/srv/irclogs.ubuntu.com/2005/12/03/#ubuntu-kernel.txt

BenCmjg59: I see one thing odd01:54
BenClook at the diff to bitblit.c01:55
BenCnm01:55
BenCI didn't see where it set err = 101:55
BenCI can't see anything that jumps out01:59
BenCpretty straight forward01:59
=== BenC [n=bcollins@richmond-209-163-125-167.dynamic-dialup.coretel.net] has joined #ubuntu-kernel
=== BenC [n=bcollins@richmond-209-163-125-167.dynamic-dialup.coretel.net] has joined #ubuntu-kernel
mjg59BenC: Reverting it seems to make usplash work03:07
mjg59So I'm wondering whether it's changing order of initialisation in some subtle way03:07
BenCwhat's the URL again?03:24
BenCnm, got it03:25
BenCno, I don't, can you past it again?03:26
mjg59http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff_plain;h=c465e05a03209651078b95686158648fd7ed84c5;hp=e764a20196f4e1b497a42fdc6e9d254e7ec290f203:26
mjg59I'd prefer it if someone else can verify that it changes the behaviour03:26
mjg59I'm not entirely convinced that I haven't buggered up my test system along the way03:27
=== ispiked [n=ispiked@unaffiliated/ispiked] has joined #ubuntu-kernel
BenCI don't have anything other than the machine that was already broken in breezy to test on03:33
BenCnothing changes when softcursor.c got moved (code wise)03:33
mjg59Hmm.03:33
mjg59Yeah. So now I'm confused.03:33
mjg59Hmm.03:34
mjg59The usplash script tries to load softcursor.ko03:34
mjg59Which is obviously going to fail03:34
mjg59I wonder if that's the problem03:34
BenChow's it going to fail?03:35
BenCdid you make the change I submitted in the bug report to fix the loading?03:35
mjg59It gets installed in a different location, I think03:35
mjg59Uhm. Which bug report?03:36
mjg59(Sorry, I've lost track of which ones are on which issues)03:36
BenCI filed a bug report on usplash about the module loading order and location03:36
BenCsoftcursor needs to be loaded first, and needs to be changed to video/console/softcursor.ko03:37
mjg59I'm actually shifting to modprobe03:37
mjg59Just let me get a fresh kernel image and I'll check03:38
BenCthat'll work, modprobe vga16fb should do the right thing03:40
mjg59Ok, modprobe doesn't fix it03:54
mjg59Hm. So now I wonder if I was finding the wrong breakage with my bisection03:55
mjg59Grr03:55
=== TheMuso [n=luke@dsl-202-173-132-131.nsw.westnet.com.au] has joined #ubuntu-kernel
mjg59I may be entirely wrong about that patch, so04:04
mjg59Oh ha ha fucking ha04:27
mjg59Making sure that fbcon is loaded before vga16fb makes it work04:27
mjg59So not a kernel issue at all04:28
mjg59PAIN04:28
infinity... Isn't that backwards?04:40
=== infinity always assumed fbcon depended on having an fb loaded, not the other way around.
infinityOh well, whatever. :)04:40
infinityIf you've figured out the magical load order, can you eiher push me a patch, or just make an upload?04:40
mjg59I've uploaded04:42
infinityYay.04:42
mjg590.1-23 should fix04:43
infinityBTW, do we have any spare slots in the pallette?04:44
mjg59No04:44
mjg59It's all taken up with brown04:45
infinityToo... Much... Brown.04:45
mjg59So, can we have the 640x400 patch? Can we can we can we?04:45
infinityI need to push aside a shade of brown for some sort of yellow/warn colour, I think.04:45
mjg59Ngh.04:45
mjg59I'd also like some artwork that used flat colours rather than trying to anti-alias with a tiny number of them04:46
mjg59I think that would be a feature04:46
infinityYeah.  Good luck convincing someone to do that.04:46
mjg59How about we kill Jeff and get someone else to spec the artwork next time?04:46
mjg59Also:04:47
mjg59BenC: Have you dropped the Toshiba acpi key magic patch?04:47
jbaileymjg59: Eh?04:48
jbaileymjg59: Of what do you accuse me? =)04:48
infinityI'm sure I could just scribble "OOBOONTOO" in a red pen in The GIMP and use that.04:48
mjg59jbailey: Other Jeff04:48
jbaileymjg59: Oh good.04:48
mjg59infinity: I'll upload it tomorrow04:48
=== infinity goes into his little corner to give mjg59 a working lrm for his ath.
