/srv/irclogs.ubuntu.com/2005/10/03/#ubuntu-kernel.txt

mkrufkyso, deadline or no deadline........12:03
=== spayne is training to become a MOTU as well!
mkrufkyMOTU ?12:03
spayneUniverse Packager?12:03
mkrufkyoh12:03
crimsununiverse/multiverse maintainers12:03
crimsunsee #ubuntu-motu12:04
mkrufkyi'm what they call an "upstream guy"12:04
crimsunyou're more than welcome to join, mkrufky12:04
mkrufkyhmmm12:04
mkrufkyuntil that last comment i was GOING to say, "heh..  i just like to write code.... i'll leave the packaging to you guys" .... but since u put it this way12:05
mkrufkymaybe i might just have to investigate ;-)12:05
mkrufkycrimsun: i dont think we've met yet, btw12:05
crimsunwell the ideal situation is of course if the coder packages it, since (s)he can deal directly with bug reports ;-)12:05
mkrufkyi am a maintainer of upstream v4l & dvb kernel subsystems12:05
mkrufkyim not *THE* maintainer... just one of the few12:06
crimsunmkrufky, excellent. Pleased to meet you. I help out with ALSA in Ubuntu.12:06
mkrufkygreat12:06
mkrufkyi told them i would help to backport newer code to their next kernel....12:06
mkrufkybut i dont think this involves any packaging12:07
crimsunprobably not, I think lamont and benc will handle that12:07
mkrufkybut there's always new stuff to learn, and im always interested12:07
mkrufkyanyway, its about that time to go home now... nice to meet u crimsun and spayne .... 12:09
mkrufkytty guys later... good luck12:09
spaynekeep in touch :)12:09
crimsunthanks, cya 'round12:09
mkrufky:-)... i can always be found in both #v4l and #linuxtv ... (and usually here as well)12:10
mkrufkyttyl12:10
=== mkrufky [n=mk@68.160.103.77] has left #ubuntu-kernel []
spayneBenC: ping me if you are back :)12:15
lamontmjg59: care to take a guess as to which acpi patch breaks ia64?12:22
lamontas in. most invasive patch, etc.12:22
lamontjbailey: any reason not to throw out initramfs for ia64 as well as hppa?12:24
spayneBenC: ?12:36
jbaileyThrow out, like not use?12:40
jbaileyThe biggest reason is that those machines are then stuck with 2.6.12 until you get it working.12:40
jbaileySo you're only buying yourself 2 weeks or so.12:41
BenCspayne?01:10
=== mkrufky [n=mkrufky@user-12lcl1s.cable.mindspring.com] has joined #ubuntu-kernel
BenCanyone know why isapnp would be reserving memory that was reserved for a PCI device?02:07
BenCuser has an 8139 that has 0xde00-0xdeff reserveed for I/O, and isapnp grabs 0xde00-0xde03 on bootup02:08
BenCif they boot with noisapnp, the device works ok02:08
BenCotherwise, it complains about not being able to reserve the io region02:08
mkrufkythis may or may not help:02:11
mkrufkyrecently, on lkml, there was talk about pci memory allocation to the wrong memory area....02:11
mkrufkysomething to do with overwriting memory thats SUPPOSED to be reserved for the bios02:12
mkrufkyi believe the talk was before 2.6.13 got rreleased, so i dont know if its related or not02:12
mkrufkybut i remember that it was GregKH that was talking about it.... you may want to search lkml02:12
mkrufkyi know that isnt the exact situation that you're describing, but it COULD be related02:13
jbaileymjg59: Around?02:13
=== calc [n=ccheney@ip70-185-4-246.ma.dl.cox.net] has joined #ubuntu-kernel
BenCmkrufky: well the same memory area is listed for 2.6.10 (where it's working) and 2.6.12 (where pnp grabs part of it)02:28
BenCthe only difference is that in 2.6.12, isapnp beats 8139 to reserving the memory, even though /proc/ioports clearly knows that the memory region is for the PCI device02:29
BenCand lspci -v shows it too02:29
mkrufkyhmm02:29
BenC16371 if you want to look at it02:30
BenCI think isapnp is just broken in this case, unless the guys bios is setup wrong02:30
mkrufkyi'll take a look... but i think this means that the bug *i* was talking about is probabl;y unrelated02:30
mkrufky...sorry02:30
BenCyeah, sounds like it02:30
BenCno problem02:30
mkrufkyhmm... he didnt include dmesg without noisapnp02:33
mkrufkyoops... i see it02:33
