| === allee_ is now known as allee | ||
| mjg59 | BenC: #180678 - seriously. Every single release I can remember. This really isn't funny any more. | 13:54 |
|---|---|---|
| alex_joni | bug #180678 | 13:54 |
| ubotu | Launchpad bug 180678 in acpi-support "dmesg: toshiba_acpi: Unknown parameter `hotkeys_over_acpi'" [Medium,In progress] https://launchpad.net/bugs/180678 | 13:54 |
| BenC | mjg59: every release it is the same reason...acpi core changes so much the patch cannot be applied | 13:56 |
| BenC | mjg59: pretty sure it was in the list of "patches that got left behind" | 13:56 |
| BenC | mjg59: we've audited the patches that got removed at hardy kernel rebase, and that's the only one | 13:57 |
| mjg59 | BenC: As far as I can tell, that mail never went anywhere public? | 13:58 |
| mjg59 | Amit replied with a subset Cc:ed to kernel-team | 13:58 |
| BenC | mjg59: I mean the list I sent you months ago | 13:59 |
| mjg59 | BenC: No, it's not in that list | 14:03 |
| mjg59 | Can you check whether any other patches have fallen through the cracks? | 14:04 |
| amitk | BenC: can you please send me this other list that you sent to mjg59 | 14:14 |
| BenC | mjg59: I told ya we just did an audit of the patches :) | 14:14 |
| BenC | amitk: it's the same one you got, minus stuff that already got checked | 14:14 |
| amitk | BenC: ok | 14:14 |
| mjg59 | BenC: As I said, that patch isn't in hardy git. It was in gutsy. It's not in the email you sent me. | 14:15 |
| mjg59 | If your audit process doesn't pick that up, then that's a problem with your audit process | 14:15 |
| mjg59 | Which is consistent with previous occasions where patches have gone missing from the lists of missing patches | 14:16 |
| mjg59 | BenC: 2d35697db5e3c4f240ac288c7c5abeb5a03c805c in gutsy | 14:18 |
| BenC | interesting...it applies cleanly...wonder why it didn't hit the rebase | 14:19 |
| mjg59 | BenC: It would be good to know if anything else falls into the same category | 14:20 |
| rtg | BenC: be sure to mark it as a SAUCE patch so we don't lose it again. | 14:20 |
| === asac_ is now known as asac | ||
| siretart | did someone already encounter bug #197006, and/or knows a workaround? | 14:59 |
| ubotu | Launchpad bug 197006 in linux-meta "NFS over Unionfs prevents updating existing files" [Undecided,New] https://launchpad.net/bugs/197006 | 14:59 |
| BenC | mjg59: toshiba_acpi patch applied | 15:16 |
| mjg59 | BenC: Thanks | 15:16 |
| rtg | tjaalton: I'm turning the ABI crank today. I'll upload -11 in a few hours in advance of Alpha 6 scheduled for next Thursday. | 15:55 |
| Kano | hi rtg , did you find my bug report | 16:12 |
| Kano | rtg: btw. the ivtv-fb driver is not working, maybe compiled against wrong header | 16:12 |
| rtg | Kano: I didn't notice it. What number is it? | 16:12 |
| Kano | #196745 | 16:12 |
| Kano | will check if i can do a fix for ivtv-fb | 16:13 |
| rtg | Kano: we are still working on the ABI mismatch which is probably what is wrong with ivtv-fb | 16:13 |
| Kano | yes the most easy way is to overwrite the included .h files with the one from the kernel | 16:13 |
| Kano | that fixes it | 16:13 |
| rtg | Kano: can't do that (debian policy). see https://wiki.ubuntu.com/KernelTeam/Sprints/Feb2008/HeadersABI for possible solution | 16:14 |
