/srv/irclogs.ubuntu.com/2010/12/16/#ubuntu-kernel.txt

=== bjf is now known as bjf[afk]
=== jjohansen is now known as jj-afk
=== pgraner-afk is now known as pgraner
=== yofel_ is now known as yofel
cactuarHi all! I submitted a bug for Maverick (#647128) and the triager there told me to try using the mainline kernels. So I looked at KernelMainlineBuilds wiki article to get the mainline kernels and try to install them. However, since my kernel panics before during boot, I am trying to install the kernel packages from a LiveCD. I figured I can do this by adding the --root=/media/*drive-uuid* option to dpkg when installing the files. Unfortunately05:20
jk-cactuar: just boot to the live cd05:21
cactuarHi jk-, I'm in the LiveCD right now.05:22
jk-mount your hdd: mkdir /mnt/foo; mount /dev/sdXX /mnt/foo05:22
jk-then chroot: chroot /mnt/foo /bin/bash05:23
cactuarOk05:23
cactuarLet me try that.05:23
jk-then you're "in" the installed distro05:23
jk-probably want to mount -t proc none /proc; mount -t sysfs none /sys05:23
jk-then apt-get or whatever05:24
cactuarAll right, I'll let you know if it works!05:25
cactuarThanks!05:25
jk-cool :)05:25
jk-no problem05:25
cactuarjk-: It didn't work ;.;05:30
cactuarRight before it terminates, I get "update-initramfs: Generating /boot/initrd.img-2.6.27-999-generic"05:31
cactuarThen a bunch of "/usr/sbin/mkinitramfs:233: cannot create /dev/null: Permission denied"05:31
cactuarAnd then "WARNING: no ldd around - install libc-bin"05:33
cactuarand then finally "update-initramfs: failed for /boot/initrd.img-2.6.37-999-generic"05:33
cactuarThough, I'm pretty sure I have libc-bin.05:33
cactuarYeah05:33
jk-cactuar: ok, from a livecd terminal (not the chroot):05:54
jk-mount /dev /mnt/foo/dev -o bind05:54
cactuarok05:55
cactuarAH!05:55
cactuarNow try again?05:56
cactuarfrom chroot?05:56
jk-yeah05:56
cactuarjk-: Ah! It finished! Thank you so much!06:06
jk-cactuar: woot :)06:06
=== smb` is now known as smb
bullgard'~$ modinfo ipw2200; filename:  /lib/modules/2.6.35-23-generic/kernel/drivers/net/wireless/ipw2x00/ipw2200.ko'. How can I establish that its associated source code file is /usr/src/linux-source-2.6.35/drivers/net/wireless/ipw2x00/ipw2200.c ?09:03
=== diwic is now known as diwic_afk
lagWould someone be kind enough to explain a "Section mismatch" please?11:31
lagI'm assuming a definition has been placed into .text when it should have been put in .data11:31
lagHow to solve?11:32
ikepanhclag: last time I have some markup to solve the warning...11:37
lagikepanhc: Hey Ike11:37
lagikepanhc: Can you explain a little more?11:37
ikepanhclag: searching the markup now11:38
lagikepanhc: Great, thanks11:38
ikepanhclag: at include/linux/init.h line 7811:39
ikepanhcah, I think maybe you shall read from head of the file11:40
lagAh, I see11:40
lagSo something must be declared as __init, or such11:40
lagikepanhc: Thanks11:41
ikepanhclag: last time I ask a __devinit function to read global variable in wrong section..11:42
ikepanhcand have the warning..11:42
lagThanks Ike, I'll look into it11:44
ikepanhc:)11:45
lagI've managed to get rid of the mismatch error, but now I get:11:50
lag"arch/arm/mach-ux500/built-in.o:(.data+0x3fec): undefined reference to `sensors1p_device'"11:50
lagHere's a grep report:11:50
lagarch/arm/mach-ux500/board-mop500-regulators.c:extern struct platform_device sensors1p_device;11:50
lagarch/arm/mach-ux500/board-mop500-regulators.c:.dev = &sensors1p_device.dev,11:50
lagarch/arm/mach-ux500/board-mop500-regulators.c:.dev = &sensors1p_device.dev,11:50
lagarch/arm/mach-ux500/board-mop500.c:struct platform_device sensors1p_device = {11:50
lagarch/arm/mach-ux500/board-mop500.c:&sensors1p_device,11:50
* lag is confused11:50
lagAnd it's only complained about at the linking stage?11:51
lagAh ...11:52
lagBingo!11:52
lagGrrrrrrr, some result12:01
lagA little help: http://paste.ubuntu.com/544393/12:01
ikepanhclag: find the solution?12:01
apwlag, are they in different modules?  as in different .ko's ?12:02
