=== caldwell_ [n=rob@cpe-69-135-204-242.woh.res.rr.com] has joined #ubuntu-kernel === caldwell_ [n=rob@cpe-69-135-204-242.woh.res.rr.com] has left #ubuntu-kernel ["Leaving"] [12:55] fabbione: util-linux built, just need to do some more testing, and then upload/merge/upload [12:57] lamont: how invasive are the hppa patches for other architectures? [12:58] I noticed most of the changes outside of hppa are related to types, do these really affect non-hppa, or is it just mostly non-changes when compiled on something other than hppa? [01:08] BenC: it was "scary" - so we left them separate. [01:08] I know that people have built i386 kernels from the hppa tree with no ill effects [01:08] dunno how recently, thouhg === lamont must really run [01:09] back on in about 4.5 hours. [05:13] BenC: i did rename some of the patches to match our name convetion. they are all at the end of the 00list-7.11 in the same order as mjg59 did push them to us [05:14] so they are pretty simple to identify [05:22] i didn't remove the old patches yet.. i tend to keep such big changes around easily reversible for at least one release [06:36] hey [06:36] hey Ben [06:37] the acpi doesn't compile, so I am figuring out why [06:37] yes i am looking at it too [06:37] got access to concordia, so I am doing builds there [06:37] the toshiba acpi is one of them [06:37] the one place I found is the toshiba [06:37] eheh so am i :) [06:38] # CONFIG_INPUT_ACERHK is not set [06:38] this needs to be N on amd64... [06:39] klibc_1.0.14-1ubuntu1_sparc.changes ACCEPTED [06:39] finally! [06:39] we can install kernels on sparc again [06:39] ok.. removing toshiba and setting ACER=n on amd64, it builds... [06:39] BenC: did you get access to davis and halley too? [06:40] no [06:40] btw, I am doing i386 build [06:41] does the toshiba need to be working on i386? [06:43] fabbione: wanna test a util-linux for me? [06:45] BenC: it doesn't build on amd64.. and it should on i386 at least [06:45] lamont: sure.. [06:45] deb http://people.debian.org/~lamont/util-linux / [06:45] it's the debian version of hwclock*.sh, I need to go make the ubuntu changes again [06:46] i guess that requires a reboot to verify that it works.. but it's ok.. [06:46] i can do it [06:46] if it's happy, I'll bump it to -6 and upload [06:47] lamont: happy to what degree? [06:47] that it doesn't hang at boot? [06:47] or that it works? [06:48] pretty much - It'd be nice to know that hwclock*.sh are spitting things out in some semblance of good format [06:48] ok... [06:48] and that ocfs2 works [06:49] lamont: ocfs2 will have to wait that i upload the next kernel [06:49] and upgrade the laptop [06:50] it's in a special setup that initrd doesn't handle and that initramfs should... [06:50] but the guy that did push the patch is good.. [06:50] fabbione: Eh? [06:51] jbailey: it's that setup in which i have /boot on IDE and / on usb.. [06:51] remember i was talking about sleep N in initrd? [06:51] because scsi over usb takes time to init.. [06:52] Hmm. [06:52] I have a sleep 2 after the PCI bus for the USB bus to populate. [06:52] Is that not enough for this? [06:52] + if / is on subsystem foo, and /boot on bar, bar doesn't get included in the initrd [06:52] jbailey: in the initrd there is no sleep [06:52] i didn't check initramfs yet [06:52] initramfs has a sleep2 [06:52] Otherwisie it goes by way too fast. [06:52] and as i said i need to upgrade it when i have time to recover in case of mess [06:52] right now i am just too busy [06:53] jbailey; since switching to initramfs i'm getting a lot of warnings during startup [06:53] jbailey; you probably know about these? [06:54] desrt: Depends on the warning. =) [06:54] lamont: is util-linux enough for the clock thingy? [06:54] lvm [06:54] or do i need to upgrade all of them [06:54] desrt: AFAIK there should be no lvm warnings. [06:54] i don't remember exactly... but it's something about can't initialise volume group sda1 [06:55] desrt: LVM tends to spit out gratuitous warnings [06:55] nod. it certainly has no effect on my system successfully booting [06:55] nope.. that's probably lvscan or something [06:56] that hits automatically all the devices [06:57] util-linux is enough for the clock [06:57] mount is enough for ocfs2 [06:57] lamont: ok.. i will be able to test ocfs2 probably tomorrow [06:58] and mount immediatly.. for regressions [06:58] if there are no regressions, can you upload and i will test after? [07:00] fabbione: yes [07:00] the best part is that the diff between debian/ubuntu is now minimized [07:01] lamont: ehehe [07:01] Setting up util-linux (2.12p-5.1) ... [07:01] Installing new version of config file /etc/init.d/hwclock.sh ... [07:01] Installing new version of config file /etc/init.d/hwclockfirst.sh ... [07:01] BenC: i will start test builds on ppc and ia64... [07:01] fabbione: When is your next kernel upload? [07:01] sparc can wait a bit given that there is no ACPI :) [07:01] jbailey: soon.. [07:02] jbailey: and it will be my last kernel upload :) [07:02] I need one with the DSDT patch in it. [07:02] jbailey: than you have more or less today to give me the patch [07:02] brb [07:03] Lovely. That's sometime in the