mjg59I have an ath?04:53
mjg59Oh, yes04:53
infinitySomeone should send me an atheros card for Christmas, so I can actually test this driver.04:55
infinityThe only LRM stuff I can test is fglrx and nvidia.04:55
infinity(Which is the largest amount of work in the package anyway, so I guess that works)04:56
=== TheMuso [n=luke@dsl-202-173-132-131.nsw.westnet.com.au] has joined #ubuntu-kernel
mjg59BenC: Can you add the diff from http://www.codon.org.uk/~mjg59/tmp/fb-sysfs.diff ?04:59
mjg59Haha.05:09
mjg59So.05:09
mjg59The reason for powernow not working on my Turion may have been the fact that I'd disabled it in the BIOS05:09
mjg59Oh deary me.05:09
infinity*cough*05:10
mjg59Which means that the only real thing missing from this machine is the Broadcom wireless05:11
mjg59Looks like it just got support for scanning. Hurrah.05:13
mjg59Association can't be too far off05:13
BenCmjg59: patch will be in -5.705:14
mjg59BenC: Rock, thanks05:15
=== ispiked [n=ispiked@unaffiliated/ispiked] has joined #ubuntu-kernel
BenCmjg59: what was the url to the softmac stuff?06:25
mjg59softmac.sipsolutions.net06:26
BenCthanks06:27
fabbionehey guys06:29
fabbioneBenC: sparc fails with oldconfig at kernel-headers otherwise we look good06:30
fabbionei think the defconfig in the tree is old06:30
fabbionedpkg-deb: building package `linux-headers-2.6.15-4' in `../linux-headers-2.6.15-4_2.6.15-4.6_sparc.deb'.06:30
BenCI just did a str8 build and it did fine06:30
fabbionethat's weird06:31
fabbionehow do you call the build?06:31
BenCnot with sparc3206:31
BenCjust directly06:31
infinityOkay, dude, we can be coworkers still ,but we can't be friends anymore.06:31
=== infinity mutters about people typing "straight" as "str8"...
BenCinfinity: I can say str8 if I want to :P06:31
fabbioneif i use sparc64 in debian/config/archmap and dpkg-buildpackage it farts06:31
BenCI'm used to typing that when I play online poker06:31
BenChabit06:32
fabbioneunknown Debian architecture sparc64, you must specify GNU system type, too at /u06:32
BenCfabbione: guess I need to test with "sparc32 dpkg-buildpackage ..."06:32
BenCsee if I can figure out what is going on06:32
fabbionewell what i did was:06:32
fabbioneuse sparc in archmap06:32
fabbioneand build using sparc64 dpkg-build...06:32
BenCthat's what I have now06:32
fabbionethat builds all the kernel except the old config for -header-06:33
fabbionei need to test out of git06:33
BenCI didn't call sparc64, but it should be the same06:33
fabbionethis was 4.606:33
BenCI need to push, have a big changeset06:33
BenC(ocfs2, and linus sync)06:33
fabbionedid you also pull from me?06:33
fabbioneperfect06:33
fabbioneif you can push, i can run a build06:34
infinityBenC : eta on -5.X?06:34
BenCif all goes well, tomorrow night06:34
infinityBenC : We really need the update-initramfs change for the -5 ABI bump, and I need to push you the vga16fb patch too, before it's too late.06:34
BenCprobably a good idea if I make tomorrow a "get everything ready for linux-meta switch to .15" day06:35
infinity(Need to build kernels and test on a couple of machines here first, before I decide to inflict it on the world)06:35
BenCincluding fixing k-p06:35
=== lamont [n=lamont@mix.mmjgroup.com] has joined #ubuntu-kernel
infinityYeah, LRM will be in by then.06:35
infinityUpdating it for the ABI bump is simple.06:35
BenClamont: !!06:35
infinitySo that's no big deal, once I've done the hard bits (fixing it to stop sucking and actually build)06:35