lagikepanhc: Check out my pastebin - I'm sure I'm missing something obvious 12:02
lagapw: No, everything is built-in12:02
lagIt's arch/arm/mach-<blar> code12:02
apwlag, are they in the same built-in.o12:03
lagapw: Yes12:04
ikepanhcahh, undefined reference..12:04
lagapw: Does the order of the Makefile make a difference? 12:04
apwlag, don't believe so no12:04
apwcan you paste the bit where the .o's are added for those two files in the makefile12:04
apwlag, ie grep board-mop500 Makefile12:06
lagI just did that, hence why I asked12:06
lagboard-mop500-regulators.o is before board-mop500.o in the Makefile12:07
apwlag where is the output ?12:07
apwi want to _see_ it12:07
lag:)12:08
lagCFLAGS_board-mop500.o += -DCONFIG_CRYPTO_DEV_UX500_HASH12:08
lagCFLAGS_board-mop500.o += -DAB8500_BAT_INTERNAL_NTC12:08
lag   board-mop500-regulators.o12:08
lagobj-$(CONFIG_MACH_U8500_MOP)+= board-mop500.o board-mop500-msp.o12:08
lagobj-$(CONFIG_MACH_U8500_PDP)+= board-mop500.o board-mop500-msp.o # until pdp-specific files are there12:08
lagobj-$(CONFIG_MOP500_SDI)+= board-mop500-sdi.o12:08
lagobj-$(CONFIG_MACH_U8500_MOP)+= board-mop500-mcde.o12:08
lagobj-$(CONFIG_MOP500_NUIB)+= board-mop500-nuib.o12:08
apwlag, what is the obj-xx prefix on the board-mop500-regulators.o stanza12:09
lagapw: Your guess is as good as mine12:10
apwlag can you pastebin the whole Makefile then12:10
lagOh I see12:10
lagapw: After -DAB8500_BAT_INTERNAL_NTC, there is a \12:11
lagSo "CFLAGS_board-mop500.o += -DAB8500_BAT_INTERNAL_NTC board-mop500-regulators.o"12:11
* apw is supprised that that is valid12:11
lagapw: Yes, looks add12:11
lagodd*12:11
lagYou still want the Makefile pastebin'ed?12:12
apwdm-raid45-objs := dm-raid4-5.o dm-memcache.o dm-region-hash.o dm-message.o12:12
apwyou normally do things like that to make an obj out of more than one obj12:12
ikepanhcin my point of view, I will guess someone cut few words from Makefile12:12
apwyeah it feels like something is not right with the makefile12:12
apwlag, its not clear that they are not meaning to do the same as is done for b-mop500-msp12:13
apwie everywhere board-mop500.o is mentioned board-mop500-regulators.o should follow it12:13
lagAh, hold on12:14
laggrep didn't pick up all lines12:14
lagMy bad12:14
lag..12:14
apwpastebin the whole thing then12:14
lagobj-$(CONFIG_REGULATOR)+= regulator-u8500.o \12:14
lag   board-mop500-regulators.o12:14
lagThat's better :)12:14
apwso is CONFIG_REGULATOR defined in your config12:15
lagYep12:15
lagOr I wouldn't get the error12:15
lagDefinition is in board-mop500.c12:15
apwlag, what is it set to ?12:16
lagI believe the error is in board-mop500-regulators.c12:16
* lag checks12:16
apwCONFIG_MACH_U8500_MOP and that is it =y or =m12:16
lagCONFIG_REGULATOR=y12:16
apwsame for CONFIG_REGULATOR12:16
apwand this one CONFIG_MACH_U8500_MOP12:17
apwfinally is there any bracketing of the struct platform_device sensors1p_device = {12:18
apwin board-mop500.c ...12:18
apwie any config round it12:18
* lag checks12:18
* apw pokes lag12:18
lagSorry, doing 100 things at once12:19
apwlag, but i am most important12:19
lagapw: You like to think that :)12:19
apwheh i know if i get bored i will go swimming and not wait for the answer12:19
lag*_MOP doesn't look set12:20
lagLet me check my *.mak12:20
* ikepanhc heads to home, see you all later12:21
lagSee ya Ike12:21
apwlag, its not obvious how _MOP can't be set else it wouldn't error would it ?12:22
lagapw: Exactly12:23
lagapw: And I swear I've checked it before (one of the first things I did)12:23
apwi would put errors in each of the files in turn and confirm they are built12:24
* lag is investigating 12:24
lagapw: Yarp 12:24
apwif that yeilds they are both compiled, i would then rename one set to something else12:24
apwand see which errors out12:24
apw#define sensors1p_device IN_FILE_ONE12:24
apwstylee12:24
lagapw: Where did you get the CONFIG_MACH_U8500_MOP option from?12:25
lagOh, the Makefile12:25
apw<lag> obj-$(CONFIG_MACH_U8500_MOP)  += board-mop500.o board-mop500-msp.o12:26