| Kano | then remove the included headers | 16:15 |
| Kano | how about that | 16:15 |
| Kano | they are not needed anyway | 16:15 |
| rtg | dunno. we'll get it figure out before Beta. | 16:15 |
| Kano | the fastest hack is just to replace the ivtv-fb headers - not the kernel ones | 16:16 |
| rtg | Kano: at the risk of sounding pedantic, file a LP bug so I can track it. | 16:18 |
| Kano | rtg: will try to fix it and then report a bug | 16:18 |
| Kano | i have got a pvr 350.. | 16:18 |
| Kano | just dont use it so often | 16:19 |
| tjaalton | rtg: ok, I have nvidia almost ready to go, targeted -11 | 16:24 |
| Kano | rtg: how about enabling vmware client modules? | 16:24 |
| rtg | Kano: which ones? seems like there is a pending LP for a sound module. | 16:25 |
| Kano | well the standard ones needed for vmplayer | 16:25 |
| Kano | i think those will not change | 16:26 |
| rtg | Kano: do you think enabling ALSA in the kernel would be sufficient? | 16:26 |
| Kano | virtual box needs always differnet modules, so i would not integrat it | 16:26 |
| Kano | i think you have alsa already enabled? | 16:27 |
| Kano | didnt you | 16:27 |
| tjaalton | zul: what about xen & nvidia? | 16:27 |
| zul | tjaalton, i havent got a chance to look at it | 16:28 |
| tjaalton | zul: ok | 16:29 |
| Kano | rtg: also did you notice that "ifneq ($(wildcard /CurrentlyBuilding),)" has to be "ifeq ($(wildcard /CurrentlyBuilding),)" in debian/rules.d/0-common-vars.mk ? | 16:40 |
| Kano | currently it is absolutely stupid that it compiles faster when there is a dir/file named /CurrentlyBuilding | 16:40 |
| Kano | dont you think so... | 16:41 |
| Kano | ps aux|grep -e -j | 16:42 |
| rtg | Kano: looks to me like it just overrides CONCURRENCY_LEVEL when on a buildd. | 16:42 |
| Kano | yes, but completely in the wrong way | 16:43 |
| Kano | why would somebody build on a multi core machine with -j1? | 16:43 |
| Kano | you could even disable that check | 16:44 |
| rtg | Kano: because there might be other builds going on the same machine. | 16:44 |
| Kano | rtg: i thought that, but then the logic is completely reversed! | 16:44 |
| Kano | as you compile with 1 thread when there is NO such file and with twice the number of cores when it is there | 16:45 |
| Kano | check it out yourself | 16:45 |
| Kano | you will at least see -j2 even with a single core when the file is there and -j1 when it is not there | 16:48 |
| Kano | must be clear | 16:48 |
| Kano | that this is wrong | 16:48 |
| rtg | Kano: this particular problem is not important. I've a zillion other issues to deal with. buildd passes happen overnight for me, so I just don't care. | 16:48 |
| Kano | rtg: but i care | 16:49 |
| rtg | start a bug report then. | 16:49 |
| Kano | you only have to remove the "n" | 16:49 |
| Kano | https://bugs.launchpad.net/ubuntu/+source/linux/+bug/197040 | 17:06 |
| ubotu | Launchpad bug 197040 in linux "reversed logic for bbuild check leads to -j1 default" [Undecided,New] | 17:06 |
| Kano | here it is | 17:06 |
| === alleeHol is now known as allee | ||
| Kano | sound/usb/usx2y/snd-usb-usx2y.o | 17:31 |
| Kano | you mean that erro? | 17:31 |
| Kano | hmm no, the error is in toshiba_acpi.c | 17:52 |
| Kano | what did you do there | 17:52 |
| rtg | Kano: I'm just getting to buld testing. looks like the acpi crack broke something. | 17:54 |