BenClamont: I have -5.7 building on hppa and ia64 now :)06:35
BenCinfinty: good deal06:36
fabbionecoll06:36
fabbionecool06:36
BenCI have 5 builds going atm06:36
infinityBenC : mdz wants me to give you a quick tour of LRM too, so you're not stuck blocking on me when the world crashes around my ears with other tasks.06:36
BenConly reason I don't have amd64 going is I haven't pushed for concordia to pull from yet06:36
infinityBenC : But we'll save that until after I've updated all the drivers.06:37
BenCinfinity: unless there's more than what I already encountered doing a few ABI bumps in breezy, then I'm good06:37
infinityBenC : Axcept for the part where I'm changing that mechanism, no.  If you never have to update a driver, LRM should be simple.06:38
BenCfabbione: the e3k is my new best friend...I can pump out a quick build in 15 minutes just to see if it compiles ok :)06:38
infinityIf you find yourself having to update one (especially nvidia or fglrx), you'll hate yourself.  So, it's best not to bother. :)06:38
fabbioneBenC: ehehe ok06:38
fabbionei didn't want to keep you awake06:38
BenCinfinity: I'll try to stay clear of actually doing work on lrm :)06:38
=== infinity looks in wonder at the new word he's created.
infinitys/Axcept/Except/06:39
BenCinfinity: my hillbilly accent rubbin' off on ya'll06:39
infinityBenC : Ugh, dude, I was talking like you for days after you left.06:40
fabbioneBenC: ok, i am going to stop building the kernel for now. Just let me know if i need to build it with sparc32 or sparc64. I can override that in sbuild per source pkg06:40
BenCwas surprised my wife didn't notice my slight french accent when I got back06:40
fabbionelol06:40
fabbioneMerci06:40
BenCI noticed I even spoke broken english a few times, in addition to the accent06:41
infinityIf it's not buildable with sparc32, you're not doing it right. :)  Period.06:41
infinityI sure as heck wouldn't special-case the powerpc buildds to build with linux64 for kernel builds.06:41
BenCinfinity: yeah, because we all know that it has to build on a 50mhz sun4c :)06:41
fabbioneehhehe06:41
fabbioneinfinity: i blame kernel-package06:42
fabbioneno really..06:42
infinityBenC : No, but the buildds themselves run in linux32/sparc32 for pretty valid reasons, and having the kernel builds not fit into my little automated world sucks. :)06:42
infinity(doubly-so, if I end up with a sparc buildd in the DC)06:42
BenCyeah, before I set the sparc buildd's to use sparc32 way back when, administering it was awful06:43
fabbioneinfinity: that's why lamont did implement an option in sbuild for that :)06:43
BenCfor what?06:43
fabbione$build_env_cmnd{'linux-source-2.6.15'} = '/usr/bin/sparc64';06:43
fabbione <- in .sbuildrc06:43
BenCah06:43
infinityOh, that's just vile.06:43
fabbioneso you can override the environment per source pkg06:43
fabbioneinfinity: no. it works06:44
fabbioneBenC: the buildd defaults to sparc3206:44
fabbioneand you can make exceptions06:44
infinityAnyhow, that wasn't lam.06:44
infinitylamont, even.06:44
fabbioneinfinity: you?06:44
fabbionebecause the default environment was alredy there06:45
fabbionethe override per source was lamont06:45
fabbionei am pretty sure about that06:45
fabbionecheck the baz changelog06:45
infinityOh, wait.  Yes, that was him.06:45
infinityIt kinda mucks with how we do the global  linux32 stuff though.  Yay.06:46
infinityTo be fair, I hacked the latter after he did the former.06:46
