kernel === anibal [n=anibal@debian/developer/anibal] has joined #ubuntu-kernel === Lure [n=lure@clj46-234.dial-up.arnes.si] has joined #ubuntu-kernel [08:40] zul: have you gotten a new packages exception for openvz? [08:41] heh, it's 3am for him. [08:41] sleep is for the weak [08:41] :P [08:41] weak from exhaustion, yes. [08:42] so openvz is going to be in the archive? [08:42] or might* [08:42] joejaxx: I didn't say that. === ajmitch would guess he hasn't got an exception - first I've heard of it === _ion [i=johan@kiviniemi.name] has joined #ubuntu-kernel [09:33] <_ion> benc: linux-source-.../debian/firmware/ipw2200/LICENSE says "Your rights to redistribute the Software shall be contingent upon your installation of this Agreement in its entirety in the same directory as the Software". Should the license be installed to /lib/firmware/...? [09:38] <_ion> Also, considering the limitations the license seems to put to the redistribution of the firmware, would restricted be a better place for it? I might be wrong, i didn't read the license carefully. === doko [n=doko@dslb-088-073-111-226.pools.arcor-ip.net] has joined #ubuntu-kernel === gurke_ [n=gurke@e178010117.adsl.alicedsl.de] has joined #ubuntu-kernel === ivoks [n=ivoks@stream.grad.hr] has joined #ubuntu-kernel === ivoks [n=ivoks@stream.grad.hr] has joined #ubuntu-kernel === heno [n=henrik@ubuntu/member/heno] has joined #ubuntu-kernel === gnomefreak [n=gnomefre@ubuntu/member/gnomefreak] has joined #ubuntu-kernel === jml [n=jml@59.167.203.115] has joined #ubuntu-kernel === pmjdebruijn [n=pmjdebru@pmjdebruijn.xs4all.nl] has joined #ubuntu-kernel === afflux [n=k@pentabarf.de] has joined #ubuntu-kernel === dystopianray [n=dystopia@adsl117-124.dyn.koalatelecom.com.au] has joined #ubuntu-kernel === dystopianray [n=dystopia@adsl117-124.dyn.koalatelecom.com.au] has left #ubuntu-kernel ["Kopete] === bdgraue [n=bdgraue@dyndsl-080-228-221-247.ewe-ip-backbone.de] has joined #ubuntu-kernel [12:08] I installed the kernelupdated to 2.6.20-14.23 yesterday, which broke my system (something with ata)... but I removed the older kernels that were still installed for cleaning up. [12:08] now i'm in a livecd [12:08] chrooted into feisty and updated to -15.25 [12:08] and I still have the same ata error (revalidation failed, n_sectors mismatch) [12:10] same here updated and now http://www.ubuntuusers.de/paste/9217/ [12:11] another confirmation here: https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.20/+bug/106063/comments/80 === mvo [n=egon@p54A654C7.dip.t-dialin.net] has joined #ubuntu-kernel [12:23] if someone need more specific, please ask--- [12:26] I expect the kernel devs are still asleep, but please hang around until they return [12:27] BenC, kylem, mjg59: ^ still problems with 15.25 it seems === m0rg0th [n=manugarg@219.64.120.181] has joined #ubuntu-kernel === ivoks [n=ivoks@161.53.50.78] has joined #ubuntu-kernel [12:44] bdgraue: could you try putting libata.ignore_hpa=0 on the kernel command line, please? [12:45] afflux: you too [12:47] going to have a try, will be back in about 10-20 minutes [12:49] cjwatson: ok, i'll try, moment please === ivoks_ [n=ivoks@38-10.dsl.iskon.hr] has joined #ubuntu-kernel [12:50] if that works, dmesg and lspci -vvnn would be greatly appreciated [12:50] possibly a new bug each to avoid confusion === cjwatson <- not a kernel developer, merely a manager of kernel developers, but since most of them are asleep right now ... [12:51] http://librarian.launchpad.net/7329378/00001.jpg is interesting, definitely suggests libata.ignore_hpa=0 should work, but seems to be applying the wrong hpa_sectors size [12:53] doesn't work for me [12:54] going to have a shower, will be back in 20 mins [12:54] a photo of the error would be good [12:54] cjwatson: doesn't work for me too [12:55] just to be clear, http://www.ubuntuusers.de/paste/9217/ is just a consequence, not the real error [12:55] we need the sort of ata debugging messages you see in http://librarian.launchpad.net/7329378/00001.jpg [12:55] might have to boot without 'quiet' on the kernel command line to get that [12:55] probably without 'splash' too [12:56] could 'debug' help? [12:56] no [12:56] looks like the module parameter stuff isn't working right :-/ [12:58] lspci -vvnn from a live CD might help [12:58] cjwatson: i'll give u the photo [12:58] gimme some time please :-) === Nafallo [n=nafallo@ubuntu/member/nafallo] has joined #ubuntu-kernel [01:00] kylem: why don't the ata_set_native_max_address* functions set ATA_TFLAG_LBA? [01:00] (not sure if that's relevant though) [01:11] cjwatson: http://phpfi.com/226658 [01:11] cjwatson: will make a photo later [01:14] cjwatson: lspci -wnn give me an lspci: invalid option -- w [01:15] bdgraue: twice a v, not a w please [01:15] lspci -vvnn [01:15] ok, afflux is using sata_nv [01:15] cjwatson: http://img383.imageshack.us/img383/5466/pict0002cs9.jpg there is the photo === mrec [n=mrechber@165.204.76.244] has joined #ubuntu-kernel === cjwatson has to go out now; if you could hang around until actual kernel developers show up, that'd be good [01:16] i'll leave in 30 mins too... [01:17] hi, I'm really curious I upgraded my notebook to ubuntu, also recompiled the kernel but it's not possible to start Xgl the whole screen just starts to flicker around ... so I guess the livecd includes some specials for that device? [01:19] bdgraue: hmm, there don't appear to be any of the hpa handling messages I was expecting; that just looks like normal boot [01:19] possibly you have enough disks that it scrolls off [01:20] cjwatson: so whats then the problem? [01:21] cjwatson: the cdrom-drives are ide [01:23] cjwatson: http://www.ubuntuusers.de/paste/9222/ hte lspci -vvnn [01:24] is there any documentation about the ubuntu i855gm setup available? for any reason I suspect that some intel engineers did some specials there [01:25] bdgraue: like I say, I'm not a kernel developer, I'm just trying to gather as much information as possible [01:26] the problem for afflux is that the HPA suppression logic appears to be setting the drive size to the wrong value on NVIDIA ATA chipsets [01:26] i hope my informations could help to solve the problem [01:26] it's not entirely clear that yours is the same problem, although you're also using NVIDIA [01:26] haven't heard any concrete non-NVIDIA reports yet [01:28] cjwatson: an other photo is here: http://img150.imageshack.us/my.php?image=pict0004gr3.jpg [01:28] kylem: converting the sector counts to hex in http://librarian.launchpad.net/7329378/00001.jpg is very interesting [01:28] kylem: it's 0x0D00000E instead of 0x0DF94BB0, and 0x2E00002F instead of 0x2E9390B0 [01:29] seems like an obvious pattern === ivoks [n=ivoks@161.53.50.78] has joined #ubuntu-kernel [01:31] mrec: No, we don't have any special 855 code [01:32] bdgraue: thanks, that's more useful [01:33] that's exhibiting the same pattern of corruption [01:33] I see, I booted from the CD and started the local X, I have the same effects [01:33] 0x1D1C5970 -> 0x1D00001E [01:33] seems like the upgrade wasn't clean [01:33] looks like sata_nv is very consistently buggering the sector count [01:35] bdgraue: could you try booting with sata_nv.adma=0 on the kernel command line? [01:35] i can [01:35] mom please [01:38] cjwatson: don't work [01:39] hmm, what exactly was your kernel command line? === Nafallo [n=nafallo@ubuntu/member/nafallo] has joined #ubuntu-kernel [01:39] mom [01:39] and could you take photos of the output with hpa messages in it again? I'd like to confirm that it's the same problem === mdz [n=mdz@ip-81-1-98-32.cust.homechoice.net] has joined #ubuntu-kernel [01:40] ok I manually removed the xorg directories and reinstalled them, now it seems to work [01:42] cjwatson: kernel /boot/vmlinuz-2.6.20-15-generic root=UUID=01b05012-0b6d-4ac6-8c72-cb146fd64f9f ro locale=de_DE sata_nv.adma=0 [01:44] cjwatson: sata_nv and libata are modules - do we have sensible code in initramfs-tools to deal with that? (I remember it being wishlist, I don't remember whether we did or not) [01:50] cjwatson: http://img219.imageshack.us/my.php?image=pict0007ro7.jpg [01:51] That's very weird. [01:53] with -15 everything is sd?. thanks :-) [01:59] cjwatson: Exactly the same pattern with bdgraue's drive [02:00] Something's squashing most of the bits [02:01] i wish i had an older kernel