| Kano | yes it does not build | 17:54 |
| === doko_ is now known as doko | ||
| tjaalton | rtg: lrm with new nvidia ready | 17:54 |
| rtg | tjaalton: I'm still build testing. it'll be a bit yet. | 17:55 |
| tjaalton | rtg: np | 17:55 |
| Kano | tjaalton: if you added .12 then you can be sure it will not be the last update | 17:56 |
| tjaalton | Kano: surprise | 17:56 |
| Kano | latest is still set to 169.112 | 17:57 |
| Kano | that can show my script with -v latest option... | 17:57 |
| Kano | DISPLAY= fakeroot install-nvidia-debian.sh -pv latest | 17:58 |
| Kano | 169.12 | 17:58 |
| rtg | mjg59: can you have a look at the acpi patch BenC applied? It doesn't build, so I've reverted it. I'm kind of on a deadline for Alpha 6. | 17:58 |
| mjg59 | rtg: Not currently, no | 17:58 |
| mjg59 | I'm up to my eyes in mobile | 17:59 |
| rtg | mjg59: ok, its gonna have to wait until after -11 is released. | 18:00 |
| mjg59 | Well, as long as it hits before final, I'm unfussed | 18:00 |
| rtg | mjg59: can do. | 18:00 |
| Kano | btw. you know that at least 2 ide modules are too much? | 18:02 |
| Kano | via82cxxx and amd74xx are useless, because you have pata_via and pata_amd | 18:03 |
| Kano | which work much better | 18:03 |
| Kano | maybe more, but you can be sure that pata drivers are tested for those | 18:04 |
| Kano | tjaalton: so what did you package 169.12 or 171.05? | 18:05 |
| Kano | btw. did you try pxe boot for your live images? | 18:09 |
| Kano | live-helper can for example use net as target | 18:10 |
| Kano | and creates a tftpboot dir | 18:10 |
| boucman | hello all | 18:12 |
| boucman | I have been skimming the bug list for some time, but can't find a bug similar to what I observe | 18:12 |
| boucman | my machine has some huge speed problems with 2.6.24 (2.6.22 is fine) | 18:13 |
| Kano | boucman: lsmod|grep -e ide -e pata | 18:13 |
| Kano | best check this | 18:13 |
| Kano | if you have got an legacy ide + pata driver loaded blacklist the old ide and run: update-initramfs -u | 18:14 |
| boucman | Kano: thx... I'm using 2.6.22 right now, are you interested in what it reports, or should I reboot in 2.6.24 to check ? | 18:14 |
| Kano | well you would need 2.6.24 | 18:15 |
| Kano | to get this output | 18:15 |
| boucman | I do get an output :P | 18:15 |
| Kano | sure, but i dont think something with pata or? | 18:16 |
| boucman | Kano: you're right | 18:16 |
| boucman | I'll reboot in a minute, but since I'm unable to log to irc with that kernel, what should I be looking for precisely ? | 18:17 |
| Kano | the problem is when you would do that now you could not boot your old kernel ;) | 18:17 |
| Kano | i would prefer those old modules removed, but.. | 18:17 |
| boucman | you mean 2.6.22 wouldn't boot anymore after doing that ? | 18:17 |
| Kano | if you blacklist the module and run update-initramfs -u (updates initramfs for current kernel) | 18:18 |
| Kano | usually the new kernel runs the devices without dma access | 18:19 |
| Kano | because 2 drivers want to work with the same device, disabling the old ones fixes that usally | 18:19 |
| boucman | Kano: ok, can you walk me throught the process (or point me to a web page ) I would be gratefull | 18:20 |