infinitySpecifically to work around the bogus requirement to have sparc32/linux32 in the chroots.06:46
infinityWhich is painfully dumb.06:46
fabbioneinfinity: after we will get the sparc buildd at the CD06:47
fabbionewe will see how it works06:47
fabbionei really don't care..06:47
fabbionejust don't break the stuff for older releases06:47
infinityI don't plan on breaking anything. :)06:48
infinityIf you have whacky sbuild configs, though, you'll have to push those to me.06:48
fabbioneinfinity: sure06:48
fabbionethat's no problem06:48
fabbionethe line you saw is basically the only custom in that config06:48
fabbionewell.. for hoary/breezy/dapper atm06:48
fabbionewe didn't build warty06:48
infinityAhh, so you haven't done these overrides for older linux source stuff?06:48
fabbioneso that's no issue06:48
fabbioneyes. only the kernels06:49
fabbione$build_env_cmnd{'linux-source-2.6.10'} = '/usr/bin/sparc64';06:49
fabbione$build_env_cmnd{'linux-source-2.6.12'} = '/usr/bin/sparc64';06:49
fabbione$build_env_cmnd{'linux-source-2.6.15'} = '/usr/bin/sparc64';06:49
fabbionethat's all the custom love i have there06:49
infinityAhh, three lines then, not one. :)06:49
infinityWell, let's see about fixing kernel-package.06:49
fabbione+ the default to use sparc3206:49
infinityIf it DTRT on powerpc, surely we can make it do the same for sparc.06:49
fabbioneoh yeah06:49
fabbionespecially because we don't care about 32bit kernels06:49
fabbionethat's where all the mess comes from06:50
infinityWell, yes, 32-bit targets on sparc are laughable, but it's technically a bi-arch system with a bi-arch toolchain, and we should treat it as such.06:50
infinityEven if the kernels will all be 64-bit.06:50
infinitySome day, someone may resurrect sun4 support in 2.6, I'll boot it once, remember how HIDEOUSLY SLOW they were, and never again. :)06:51
fabbioneerrr06:54
=== BenC remebers the 28 hour builds for libc6 on his Sparc LX
fabbionewe already killed support for sparc32 in glibc06:54
fabbionethere is really no point in sparc3206:54
fabbioneat all06:54
BenClibc is v9 optimized?06:55
fabbioneif somebody will resurrect it... well they are welcome to patch everything back06:55
fabbioneBenC: you need ask jbailey but i think they are06:55
fabbione      - debian/sysdeps/sparc.mk: Set libc_configure_build=sparcv9-linux06:56
fabbione        set sparc64_configure_build=sparc64-linux06:56
fabbione    * Drop support for Neanderthal Sparc systems06:56
fabbione      - debian/sysdeps/sparc.mk: Drop sparcv9 package06:56
fabbione        Default to nptl for main build and sparcv9b06:56
fabbione      - debian/control.in/opt: Drop sparcv9 package06:56
fabbione      - debian/control: Regenerate06:56
fabbioneso i think they are optimized06:56
infinityIs that true in Debian too?06:57
fabbioneno06:57
fabbioneonly ubuntu06:57
infinityKay, that makes sense.06:57
fabbionesparc is our pet arch.. mmk?06:57
fabbioneno really we can do stuff in here that Debian can't06:57
fabbionelet's do i06:57
fabbioneit06:57
infinity<nod>06:58
infinityI need to get a reasonably speedy Ultra in my house.06:58
fabbionei need to get elmo and znarl to look at buildd prices and get them to the datacenter06:59
fabbioneso i can finally use my sparc for debugging and testing06:59
infinityWe should be able to get a set of craptastic Sunfire V240s for next to nothing.07:01
infinityShould fit well with the buildds we already have.07:01
fabbionei think elmo was looking in something like that07:01