left on the sys :-( [02:02] (Oh, sorry, you'd already commented on that) [02:02] :-) [02:13] ah === tuxmaniac [n=aanjhan@unaffiliated/tuxmaniac] has joined #ubuntu-kernel [02:15] mjg59: initramfs-tools> I'm not sure [02:17] bdgraue: https://launchpad.net/ubuntu/+source/linux-source-2.6.20/2.6.20-14.22 - "Builds of" drop-down in the top-left, you can get old packages from that [02:19] mjg59: can you tell whether it's sata_nv or pata_amd in use? I couldn't quite tell [02:21] I think it's sata_nv in bdgraue's case at least [02:21] BenC: could you push the various -15 git branches? [02:24] mjg59: I don't quite know how module options work. What would initramfs-tools need to do? [02:26] mjg59: the way that the top three bytes are being consistently copied to the bottom three bytes (kernel code then adds one to it) is particularly suspicious [02:27] I'm assuming it's LBA48 since at least some of the sizes in question exceed 28 bits [02:28] it says "lba48" in the screenshot too === tuxmaniac [n=aanjhan@unaffiliated/tuxmaniac] has joined #ubuntu-kernel [02:33] shouldn't there be a 0x0f masking in the set_native_max_address_ext? That bit looks suspicious. [02:33] Mithrandir: where? [02:33] libata-core.c [02:33] no I mean where 0x0f [02:33] tf.lbah = new_sectors, etc [02:34] I thought lbah would only get four of the bits set there? [02:34] or is lba48 sane? [02:34] Mithrandir: the reason set_native_max_address has a 0x0f mask is because it's the top 4 bits of the LBA28 [02:34] but LBA48 divides exactly by 8 [02:34] ok, false alarm then. [02:34] I still very strongly suspect adma here [02:35] quite tempted to build a kernel with adma disabled [02:35] in fact, yeah, why don't I kick that off on ronne === zul [n=chuck@CPE0006258ec6c1-CM000a73655d0e.cpe.net.cable.rogers.com] has joined #ubuntu-kernel [02:35] try that, then? You have the ccache from yesterday, so it should be fast enough? === Mithrandir goes out for a little bit. [02:35] I'm not sure it was ccached unfortunately === Amaranth [n=travis@ubuntu/member/Amaranth] has joined #ubuntu-kernel [02:36] one build shouldn't take all that long even from scratch though === cjwatson looks suspiciously at nv_adma_tf_to_cpb === tuxmaniac [n=aanjhan@unaffiliated/tuxmaniac] has joined #ubuntu-kernel [02:40] cjwatson: hoe can i get the old kernel in the chrooted system? [02:40] how [02:41] get => download or get => install? [02:42] download and install with dpkg or something else or install via apt if possible, cjwatson [02:44] find URL, use wget to fetch it, dpkg -i [02:45] ok, thx [02:47] bdgraue: would you be able to help out with testing unofficial fixes at some later point? [02:47] hmm I noticed that ubuntu feisty doesn't support the highest resolution on my notebook, how about trying to detect the notebook and patching the bios to get it work? [02:48] cjwatson: if i have time, yes [02:48] (I tested it with the temporary biospatch, and it seems to work fine.. and it looks way better) [02:48] i would be an honor to help [02:49] I'm still trying to figure out what the hell's going on, and I'm planning to be out this afternoon [02:49] kylem may stand a better chance once he wakes up [02:51] 3 ours from now i have to go... then i have no more time :-( [02:51] then only tomorrow [02:56] cjwatson: do i have to install some of the modules too, or only image and headers? [02:57] from the older kernel... [03:02] bdgraue: just the image, and you may need to go and hunt down a matching linux-restricted-modules too if you rely on non-free video drivers or whatever [03:04] hunt down a matching linux-restricted-modules? [03:05] cjwatson: what does that mean, i have to find one, somewhere? === ThiefOfBaghdad [n=aanjhan@unaffiliated/tuxmaniac] has joined #ubuntu-kernel === Keybuk [n=scott@wing-commander.netsplit.com] has joined #ubuntu-kernel === rpereira [n=rpereira@ubuntu/member/rpereira] has joined #ubuntu-kernel === bdgraue_ [n=bdgraue@dyndsl-080-228-221-247.ewe-ip-backbone.de] has joined #ubuntu-kernel [03:38] cjwatson: thx, i have the old 2.6.20-14.22 running [03:42] cjwatson, test kernel building [03:58] kylem: cool, what are you trying? [03:59] cjwatson, what you