apwthose files are not made lest it is set12:26
lagapw: It's in the defconfig I'm using: arch/arm/configs/mop500_defconfig:CONFIG_MACH_U8500_MOP=y12:27
lagapw: I wonder why it's not in build/.config12:27
apwlag, not there at all ?12:28
apwlag, for it not to be there at all, the Kconfigs which are active in your configuration do not define it at all12:28
lagapw: arch/arm/mach-ux500/Kconfig:config MACH_U8500_MOP12:30
apwlag, is that file included, and is that entry conditional12:31
apwpaste the config stanza12:31
lagIt's not conditional 12:33
lagapw: I'm trying a different defconfig12:35
lagapw: Both should work, hopefully this one works better12:36
apwif it works it implies there is a dependancy between two config options which is not represented correctly12:36
lagapw: It does12:36
apwlag, it works or it imples ...12:38
lagapw: Implies 12:38
lagapw: Still building (which is a good sign at least)12:39
apwhrm12:39
lag:o12:48
lagCompiled 12:48
=== fddfoo is now known as fdd
=== diwic_afk is now known as diwic
* apw_ waits on -rc6 builds13:52
ckingapw_, just crack that whip harder13:54
apw_heh yeah ...13:55
tgardnerapw: what are you building on?13:55
* smb remarks that apw is whipping from the gym13:55
apw_tangerine ... though they may have finishednow13:56
tgardnerapw: yeah, load average is 0.3513:56
apw_cool13:56
jjohansensmb: have you looked at Bug #68283113:56
ubot2Launchpad bug 682831 in plymouth (Ubuntu) (and 1 other project) "lost console output early in boot (affects: 1) (heat: 10)" [Medium,Fix released] https://launchpad.net/bugs/68283113:56
smbjjohansen, no13:56
* tgardner thought jjohansen is on vacation13:57
jjohansensmb: okay, me neither13:57
smbjjohansen, But I can look a little today13:57
smbThough from tomorow /me is on holidays :-P13:57
jjohansentgardner, soon I am just trying to finish up with the review of jamie's stuff I promised him13:57
* smb moans14:00
smbSo jjohansen is still doing yesterday even as it is well today now14:00
jjohansensmb: I wouldn't call it critical, I can look at it next week, it is blocking smoser from shipping lucid amis using pv-grub which is nice to have, but is really a new feature for lucid14:00
* apw_ wonders which day jjohansen is on14:00
smbjjohansen, At least I can give it a first stab 14:01
smbSaw some of his posts with half a mind but did not really look close14:01
jjohansenapw: well wednesday's work but I uhm slept last night :)14:01
apw_jjohansen clock is mighty eratic, i suspect we need to get him a radio locked clock for xmas14:02
jjohansenhehe14:02
smbHe is the living example of the relativity theory. His time is heavily relative14:03
jjohansenhehe, I actually run through cycles, there are times when I just sleep in nap segments of an hour or two and others when I sleep long nights14:04
apw_that would kill me i recon14:05
tgardnersmb, creating  a /boot allowed me to RAID0 everything else.14:08
smbtgardner, I would have been a bit surprised if not. But good to hear you succeeded. Without reinstalling?14:09
jjohansenso smb my git bisect problems from the other night where from a corrupt fs14:10
tgardnersmb, oh, it was already trashed, so I needed to reinstall14:10
bullgard'~$ modinfo ipw2200; filename: /lib/modules/2.6.35-23-generic/kernel/drivers/net/wireless/ipw2x00/ipw2200.ko'. How can I establish that its associated source code file is /usr/src/linux-source-2.6.35/drivers/net/wireless/ipw2x00/ipw2200.c ?14:10
smbjjohansen, Doh! I am not sure I want to know how that could happen.14:10
jjohansensmb: I didn't chase it that far, running fsck spewed lots copious errors14:12
mjg59bullgard: There's no fundamental way to do so14:13
mjg59bullgard: But the path in /lib/modules will correspond to the directory the driver lives in in the source tree14:13
* smb does not think he understands "establish" the right way. One could copy a tree there manually, but would then need to take care of the right version14:14
apwyay they built14:15
smbAnd the linux-source deb package is a bit useless (imo) even if it would extract/install into the right directory14:15