infinityThe 240 is a lot less crap than the 100/120s I had to work with at my last job, at least.07:03
=== chmj [n=chmj@wbs-146-132-171.telkomadsl.co.za] has joined #ubuntu-kernel
=== CataEnry [n=Enrico@host204-30.pool8248.interbusiness.it] has joined #ubuntu-kernel
CataEnryhi all01:11
=== jbailey [n=jbailey@modemcable139.249-203-24.mc.videotron.ca] has joined #ubuntu-kernel
=== zul [n=chuck@ubuntu/member/zul] has joined #ubuntu-kernel
zulheylo02:23
fabbionehey zul02:24
jbaileyHeya Chuck!02:29
jbaileyzul: Ottawa's even pretty in snow, it's kinda sick.02:29
zulmeh...its ok :)02:54
zulhey guys how was your weekend?02:55
=== mgalvin [n=mgalvin@ubuntu/member/mgalvin] has joined #ubuntu-kernel
mjg59infinity: Can't remember if I asked this last night - what's the situation with the 640x400 thingy?03:37
zulhmmm...my touchpad is foobared03:48
mjg59On 2.6.15?03:50
mjg59Known03:50
zulyeah i know03:53
mjg59I blame Scott03:54
zulheh.03:59
mjg59Or Marco, or GregKH04:01
=== mkrufky [n=mk@68.160.103.77] has joined #ubuntu-kernel
jbaileyzul: <Keybuk> http://people.ubuntu.com/~scott/packages/new-udev04:06
jbaileyFrom #ubuntu-boot not too long ago.04:06
=== BenC [n=bcollins@richmond-209-163-125-167.dynamic-dialup.coretel.net] has joined #ubuntu-kernel
zulhmmm..04:43
makxzul: link to your weblog?04:48
makxdo you have modular mousedev04:48
makx?04:48
zulmakx: i dont have a weblog04:58
makxah ok wasn't you that said to respond to davej, zul?04:59
zuldont have time to05:06
mkrufkyzul: you closed bug 15395, and ive been busy -- didnt have enough time to write you back05:10
mkrufkyanyhow, yes... i confirm that not only is it not an ubuntu bug (it was an upstream bug) it is also fixed now05:10
mkrufkythe problem was due to missing dvb headers for users trying to install v4l modules from cvs -- it is a moot point now, as we have officially merged dvb and v4l cvs repositories05:11
mkrufkyside note: we are also merging ivtv and pvrusb2 and usbvision drivers into v4l ... so such problems will be nothing but history soon05:12
zulok cool..05:17
mkrufkyya, sorry for the 2-week delay in my response ;-)05:22
zulmeh...its ok i guess ;005:30
=== netjoined: irc.freenode.net -> brown.freenode.net
=== ispiked [n=ispiked@unaffiliated/ispiked] has joined #ubuntu-kernel
=== TheMuso [n=luke@dsl-202-173-132-131.nsw.westnet.com.au] has joined #ubuntu-kernel
=== BenC [n=bcollins@debian/developer/bcollins] has joined #ubuntu-kernel
=== BenC_ [n=bcollins@richmond-209-163-125-167.dynamic-dialup.coretel.net] has joined #ubuntu-kernel
BenC_lamont: ping06:37
=== JaneW [n=JaneW@wbs-146-174-135.telkomadsl.co.za] has joined #ubuntu-kernel
zulhmmm...somtimes i can be such a bastard07:30
=== lamont [n=lamont@mib.fc.hp.com] has joined #ubuntu-kernel
BenClamont: hey, when can ia64 and/or hppa use initramfs?08:13
BenClamont: btw, I have 2.6.15-5.7 running on my i2k08:13
=== janimo [n=jani@Home03207.cluj.astral.ro] has joined #ubuntu-kernel
makxBenC: hppa needs an klibc patch.08:19
makxhaven't uploaded it yet to Debian due to troubles with missing asm symlink in linux-headers08:19
BenCwhat about ia64?08:19
makx" inflate code misreads magic number" according to jbailey.08:20
jbaileyklibc has the patch in Ubuntu.08:21
makxaah you added it :)08:22
makxneed to sync with that badly.08:22
jbaileyI need to make lkh not suck in Debian badly.08:23
makxhehe08:23
BenCjbailey: so I can safely switch ia64 to use initramfs, or has it not been tested?08:24