pointed out above (setting TFLAG_LBA) [03:59] ok [03:59] I've got a build running that turns off adma [03:59] (in sata_nv) [03:59] we're still seeing issues there? [04:00] the only bug i saw in my mbox in this dumping was the dude lying about running 15.25 :/ [04:00] kylem: yes, consistently with sata_nv [04:00] bug#? [04:00] something is blatting the high-order-bytes of the LBA48 over the top of the low-order three bytes [04:01] er dunno about a bug number, but see e.g. http://librarian.launchpad.net/7329378/00001.jpg [04:01] ok. [04:01] it's all in scrollback here [04:02] there was another example as well - in both cases LBA48 0000AABBCCDD was transformed into 0000AA0000AA (+1) [04:03] in the middle of ata_exec_internal so presumably by the driver [04:03] hm, ok. [04:03] I've only heard of it with nv so far [04:04] but I'm not fully up to date on bugs [04:06] ergh. [04:08] I also think module options passed on the kernel command line just plain don't work in feisty [04:17] damnit, my build failed for no obvious reason [04:17] + xargs strip --strip-debug [04:17] make[1] : *** [install/linux-image-2.6.20-15-generic] Error 1 [04:17] + xargs strip --strip-debug [04:17] make[1] : *** [install/linux-image-2.6.20-15-generic] Error 1 [04:17] (sorry for dup) [04:17] ok, screw this, I'm going out. call me if it's urgent === ubeyde [n=ubeyde@88.224.185.152] has joined #ubuntu-kernel === ubeyde [n=ubeyde@88.224.185.152] has left #ubuntu-kernel [] === EtienneG [n=etienne@ubuntu/member/EtienneG] has joined #ubuntu-kernel === defendguin [n=supertux@ip70-180-86-93.no.no.cox.net] has joined #ubuntu-kernel === ivoks [n=ivoks@ivoks.grad.hr] has joined #ubuntu-kernel === reitblatt [n=mark@resnet-50-180.dorm.utexas.edu] has joined #ubuntu-kernel === macd [n=d@adsl-156-75-197.msy.bellsouth.net] has joined #ubuntu-kernel === Aikurn [n=aikurn@155.Red-88-2-203.staticIP.rima-tde.net] has joined #ubuntu-kernel [05:22] hi === Lure [n=lure@clj46-234.dial-up.arnes.si] has joined #ubuntu-kernel [05:28] kylem: do we have an issue with -15.25? [05:28] still reading scrollback [05:30] looks like it [05:31] and it looks like libata.ignore_hpa isn't making it to the module in some cases. [05:31] how does the kernel handle cmdline opts to modules? [05:32] no clue. it didn't change. [05:32] I thought module.XXX would only work for built in modules [05:32] that was fixed in 2.6, AIUI. [05:32] i thought the modprobe handled it by inspecting /proc/cmdline [05:33] at least I remember rlove and rusty claiming so at LCA 2004 during the kernel hacking tutorial. [05:33] Module parameters for modules that are built into the kernel image [05:33] are specified on the kernel command line with the module name plus [05:33] '.' plus parameter name, with '=' and value if appropriate, such as: [05:34] I don't think cmdline for actual modules is supported [05:34] It used to be you couldn't pass options to built-in modules, and that's probably what they fixed [05:35] there's more than two ways you can do it, i have no idea how universally stupid it is if they haven't implemented it. [05:35] kylem: so what's the extent of this bug? [05:36] BenC, some controllers (nv apparently?) are reporting bogus sizes back from ata_set_max_address{,_ext} [05:37] what's with these 1Meg HPAs [05:38] appears to be fairly common on sata disks. [05:39] it's weird, the size shows in "native size increased" is not the same size as the one read in ata_hpa_resize [05:39] which means it's not the one set in n_sectors_boot [05:39] the native sized increased size is the one returned value read out of the taskfile [05:40] whatever size it was increased to is the one we need to put in n_sectors_boot [05:40] then we need to use that value to store the n_sectors_boot [05:40] err, I mean the n_sectors [05:40] not until we figure out why it's all friggity fucked. [05:41] Not sure why we are getting a different result than what we ask for [05:41] kylem: maybe we should fail to change n_sectors if the results for the set isn't what we asked for? [05:42] colin pointed out that the code wasn't setting ATA_TFLAG_LBA [05:42] in set_max [05:42] I thought you said that was ok because of 