BenCdoesn't seem that isapnp was disabled like i thought it would be02:35
BenC2.6.10 dmesg shows isapnp just not doing anything since there are no pnp devices02:35
BenCjust the parallel port02:35
BenCat 0x378, like it should be02:35
mkrufkyya02:35
mkrufkyhmmm... what a mess02:37
BenCmaybe I need to disable pnpacpi02:37
mkrufkythat could be something to try02:38
mkrufkyhmmm.... it's not really clear what is going on in the inline dmesg output... i assumed that was 2.6.12 without the noisapnp02:41
mkrufkyand then you said, "doesn't seem that isapnp was disabled like i thought it would be"02:42
mkrufkyi assume that you're referring to:02:42
mkrufky[4294670.791000]  isapnp: Scanning for PnP cards...02:42
mkrufky[4294671.145000]  isapnp: No Plug & Play device found02:42
BenCthat's 2.6.1002:43
mkrufkyah02:43
mkrufkyso does he show us 2.6.12 nywhere?02:43
BenCthere's a 2.6.12 inline, original, and a 2.6.12 with isapnp (attachment)02:43
BenC...02:43
mkrufkylooks like he is only showing 2.6.12 with noisapnp02:43
BenC"dmesg output whit noisapnp"02:43
BenCthat's all we need02:44
BenC2.6.10 works02:44
BenC2.6.12 with noisapnp I meant02:44
BenCjust asked him to try some other options02:44
mkrufkyok02:44
=== TheMuso [n=luke@dsl-202-173-132-131.nsw.westnet.com.au] has joined #ubuntu-kernel
fabbionemorning guys05:40
=== crimsun [n=crimsun@pdpc/supporter/silver/crimsun] has joined #ubuntu-kernel
=== JaneW [n=JaneW@wbs-146-181-185.telkomadsl.co.za] has joined #ubuntu-kernel
=== Yagisan [n=jamie@60-240-35-96.tpgi.com.au] has joined #ubuntu-kernel
=== Yagisa1 [n=jamie@60-241-38-168.tpgi.com.au] has joined #ubuntu-kernel
=== spayne [n=spayne@i-195-137-120-148.freedom2surf.net] has joined #ubuntu-kernel
spaynemornin' all09:01
spayneBenC: ping09:01
fabbionespayne: BenC is asleep at this time09:06
spaynefabbione: where abouts is he?09:08
fabbioneUS09:08
fabbioneso i guess he will be around in 6 hours + o -09:09
spaynefabbione: thanks09:09
spaynefabbione: i'm just darn excitied to get this bug squashed09:09
spayne:)09:09
infinityspayne : Which bug is that?09:17
=== chmj [n=chmj@196.36.161.235] has joined #ubuntu-kernel
spaynethe ndiswrapper one09:22
infinityOh, dang.  I was hoping it would somehow relate to USB being goofy.09:23
spayneright, well09:24
spaynethe newest version released yesterday (1.4rc1) fixes the bug as the whole USB layer has been rewrittrn to work better with the 2.6.12 kernel as well as fixing hundreds of bugs since 1.209:25
spayne1.3 was never released09:25
spayneso...09:26
spaynethe solutions arer09:26
spaynea.) Do Nothing - Will cause problems when Breezy released as many others way have this sort of problem which is fixed in 1.4rc109:26
spayneb.) Wait until 1.4 is released - This could be a week or just a few days09:27
spaynec.) Put it into the Kernel - not a good idea as it has had little testing and bugs will/may be found09:27
spayned.) Make a Seperate Universe Packaging for 1.4 - This seems a good comprimise as this could be updated before Breezy is released with 1.4 final and means those who need it can use it09:28
spayneinfinity: what do you think?09:28
infinityHrm.  How would the seperate universe package work?... dpkg-divert the USB modules from the kernel?09:33
infinityIcky, but I guess it's not technically difficult.09:33
spaynea single package which will install both the modules and the utils09:33
infinityRight, but we also have USB module in the kernel package, so you'd need to divert those elsewhere.09:33
Mithrandirthat'll be fun09:35
spaynewell, what are your suggestions?09:36
spayneinfinity: what would you do?09:41
Mithrandirdiverting the modules will work, it's just ugly.09:42
spaynebrb09:43
infinityHey, if diverting modules works for lrm-nvidia-legacy (it diverts modules from lrm), it can work for this, too. :)09:44
spaynei wouldn't know where to begin09:44
spaynei'm JUST learning how to package from Debian :-)09:44
spaynes/Debian/Ubuntu09:44
infinityI do have an interest in having the USB stuff not be buggered, but perhaps the better answer is to get it in the kernel proper and get VERY widespread testing in the next week.09:45