BenCif it hasn't been tested, I can surely do it, and make the change to kernel-package with this update-initramfs change I'm about to do08:28
jbaileyBenC: Well, it doesn't boot. =)08:30
jbaileyBenC: My initial suspicion is that the READ_BYTE or some such macro isn't right.08:30
jbaileyI had LaMont add some printk's to give us the value of those bytes, and they were complete whacked.08:31
jbaileyIt's probably easy enough to troubleshoot, but my ia64 isn't local.08:31
BenCI might just wait till the weekend and take a look at it then08:35
BenCjbailey: -5.7 will be using update-initramfs, fyi08:36
jbaileyBenC: Nice, thanks!08:53
fabbionehave a nice evening guys09:01
fabbionecya tomorrow09:01
fabbioneor very late today :)09:01
crimsun'night, fabbione 09:01
fabbionenight09:02
janimoBenC, I don't know if LaptopTesting issues have different priority but here's https://bugzilla.ubuntu.com/show_bug.cgi?id=2008509:13
janimoit makes a test laptop boot09:13
BenCis that the acpi return value ignore patch?09:14
janimoyes09:14
BenCif so, it's in -5.709:14
janimoit is confirmed upstream09:14
janimothanks!09:14
CataEnrybye all09:14
BenCsure thing09:14
BenCjbailey: ping09:18
jbaileyBenC: pong09:19
BenCjbailey: update-initramfs has a bug09:19
BenCthe -d option requires an argument, yet it never sets version to the argument, so it always fails09:20
jbaileyBenC: It's infinity's package now, but if it's an obvious fix, I'd say just upload it.09:20
BenCok09:20
jbaileyYou're the most important consumer of update-initramfs, so bend it to your needs.09:20
=== BenC adds another package to his list of "get the damn thing ready for linux-meta"
zulhehe09:21
BenCnm, I was just using it wrong09:24
BenCthe wiki page for the spec said just -d, and it needs to be -d -k <ver>09:24
makxBenC: which wiki page?09:25
BenChttps://wiki.ubuntu.com/InitramfsUpdates09:25
=== janimo [n=jani@Home03207.cluj.astral.ro] has left #ubuntu-kernel []
lamontjbailey: BenC  - ia64 initramfs problem is an elilo issue, forwarded upstream09:53
jbaileylamont: Cool.  Is it just setting the load address too low?09:54
lamontnfc09:56
lamontwhich reminds me... need to see what we did for breezy elilo09:56
=== doko [n=doko@dslb-084-059-082-035.pools.arcor-ip.net] has joined #ubuntu-kernel
zulBenC: you know those unknown key pressed errors that some people are getting in their dmesg?10:05
BenCzul: yeah10:10
BenCI get it sometimes too10:10
BenClamont: I've fixed the hppa64-gcc thing (it needed to be hppa64-linux-gnu-gcc-3.4 for my breezy install), however, hppa64 is a nobuild from git right now10:11
BenCa lot of missing headers referenced for the compat layer10:11
BenCzul: do you have a fix for it?10:12
zulBenC: yeah remove the prink that is causing that, its useless10:14
BenChehe, put it in yout git so I can pull it :)10:15
zulill make a patch for you tonight10:15
jbaileyBenC: From the parisc git, or from Linus?10:23
BenCI pulled parisc git hoping to get things working10:27
BenCbut it seems to only get hppa32 working, and not hppa64 :/10:27
=== jbailey pings kyle to tell him.
BenCcool, the update-initramfs stuff works...install, upgrade/reinstall, and remove/purge10:31
BenCdid the reboot-notifier too10:31
=== jbailey ^5's BenC
=== zul disklikes the reboot-notifier
zulanyways home for me10:42
=== BenC_ [n=bcollins@richmond-209-163-125-167.dynamic-dialup.coretel.net] has joined #ubuntu-kernel
=== mkrufky [n=mk@68.160.103.77] has left #ubuntu-kernel []

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