| Kano | well i dont think that ubuntu ppl did that before, thats what i did for kanotix users ;) | 18:20 |
| boucman | ok, so | 18:22 |
| boucman | 1) blacklist the module | 18:22 |
| tjaalton | Kano: 169.12 | 18:22 |
| boucman | 2) reboot with .24 kernel | 18:22 |
| boucman | 3) run "update_initramfs -u" | 18:22 |
| boucman | right ? | 18:22 |
| Kano | no you forget the major step | 18:22 |
| boucman | oh, | 18:22 |
| Kano | the blacklisting | 18:22 |
| boucman | that was step 1, but I'm not sure how to do that | 18:23 |
| Kano | show me lsmod|grep -e ide -e pata | 18:23 |
| Kano | after reboot | 18:23 |
| boucman | Kano: ok, I'll reboot, do that and come back. | 18:23 |
| boucman | then we'll discuss how to fix | 18:23 |
| boucman | and thx a million for your help | 18:23 |
| boucman | brb | 18:23 |
| boucman | back | 18:27 |
| Kano | so check lsmod|grep -e ide -e pata | 18:28 |
| Kano | also | 18:28 |
| Kano | cat /etc/fstab | 18:28 |
| Kano | cat /proc/cmdline | 18:28 |
| boucman | video 23316 0 | 18:28 |
| boucman | output 5632 1 video | 18:28 |
| boucman | ide_core 136344 1 amd74xx | 18:28 |
| boucman | pata_amd 16772 0 | 18:28 |
| boucman | pata_acpi 9856 0 | 18:28 |
| boucman | libata 176304 4 pata_amd,sata_nv,ata_generic,pata_acpi | 18:28 |
| Kano | to be sure that you use uuids | 18:28 |
| boucman | that's under .24 (i'm back to .22 right now) | 18:28 |
| Kano | the problem is definitely amd74xx | 18:28 |
| boucman | ok | 18:28 |
| Kano | why dod you boot the .22 kernel? | 18:28 |
| Kano | doesnt it boot at alls? | 18:29 |
| boucman | it does, but usually it takes approx 5' just to start pidgin | 18:30 |
| boucman | so not really usable | 18:30 |
| Kano | cat /proc/cmdline | 18:30 |
| Kano | show this | 18:30 |
| boucman | root=UUID=aefdb9c5-c42d-4de4-863d-e8db4afcd1e8 ro quiet splash | 18:30 |
| Kano | ok, should be no problem to use the fix | 18:30 |
| boucman | :) | 18:30 |
| Kano | which 2.6.24 do you want to fix | 18:30 |
| boucman | main one, I guess | 18:31 |
| Kano | -10? | 18:31 |
| boucman | yes, that's the latest one | 18:31 |
| Kano | ls /boot/vmlinuz-2.6.24-* | 18:31 |
| boucman | (so far :P ) | 18:31 |
| Kano | show this | 18:31 |
| boucman | /boot/vmlinuz-2.6.24-10-generic | 18:31 |
| Kano | echo blacklist amd74xx > /etc/modprobe.d/custom-blacklist | 18:31 |
| Kano | update-initramfs -uk 2.6.24-10-generic | 18:32 |
| Kano | then you should be able to boot new kernel | 18:32 |
| boucman | running... | 18:33 |
| boucman | ok, done | 18:33 |
| boucman | I'll be back in a minute (hopefull with .24) thx again for your time | 18:33 |
| Kano | i hope you did that after sudo -i | 18:33 |
| Kano | then it should work | 18:34 |
| boucman | yeah, I used the "administrator terminal" not the normal one | 18:34 |
| Kano | rtg: at least it compiles with reverted toshiba | 18:35 |
| Mike_Feravolo | Hello People | 18:36 |
| Mike_Feravolo | I loaded the lastest hardy update and my dual amd opteron 840 will not boot anymore | 18:37 |
| Kano | Mike_Feravolo: i guess because of amd74xx kernel module in initrd ;) | 18:38 |
| boucma1 | hmm | 18:38 |
| Mike_Feravolo | i just what to make sure that someone knows | 18:38 |
| Kano | boucma1: didit boot | 18:38 |
| boucma1 | Kano: yes, but pidgin froze at startup like it used to, I'll start a few app to see if it's better than previously | 18:39 |