infinityOn my machine, USB seems to magically take out my PCI NIC, which is pretty cool.09:45
spayneinfinity: shall we see what BenC says? (or do you have the power?09:46
infinityIt'll be up to BenC as a first line approval, and mdz to swat it down after that.09:47
infinityBoth of whom ar ein the US, so wait for them to wake up and get to work. :)09:47
spaynei got a day off school :) Speech Day this evening :)09:47
infinityThis close to a release, we're obviously rather cautious, but this is a pretty nasty regression for some.09:47
spaynei know09:48
infinityI just wish I'd noticed it earlier.  I hadn't booted or upgraded that machine for a couple of months, so I have no idea when the bugs started hitting it.09:48
spaynesigh - it is very difficult09:48
spaynebut 1.4rc1 was only released yesterday which solved these bugs09:48
spaynelook at http://sourceforge.net/mailarchive/forum.php?thread_id=8321710&forum_id=3647109:49
infinityOh, wait, you were just talking about ndiswrapper-usb, not the general kernel USB layer.09:50
infinityThat may not be my bug anyway, then. :)09:50
spayneyes09:50
infinitySince ndiswrapper shouldn't even be loaded on my system.09:51
spaynelol09:51
=== spayne thought he had a mutal friend
infinityIncluding an updated ndiswrapper would be much less scary than an updated USB layer.09:51
spayneoh yeh09:51
spaynejust i think this is importnat - the Laptop Testing Scheme09:51
spayneso laptops should work better09:52
spayneUSB works fine for me :)09:52
spaynefor everything else09:52
spaynejust moving to my workstation09:54
spaynebrb09:54
=== spayne [n=spayne@i-195-137-120-148.freedom2surf.net] has joined #ubuntu-kernel
spayneback09:58
spayneinfinity: all we can done is wait10:09
=== Yagisan [n=jamie@60-241-39-173.tpgi.com.au] has joined #ubuntu-kernel
=== spayne [n=spayne@i-195-137-120-148.freedom2surf.net] has joined #ubuntu-kernel
spayneback11:29
=== jorgp2 [n=jorgp@bnet-dial-224.bartnet.net] has joined #ubuntu-kernel
=== spayne| [n=spayne@i-195-137-120-148.freedom2surf.net] has joined #ubuntu-kernel
spayne|ping anyone01:15
=== \sh [n=sh@server3.servereyes.de] has joined #ubuntu-kernel
=== doko_ [n=doko@dsl-084-059-079-189.arcor-ip.net] has joined #ubuntu-kernel
BenChey spayne02:17
fabbionehey BenC 02:17
spayneBenC: at alst! been waiting all morning :-)02:18
fabbionepitti said to wait with that security patch02:18
fabbioneBenC: linus rejected it02:18
spayneBenC: did you get my emails?02:18
spayneBenC: i spent around two hours talking with the main ndiswrapper about this problem02:19
spayneBenC: what do you think?02:21
fabbioneBenC: sparc is fully installable now02:23
fabbione(at least ubuntu-desktop is)02:23
fabbioneBenC: are you aware of any problem runnign discover on it?02:24
fabbioneit seems that it can't find PCI video cards...02:24
fabbione(discover1)02:24
spayneubuntu works on sparc?02:24
fabbioneyes02:24
fabbioneas of today02:24
fabbionebut it's not a supported architecture02:25
spaynewill this be another archiecture02:25
spayneoh right02:25
fabbioneit's there.. but it won't be supported02:25
spaynethis isn't going to be another debian>02:25
BenCfabbione: only thing to make sure of is that ubuntu discover or upstream discover has the sbus patches02:29
\shhttp://bugzilla.ubuntu.com/show_bug.cgi?id=16539 <- if somebody can confirm this...it's really strange02:29
BenCfabbione: right now the installer doesn't autodetect the network driver and scsi driver on my e3k02:29
BenCspayne: I think updated ndiswrapper will have to wait02:30
fabbioneBenC: ok02:30
spayneBenC: not even an extra package idea?02:30
fabbione\sh: buy better hw02:32
\shfabbione: har har ;) that's ogra not me02:32
\shfabbione: the sony is new :)02:33
\shfabbione: and it's not mine actually :(02:33
BenCspayne: well, I don't have time for the extra package, so someone else would have to do it02:33
spayneBenC: what about if i write a Wiki doc (i am joining the Wiki team) on this problem and how to install ndiswrapper 1.4rc102:34