apwsmb, i think he means how can you know the source file name from the name of the module, and there is no clear way to do that14:16
smbAh no, true. Only that the files for it are in the same relative directory14:16
apwyep ... the Makefile contains the mapping and its not trivial to script extract either14:17
smbIt sounded a bit like the target is oprofile, where you need the sources in some quite restricted target space14:17
apwsmb, ahh, yes an issue indeed.  i wonder if the debug versions of the .ko in the .ddeb would have the real source file names in, they are not stripped afterall14:18
bullgardmjg59: Thank you very much for your information and help.14:19
smbapw, Oh, right. Actually the unstripped modules might be what is actually needed14:19
apwbullgard, it is possible the debug versions of the .ko may contain more accurate filename information14:22
apwthough i have not tried to extract it14:23
jjohansensee you guys monday14:23
smbjjohansen, Next year! :)14:23
jjohansensmb: :), if there is anything I need to look at next week send me an email14:24
smbjjohansen, I hope to have all knowledge in the corresponding bug reports. And if you need to look at something it likely will find you. :)14:25
bullgardapw: Thank you.14:25
apwsmb, http://reports.qa.ubuntu.com/reports/jfo/kernel-buglist-by-team.html14:29
tgardnerapw, did 2.6.37-9.23 fix Sarvatt's problem with RO/NX and warm booting?14:36
apwtgardner, i have not heard anything back now you mention it, but he is more in west coast time isn't he?14:37
tgardnerapw, I think so. I just wanted to make sure his nick was flashing at him when he gets to work.14:38
Sarvatttrying it now14:38
apwtgardner, indeed ... and why i was repeating his name too :)14:38
tgardnerapw, don't forget to push your -rc6 bits14:40
apwtgardner, already pushed14:40
tgardnerack14:40
apwtgardner, you were keen that master-next was buildable so i was holding till the builds were done14:41
tgardnerapw, you gonna upload soon?14:42
apwtgardner, was thinking i might today or early tommorrow, sooner if there is nothing pending14:42
apwtgardner, you got something in the pipe i should be waiting for ?14:43
tgardnerapw, not really, but I was thinking the ext4 corruption fix might be important.14:43
tgardnerthough it does appear to be unique to certain mount options14:44
apwtgardner, yeah that does sound like a pile of snakes waiting to bite one on the ass ... i'll aim for sooner rather than later given that14:45
* apw tries booting it first :)14:45
tgardnerwhat a novel idea14:45
apwtgardner, you know me ... always pushing the envelope14:46
apwSarvatt, this clickpad thing ... do you know why its going to conflict (according to hendryk) ?  is that cause it will be upstream, or cause its all changing14:47
Sarvatttgardner: grub still hangs with the grub loading message after rebooting in 2.6.37-9-23-generic, the ftrace warnings are gone though14:47
tgardnerSarvatt, well, thats something I guess.14:48
apwSarvatt, thats sucky ... but yes a step, now its hard to diagnose :(14:49
Sarvattoh boy, reflashed the bios and lost the NX stuff and no joy fixing it14:54
Sarvattapw: no I don't know what he's referring to, don't see anything here http://git.kernel.org/?p=linux/kernel/git/dtor/input.git;a=shortlog;h=refs/heads/next15:06
apwSarvatt, i think i'll ignore that too then as we will be going therre soon15:08
tgardnerapw, should we drop ubuntu/rtl8192se in favor of drivers/staging/rtl8192e ?15:59
apwtgardner, aledgedly FOOe and FOOse are different and need different drivers16:00
tgardnerapw, same PCI IDs are in both drivers16:01
apwtgardner, will check into it16:02
=== diwic is now known as diwic_afk
=== diwic_afk is now known as diwic
Sarvattsmb: you had an AOA150 didn't you? do you have natty on it?16:50
smbSarvatt, An AOA110 but hell no. I want something stable. :-P16:51
Sarvattah same thing with an SSD, and darn, need to find someone else with one to see if https://bugs.launchpad.net/ubuntu/+source/linux/+bug/686705 affects them too16:52