| Mike_Feravolo | i have been running the hardy alpha for a couple of months and this is the first major snarl | 18:39 |
| boucma1 | and login was very slow... | 18:39 |
| Kano | boucma1: but at last the hd should be fast, you could check that | 18:40 |
| Mike_Feravolo | kano- hangs on the UBUNTU splash screen | 18:40 |
| boucma1 | Kano: yes, it seems to work better but still very slow | 18:40 |
| Kano | boucma1: i only use the kenrel not the rest | 18:40 |
| boucma1 | Kano: well, pidgin start time was my measure of perf, but it applies to everything | 18:41 |
| boucma1 | it's usable now, at least | 18:41 |
| Kano | boucma1: fdisk -l | 18:41 |
| Kano | then you will see only sd? | 18:41 |
| Kano | you can then check performance with | 18:41 |
| Kano | hdparm -tT /dev/sd? | 18:41 |
| Mike_Feravolo | I am not near that system now, but i will try a few things before reloading gusty - Thank You | 18:42 |
| Kano | Mike_Feravolo: can you boot antoher kernel? | 18:42 |
| boucma1 | /dev/sda: | 18:42 |
| boucma1 | Timing cached reads: 1604 MB in 2.00 seconds = 801.97 MB/sec | 18:42 |
| boucma1 | Timing buffered disk reads: 220 MB in 3.00 seconds = 73.27 MB/sec | 18:42 |
| boucma1 | (no idea if that's good | 18:42 |
| Kano | sure, without dma it would be less than 10 the 2nd one | 18:43 |
| boucma1 | ok | 18:43 |
| Mike_Feravolo | yes the gusty 64 live cd boots, so does dsl, slax and other 32bit distros | 18:43 |
| Kano | well you only need a live cd to chroot to it | 18:44 |
| Kano | if your install is 64 bit you need that of course | 18:44 |
| Mike_Feravolo | kk | 18:44 |
| Mike_Feravolo | thanks | 18:44 |
| boucma1 | ok, thx a lot, I'll use it like that | 18:44 |
| boucma1 | bye all | 18:44 |
| Kano | rtg: i hope you get how bad the amd* module is... | 18:45 |
| Kano | via is not needed too | 18:45 |
| Kano | blacklisting is a bit dangerous to older kernels,but easy to verify | 18:47 |
| Kano | best disable the modules you dont need | 18:47 |
| Kano | via+amd are the most common ones | 18:50 |
| Kano | rtg: disable at least | 19:10 |
| Kano | grep BLK_DEV /boot/config-2.6.24-10-generic |grep -e AMD -e VIA | 19:10 |
| Kano | CONFIG_BLK_DEV_AMD74XX=m | 19:10 |
| Kano | CONFIG_BLK_DEV_VIA82CXXX=m | 19:10 |
| Kano | to avoid huge problems | 19:10 |
| Kano | amd is critical | 19:10 |
| Kano | rtg: also you could use -jX for lum too | 19:33 |
| Kano | rtg: http://paste.debian.net/50205 | 19:37 |
| Kano | thats the problem with ivtv-fb | 19:37 |
| Kano | rtg: what you try is to use ivtv-driver.h | 19:58 |
| Kano | rtg: thats basically <linux/ivtv.h>, but then you have to fix the driver itself | 19:58 |
| rtg | Kano: is there a patch for ivtv-fb ? | 19:59 |
| Kano | well not an easy one | 19:59 |
| Kano | 1.0.3 is only up to 2.6.23 | 19:59 |
| Kano | where did you get the driver? | 19:59 |
| rtg | dunno, probably inherited it from Feisty/Gutsy | 20:00 |
| === clever_ is now known as clever | ||
| Kano | will check if the driver can be used from v4l hg | 20:02 |
| Kano | maybe you need then ivtv + ivtvfb from hg | 20:03 |
| Kano | hmm that will also need m52790 | 20:10 |
| rtg | Kano: https://bugs.edge.launchpad.net/ubuntu/+source/linux-ubuntu-modules-2.6.24/+bug/197089 | 20:16 |