0x40? [05:42] i know. [05:42] worth a try I guess [05:43] I wish I had a drive with HPA to test this better [05:43] my guess is that some of the taskfile dispatch paths are setting it, and some aren't depending on the value in that field [05:43] which is why it works fine on so many systems [05:45] well the set obvious works...the n_sectors mismatch shows that ata_id_n_sectors() is returning the value of the size we did in set [05:45] yeah. [05:50] kylem: Is it ever possible that if our set succeeds we should get a different value that what we asked it to set? [05:50] or I mean, should it be possible === ivoks [n=ivoks@ubuntu/member/ivoks] has joined #ubuntu-kernel [05:51] can we assume that since the set succeeds, and from these reports we can see that it succeeds, could we just set n_sectors to the value we set instead of relying on this return value? === johanbr [n=j@blk-224-156-151.eastlink.ca] has joined #ubuntu-kernel [05:52] ew [05:52] assuming this TFLAG_LBA doesn't work [05:52] that sounds risky. [05:52] i don't think so, perhaps we should set it to the value in ata_id_n_sectors afterwards [05:52] kylem: Does that get updated immediately? [05:52] aside from setting the max sectors limit, that's all that really changes [05:52] yes [05:52] then that might be best [05:53] the only thing we've seen not to be reliable so far is the return value from the set call...it seems we can definitely rely on everything else [05:53] yeah, the 0x40 definitely is the bit needing set unless some deeper bit is unsetting it based on the value of tf->flags [05:53] I don't like it, but it may be the way to go, if we can't find the problem with the return value of set [05:54] oh. [05:54] we have to re-identify the drive of course. [05:54] since the id is cached. [05:54] that's why we used the return value. [06:03] that's what I thought....else the whole n_sectors_boot thing is pretty bogus :) [06:04] what's the call to id the drive? [06:06] rc = ata_dev_read_id(dev, &dev->class, ATA_READID_POSTRESET, [06:06] dev->id); [06:16] builds almost done (concurrently for i386 and amd64) [06:16] Do we have testers? === tuxmaniac [n=aanjhan@unaffiliated/tuxmaniac] has joined #ubuntu-kernel === lazka [n=lazka@85-124-173-127.dynamic.xdsl-line.inode.at] has joined #ubuntu-kernel === heno tries to round up some testers in the Feisty section of the forums === trejack [n=tre@24-119-246-56.cpe.cableone.net] has joined #ubuntu-kernel [06:30] I was told to check in here if I was having problems with the 2.6.20-15 kernel [06:31] only if it's -15.25 === sertmann [n=sertmann@26-63-150-213.static.dsl.webpartner.net] has joined #ubuntu-kernel [06:31] It is [06:31] me2.. [06:31] cat /proc/version_signature [06:31] or dpkg -l linux-image-2.6.20-15-generic [06:32] heya, just responding to a call for sata_nv users with 2.6.20-15 kernel problems come here [06:32] ii linux-image-2. 2.6.20-15.25 Linux kernel image for version 2.6.20 on x86 [06:33] amd64 or i386? [06:33] what's the problems you are having? [06:33] amd64 [06:33] Haven't been able to boot since upgrade from -13 to -14 [06:33] ditto [06:33] do you have screenshots of the problems so we can make sure it is the same thing? [06:33] well, i can boot into the -13 kernel still [06:33] or maybe gain more info? [06:34] sertmann: do you have the latest -15.25? [06:34] Get ata2.00:qc timeout(cmd0xef) [06:34] trejack: is this macbook? [06:34] the latest one available from apt anyway [06:34] Then ata 2.00: failed to set xfermode (err_mask=0x4) [06:34] do you guys need anyone with "failed to set xfermode"? no nv_whatever here :P [06:34] No its ASUS P5NSLI [06:34] sertmann: please make sure it is -15.25, so we don't waste time on a bug that may already be fixed [06:35] Finally get to busybox shell [06:35] BenC: that's what it says menu.lst anyway [06:35] sda and sdb are listed in /dev [06:35] but not sda1,sda2,etc [06:35] sertmann: menu.lst shows -15, it doesn't say it's -15.24 or -15.25 [06:36] sertmann: please check, it's very important [06:36] transcribed this from recovery mode: from https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.20/+bug/106063/comments/88 [06:36] /boot/initrd.img-2.6.20-15-generic [06:36] how do i check the exact version? [06:36] sertmann: I gave you two commands already :/ [06:36] sertmann: dpkg -l linux-image-2.6.20-15-generic [06:38] my drives are ide, not SATA, but motherboard does have SATA controllers. === jpon [n=jpniane@neu67-3-82-239-80-181.fbx.proxad.net] has joined #ubuntu-kernel [06:39] BenC: from that output it does look installed, it lists it anyway [06:39] sertmann: can you paste it please [06:39] sertmann: I need to see the version, not whether it is installed or not [06:39] Desired=Unknown/Install/Remove/Purge/Hold [06:39] | Status=Not/Installed/Config-files/Unpacked/Failed-config/Half-installed [06:39] |/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad) [06:39] ||/ Name Version Description [06:39] +++-==============-==============-============================================ [06:39] ii linux-image-2. 2.6.20-15.25 Linux kernel image for version 2.6.20 on x [06:40] ok, that's the info we needed [06:40] thanks, so you have the latest and are still having problems [06:40] yes [06:40] yes [06:41] can you guys pastebin dmesg from the last kernel that worked for you [06:41] and also your lspci -vvn output [06:41] http://people.ubuntu.com/~kyle/kernels/feisty/linux-image-2.6.20-15-generic_2.6.20-15.25_amd64.deb [06:42] that's for anyone having "mismatch" messages in your dmesg [06:42] please test kyle's images [06:42] https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.20/+bug/106418 has it the dmesg, hang on while I do lspci [06:42] sertmann: see http://paste.ubuntu-nl.org/ === felip1 [n=felipe@host28-99-dynamic.7-87-r.retail.telecomitalia.it] has joined #ubuntu-kernel [06:44] trejack: please test kyle's kernel image(s) [06:44] http://paste.ubuntu-nl.org/15606/ [06:44] i have a sata_nv controller in this box, but it's not used, and my ide hd are being seen as sd* [06:45] felip1: then you don't have a bug [06:45] ok BenC :) [06:45] I wish I had a SATA so I could help debugging :-) [06:45] except my feisty isn't booting [06:45] http://paste.ubuntu-nl.org/15607/ for my lspci [06:46] bah, lspci output is to long for the terminal === mdz [n=mdz@yttrium.canonical.com] has joined #ubuntu-kernel [06:46] should i just paste the lines i do get? [06:46] felip1: then let's get down to the real bug [06:46] BenC: do you already know of any issues with edgy->feisty and non-sata hd? [06:46] felip1: if your drives show up, why aren't they mounting [06:46] BenC: I'm not using amd64, try it anyway? [06:46] sertmann: send it to a text file with > output.txt [06:47] trejack, no [06:47] felip1: does your root= in /boot/grub/menu.lst shows /dev/... or UUID=...? [06:47] http://people.ubuntu.com/~kyle/kernels/feisty/linux-image-2.6.20-15-generic_2.6.20-15.25_i386.deb [06:47] trejack: No, use the one kyle just pasted [06:47] BenC: UUID [06:47] trejack, ^-- i386 image is there [06:47] OK - I'll report back [06:47] felip1: does your /etc/fstab reference any /dev/hdX devices? [06:48] uhm good point, i guess so... let me check ;) [06:48] felip1: they should use UUID= as well [06:48] yes theys are [06:48] sertmann: lspci -vvn > lspci.txt, and post that [06:48] got it here: http://paste.ubuntu-nl.org/15608/ [06:50] BenC: how are we doing? [06:50] sertmann: Please try kyle's images [06:50] I've just had to spend the past several hours buying a new laptop because mine has quit [06:50] BenC: they use UUID in fact === lazka [n=lazka@85-124-46-101.dynamic.xdsl-line.inode.at] has joined #ubuntu-kernel [06:50] felip1: Boot to recovery mode and see if you can get a digital photo of the final screen === trejack [n=tre@24-119-246-56.cpe.cableone.net] has left #ubuntu-kernel [] [06:51] BenC: ok === felip1 [n=felipe@host28-99-dynamic.7-87-r.retail.telecomitalia.it] has left #ubuntu-kernel [] [06:51] mdz: white smoke and blue flames? :-) [06:52] mine did, was kind of scary :-P [06:52] ok; downloading now [06:52] no, the display is failing [06:53] ouch [06:53] mdz: we are chasing more sata_nv bugs :/ [06:54] mdz: I'm almost to the point of saying that sata_nv just doesn't get hpa resize rights [06:54] heh, it's *not* the driver [06:54] that's not how these things work... [06:54]