spayneBenC: and something in the release notes?02:34
BenCspayne: an errata would be called for, yes02:41
BenCprobably talk to mdz about it02:41
spaynewill do02:41
spayneyou mean an update once breezy is released?02:42
BenChopefully02:42
fabbionesorry BenC .. you said that discover1 in Debian fixes the sparc problems_?02:43
BenCwant the release to be solid and tested, and we can try to get some testing on ndsiwrapper v1.4 afterwards and backport it to breezy's kernel02:43
BenCfabbione: yes, I know it works02:43
fabbioneok thanks02:43
BenCdon't know if upstream got the patches02:43
fabbionei have pci devices so i can check02:43
fabbionehere it detects nothing02:43
BenCodd, it used to detect PCI on my blade 10002:44
fabbionethe one from debian works02:50
fabbionethe one on ubuntu segfaults badly02:50
fabbionethe point is to figure how intrusive the changes are02:50
fabbioneour version is very old02:50
fabbioneotherwise we need to think something really hoorible to get it into the archive02:52
fabbionehell everything is working02:52
fabbionewe can't crap out on one package02:52
BenCit wouldn't be so bad if the installer would add the modules I manually selected to /etc/initramfs/modules03:03
BenCatleast the storage module03:03
=== Yagisan [n=jamie@60-241-39-173.tpgi.com.au] has left #ubuntu-kernel []
\shBenC: hmmm....#16539 <- AltGR is also not working on the console...so which is the right package?03:22
=== Yagisan [n=jamie@60-241-39-173.tpgi.com.au] has joined #ubuntu-kernel
BenC\sh: console keyboard mappings are handled by the kernel, but it's a userspacr tool that loads the keymap based on what keyboard you selected during install03:28
BenCwhich I _think_ is console-tools03:28
infinityconsole-tools and console-data, yes.03:29
infinity(tools does the work, data has the keymaps)03:29
BenCthen it's probably -data that has the bug, but I'll let the maintainer take care of that03:30
\shBenC: funny thing about it, that it's working on my horrible broken breezy install...but not on a new warty->hoary->breezy upgrade...I'll do a new install of breezy tonight on my other laptop so if this is issue reoccur then I can confirm this issue for sure...03:30
BenC\sh: sounds like an upgrade issue with console-data03:31
BenC\sh: you should add that to the bug report, it would be very helpful in pinpointing it03:31
infinity\sh : Can you fix it with dpkg-reconfigure console-data?03:31
infinity(Note that after reconfiguring it, you may have to reboot, as the kernel is sometimes a bit anal about hanging onto a loaded keymap)03:32
\shinfinity: I don't have the sony here anymore (colleague went home...tomorrow I can check it) and this evening with the new install..yes..03:32
infinityYes, reconfiguring it works?03:33
infinityOr, "yes, this yes doesn't belong at the end of my sentence, I just put it there for dramatic effect"?03:34
\shinfinity: that I can test it this evening -> yes , right now I don't have the possibility..03:34
\shinfinity: that the point...dramatic effect ,-)03:34
infinityAlright.  If you can get the output of "debconf-show console-data" on a broken system, then try to dpkg-reconfigure it, and if that fixes your issue, the debconf-show afterward too?03:36
\shinfinity: will do 03:37
infinity(Attach both to the bug)03:37
=== Yagisan [n=jamie@60-241-39-173.tpgi.com.au] has left #ubuntu-kernel []
=== BenC [n=bcollins@debian/developer/bcollins] has joined #ubuntu-kernel
=== JaneW [n=JaneW@wbs-146-156-89.telkomadsl.co.za] has joined #ubuntu-kernel
\shBenC: I just did a new install...this strange keyboard behaviour doesn't appear for me...(todays iso)07:49
\shBenC: so something else must be wrong...upgrade issues?07:49
fabbionean upgrade issue doesn't affect the kernel07:50
fabbionethe kernel image is the same either via installing or upgrading07:50
\shfabbione: yes I know...so could it be the console-*? during upgrade?07:51
\shhmmm....w807:51
\shI'm using uk layout on my laptop...lets check after dinner how it behaves with german keyboard layout ;)07:51