ubot2Launchpad bug 686705 in linux (Ubuntu) "System hangs at GRUB loading screen every warm boot since 2.6.37-3.11 seemingly due to nx-emu patch (affects: 1) (heat: 441)" [Undecided,New]16:52
smbSarvatt, I normally do external installations on sd cards, but I run out of them. 16:54
Sarvattmaxb: running natty on your aspire one? :)16:55
Sarvattsmb: no worries, I think vish has the same machine and is running natty, will ask him to test16:56
maxbSarvatt: no.... but it's probably about the right time in the release cycle to give it a try16:56
maxbbefore I go home for christmas and am away from decent internet for a while :-)16:56
arasconklin, hey!16:57
sconklinara: hey!16:58
maxbHmm16:59
maxbNeed to get 747MB of archives16:59
* maxb hugs office internet connection, downloading at 1,890kB/s17:00
Sarvattmaxb: can't say I'd recommend it unless you want some pain :)17:00
maxbPart of the function of my netbook is to be a machine I can break :-)17:01
Sarvattsame as mine then :) when you get it all done would you mind seeing if you hang at grub after booting the natty -generic kernel and then rebooting?17:02
maxbSo, boot natty kernel once, restart without a power off, see what happens?17:04
apwmaxb, expect unity to be utter crap right now, the bug list is immense, including not starting, not displaying self, not showing menus unless you right click on the desktop first, screen saver not working ... etc17:13
apwall of the above at random and in combination of course17:14
maxbapparently even the dist-upgrader is broken right now :-)17:14
apwmaxb, _yes_ ship it17:14
jMCgmeh.. still no UTRACE support :-/17:17
jMCgMaybe I should start something like a feature request..17:17
maxbSarvatt: "About 4 hours remaining" .... ping me tomorrow :-)17:28
=== BenC__ is now known as BenC
=== charlie-tca__ is now known as charlie-tca
apwjMCg, that tends to help18:02
apwjMCg, if a feature is not upstream then we won't get it without someone needing it and proposing how we might get it and maintain it18:08
vanhooftgardner: heya18:15
tgardnervanhoof, dude18:16
vanhooftgardner: this new linux-firmware that's uploaded, did you test out the 6150 or 6250 card on maverick, or just natty?18:16
vanhooftgardner: trying to get some maverick testing done today18:16
tgardnerlucid, maverick, and natty18:16
tgardnermaverick and lucid might need compat-wireless. i can't keep 'em all straight18:17
vanhoofyeah thats what im guessing18:17
vanhoofcool, ill try it out18:17
* vanhoof was hoping for a 'just works' ;)18:18
tgardnervanhoof, try the stock maverick kernel first. lucid will definitely need compat-wireless18:18
vanhooftgardner: will do, thanks!18:19
=== smoser is now known as smoser`
=== smoser` is now known as smoser
JFotgardner, do you know who looks after fglrx-installer (Ubuntu)?20:23
tgardnerJFo, not offhand. who was the last changelog entry?20:23
JFono idea. I'm looking at a bug from skaet 20:24
JFoand I confess I have not heard of the package before20:24
tgardnerJFo, which release?20:24
JFonatty is what is targeted20:25
JFoit is bug 56643720:25
ubot2Launchpad bug 566437 in fglrx-installer (Ubuntu Natty) (and 1 other project) "package fglrx 2:8.723.1-0ubuntu2 failed to REMOVE: error exit status 2 - dpkg-divert: mismatch on package - while removing the package (affects: 112) (dups: 27) (heat: 458)" [High,Triaged] https://launchpad.net/bugs/56643720:25
JFopackage is owned by Ubuntu Developers, so that is no help20:26
JFooh20:26
tgardnerJFo, gimme a bit...20:26
JFohmmm20:26
JFotgardner, ok20:26
charlie-tcathat's the ati video package, isn't it?20:26
tgardnerJFo, its probably alberto milone, but I'm checking for sure20:26
JFotgardner, k20:27
tgardnercharlie-tca, yep20:27
tgardnerJFo, ah, Sarvatt is your man.20:35
tgardnerhe touched it last20:35
JFoah hah!20:46
* JFo assigns blame ;-)20:46
maxbSarvatt: For me, grub started to hang only after powering off for a while21:36

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!