| ubotu | Launchpad bug 197089 in linux-ubuntu-modules-2.6.24 "ivtv oops" [Medium,Confirmed] | 20:16 |
| Kano | well disalbing is not really fun. as you could even use a new xorg driver with it or for pvr350 output plugin | 20:17 |
| Kano | btw. the -j bug is in lum too | 20:18 |
| Kano | for debugging -j1 is more easy but in that case you can set CONCURRENCY_LEVEL=1 if needed | 20:18 |
| Kano | rtg: maybe it is much more easy | 20:29 |
| Kano | you have that module already | 20:30 |
| Kano | in the main kenrel | 20:30 |
| Kano | but you miss the right firmware | 20:30 |
| rtg | Kano: attaching a patch to the LP report increases the likely hood that it will get fixed . | 20:31 |
| Kano | rtg: you can remove the code completely from lum | 20:31 |
| Kano | the module is default and enabled in the kernel | 20:31 |
| Kano | you only need the correct firmware | 20:31 |
| rtg | which needs to be updated in lum | 20:32 |
| Kano | hmm | 20:32 |
| Kano | when this is already in lum just disable the module -you can completely remove the code | 20:33 |
| Kano | http://paste.debian.net/50210 | 20:33 |
| Kano | when i remove lum | 20:33 |
| Kano | grep -i ivtv /boot/config-$(uname -r) | 20:34 |
| Kano | CONFIG_VIDEO_FB_IVTV=m | 20:34 |
| Kano | CONFIG_VIDEO_IVTV=m | 20:34 |
| Kano | standard kernel | 20:34 |
| rtg | is the firmware the same version as is already in lum? the file names are the same as the web site. | 20:34 |
| Kano | didd not try, but look | 20:34 |
| Kano | isnt that nice | 20:34 |
| rtg | its certainly an easy fix. | 20:35 |
| NthDegree | where can I find a list of .patch files used in the Ubuntu Gutsy kernel? as in each individual patch, like for example just ubuntu's apparmor patches or just ubuntu's dazuko patches | 20:43 |
| Kano | rtg: yes and it seems to work | 20:45 |
| rtg | NthDegree: https://wiki.ubuntu.com/KernelGitGuide | 20:45 |
| Kano | http://paste.debian.net/50212 | 20:46 |
| rtg | NthDegree: search the log for 'SAUCE:'. Thats most of our special patches. | 20:46 |
| NthDegree | hmmm | 20:46 |
| Kano | so you just have to disable some legacy ide modules | 20:47 |
| rtg | Kano: I'm still researching that | 20:47 |
| NthDegree | rtg: so there's no chance of just grabbing things like a linux-2.6-apparmor.patch or the like that applies to the Ubuntu kernel then? | 20:47 |
| Kano | rtg: kanotix thorhammer has a kernel completely without config_ide set. only very few pcs could not boot it | 20:48 |
| Kano | the 2 modules i told you before are definitely verified | 20:48 |
| Kano | and share the same pciids | 20:48 |
| rtg | Kano: its likely that I'll blacklist them so that the binaries are still built, just not installed. | 20:48 |
| Kano | no good idea | 20:49 |
| Kano | because when somebody executes update-initramfs -uk all | 20:49 |
| Kano | then older 2.6.22 kernel will never be bootable | 20:49 |
| Kano | isnt something like that even triggered by module-init-tools? | 20:51 |
| Kano | you can use that as hotfix | 20:51 |
| Kano | but no for permanent use | 20:51 |
| rtg | Kano: start an LP report so we don't lose track, but its gonna have to wait until after Alpha 6 and I'm gone Mar 4-27 so you'll have to deal with the other kernel folks | 20:52 |
| zul | hmmm... https://wiki.ubuntu.com/ZenKernel | 21:13 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!