fabbioneyes.. console-data relies on config files07:54
fabbionesomething the kernel doesn't for the keyboard07:55
\shsounds like I have to reinstall the sony tomorrow07:56
\shor take it with home over the weekend...to check properly what's wrong...07:56
=== crimsun [i=nobody@warped.bluecherry.net] has joined #ubuntu-kernel
lamontSCORE!09:10
lamontif I start with our kernel, and debian's 2.6.12 config, and hit return to every question, it appears to boot.  Now to find the evil option09:10
jbaileylamont: Still initrd or with initramfs too?09:15
lamontinitrd09:15
lamontjbailey: first we get a booting kernel, then we worry about initramfs09:15
lamontsince I have a trivial workaround for the latter, and no workaround for the former09:15
jbaileyNo prob, I'm just tracking what's going on. =)09:15
fabbioneBenC, jbailey: ping?09:19
jbaileyfabbione: 'sup?09:20
fabbionemeh09:20
=== mkrufky [n=mk@68.160.103.77] has joined #ubuntu-kernel
=== BenC [n=bcollins@debian/developer/bcollins] has joined #ubuntu-kernel
fabbionehey BenC 09:42
fabbioneBenC: can you please test on the fly the discover1 pkg and lib from http://people.ubuntu.com/~fabbione/09:42
fabbioneit has the sparc bug fixes09:42
fabbioneit works for me09:42
fabbionebut i need you to check with multiple sbus thing09:42
fabbioneBenC: please?10:06
fabbionecan you test it10:06
BenChold a sec10:13
fabbionesure10:15
fabbioneif you can test10:15
fabbionediscover all10:15
fabbionewith both the old and the new10:15
fabbioneyou should be able to see the difference10:15
lamontbtw, CONFIG_ACPI_TC1100 is going to need to be added in more than just i386/amd64 configs....10:16
lamontand I assume it wants to be a module everywhere10:17
lamontany reason I shouldn't make it =m on hppa/ia64/sparc/ppc?10:18
fabbionesparc doesn't have ACPI?10:18
fabbionelike ppc...10:18
lamontah, that could do it...10:18
lamontia64 does though, but I'm overhauling that config atm10:18
fabbione:)10:19
lamont# unless you want to implement ACPI on PA-RISC ... ;-)10:20
lamontconfig PM10:20
lamont        bool10:20
=== lamont giggles
BenCfabbione: waiting for the e3k to boot10:22
fabbioneBenC: no rush10:22
fabbionei am pretty sure the fix is right10:22
fabbionebut i want to be triple sure since discover1 is main10:22
fabbioneso one upload > two uploads10:22
BenCwow, minicom reports it was compiled on my birthday10:24
BenCI feel special :)10:24
fabbioneahahah10:25
BenCok, the new one sees my happy meal10:29
BenCbut neither listed my esp scsi10:29
fabbionethat's weird10:30
fabbionebecause i have SBUS stuff here10:30
fabbioneand the new one can see them10:30
fabbionethe SBUS code is identical to debian10:31
BenCit sees my scsi drive10:31
BenCbut esp is already loaded10:31
fabbioneBenC: can you check using the debian one from sid10:36
BenC--module only shows sunhme10:36
fabbioneand see if it can actually see the esp?10:36
BenCyeah10:36
fabbionei am curious.. becuae they might as weel force esp to load10:36
BenC1.7.13 doesn't, and it sees my sunhme10:39
BenC1.7.7 from debian doesn't see either one10:39
fabbioneso the output from 1.7.13 is the same as mine...10:40
fabbioneright?10:40
BenCyes10:40
fabbioneperfect10:40
BenCesp is listed in the discover db10:40
fabbionein both of them?10:40
fabbionei mean both debian and mine.10:40
fabbionei wonder if that could be a problem solved by discover1-data10:41
=== fabbione uploads discover1 in the meantime
fabbione# wrapper for discover command that can distinguish Discover 1.x and 2.x10:47
fabbionediscover_hw () {10:47
fabbione        1)10:48
fabbione                case "$SUBARCH" in10:48
fabbione                  sparc/*) sbus=",sbus" ;;10:48
fabbionethe hack is in hw-detect10:48
fabbione(in the installer10:48
fabbionei will need to talk with Kamion about it10:49
fabbioneanyway i am off for a sleep10:49
fabbionethanks Ben10:49
=== mkrufky is now known as mkrufky-isnt-her
=== mkrufky-isnt-her is now known as mkrufky-not-here

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