next 24 hours, right? =) [07:07] fabbione: 69 lines of diff -u atm, the first 39 of which are changelog [07:07] hrm... oops. === lamont grumbles at the 2 hour glibc build time [07:15] fabbione: I'm going to sleep now, but when I wake up, I'll see about uploading [07:15] or will you be done playing with ocfs2 then? [07:17] Yup, bedtime here too. [07:18] fabbione: When I get up in the morning I'll do a test build of the kernel. I think I have a patch that will work (It's one I tested before, but I want to do another round of testing) [07:18] fabbione: I need to know how you do your massive ccache speedups, though, if possible. The build time on the kernel is bloody forever. === jbailey -> sleep [07:31] re [07:31] lamont-away: i hope i will be able to test ocfs2 before you wake up [07:32] jbailey: i just use ccache in the normal way [07:32] the speed up is done by export CONCURRENCY_LEVEL in your env [07:32] that's like passing make -j$VALUE [07:32] if your ccache is populated, you will switch to disk I/O [07:32] that's up to your disk at that point [07:32] and yes.. within the next my 24hours [07:33] that means when i will wake up tomorrow there must be a patch [08:06] RIDE THE SPLIT! GO FEENODE! === smurfix [n=smurf@debian/developer/smurf] has joined #ubuntu-kernel [08:18] mjg59: up till now the only failure i get is the toshiba ACPI on amd64 [08:18] it builds on i386 tho [08:19] i still need to hit ppc64 [08:19] ia64 looks good [08:20] hmmm [08:20] i386 just farted [08:36] the toshiba patch is borked [08:41] drivers/acpi/toshiba_acpi.c: In function `toshiba_acpi_init': [08:41] drivers/acpi/toshiba_acpi.c:878: error: `status2' undeclared (first use in this function) [08:41] drivers/acpi/toshiba_acpi.c:878: error: (Each undeclared identifier is reported only once [08:41] drivers/acpi/toshiba_acpi.c:878: error: for each function it appears in.) [08:42] lamont-away: it seems that hwclock did work fine here. === chmj [n=chmj@196.36.161.235] has joined #ubuntu-kernel [10:22] lamont-away: OCFS2 test passed [10:22] you are good to go === JaneW [n=JaneW@wbs-146-136-127.telkomadsl.co.za] has joined #ubuntu-kernel [11:44] fabbione: 12708, our linux-source builds are fine building XFS modules? [11:45] doko: our kernel builds with 3.4 [11:45] and we build XFS [11:45] i have no idea with 4.0 [11:45] last build i did was a long while back [11:55] fabbione: hmm, where do I set the compiler version for the debian build? [11:56] doko: to do what? [11:56] right now if you want to use the default compiler just stop applying the patch that forces gcc-3.4 [11:56] ok [11:58] mjg59: wake up dude... [11:59] bah, doesn't get to the modules ... [11:59] told you... [12:00] we don't support kernel compilation with 4.0 [12:00] that's it.. [12:00] he needs to get over it [12:00] hmm, close the report? [12:01] doko: NOTABUG [12:01] it's not a supported combination [12:02] done === Seveas [n=seveas@seveas.demon.nl] has joined #ubuntu-kernel [01:02] *** Warning: "acpi_ec_read" [drivers/acpi/i2c-acpi-ec.ko] undefined! [01:03] this is on amd64 [01:03] mjg59: WAKE UP DUDE? === JaneW [n=JaneW@wbs-146-136-127.telkomadsl.co.za] has left #ubuntu-kernel ["Bye"] === michele [n=michele@host147-90.pool8540.interbusiness.it] has joined #ubuntu-kernel [01:19] AHHHHHHHH [01:19] dpatch sucks [01:20] the toshiba thing doesn't work, because the patch doesn't apply, and there is no error reported [01:28] fabbione: Ho [01:28] 2 dpatches generate .rej [01:28] and dpatch doesn't catch them [01:28] i noticed by mistake [01:29] Oh. Fun. Which ones? [01:29] one is the toshiba.. [01:29] i am trying to find the other.. [01:30] in fact the reject bit was int status2; [01:30] that's exactly why we were getting the above failures [01:30] That is weird. It applied fine here. [01:30] Bah. [01:30] yeah it applies fine here too [01:30] expect that there is a .rej [01:30] now it did build perfectly.. [01:31] Weird [01:33] now it builds fine :) [01:33] Hurrahj [01:36] jbailey: is suspend to disk supposed to work with initramfs? It seems extraordinarily unstable for me. [01:37] MUHA MUHA MUHA [01:37] Mithrandir: works on my laptop. What problem are you having? [01:37] the last rej hunk is the missing exported symbol that was causing the above [01:37] jbailey: spits out "a" six times then sits around for a while doing nothing. [01:37] Mithrandir: At what point? [01:38] jbailey: just after loading the kernel. Checking if I can reproduce it now [01:38] Mithrandir: Basically, all suspend to disk does is pokes the major:minor into /sys/power/resume or something like that. [01:38] Mithrandir: "a"s are from usplash [01:38] Mithrandir: Try without quiet/splash [01:38] Mithrandir: So from a running system, check to make sure it's set to where your swap is. [01:39] Mithrandir: If it is, all further problems are not initramfs. [01:39] jbailey: initramfs is still too fast for / over scsi over USB [01:40] well, now it resumed (As in, got the state off disk), then gave me a screen reverse-video screen with the cursor blinking on it. [01:41] Mithrandir: Can you ssh to it? It would be worth knowing whether that's a video refresh issue or a fucked-machine issue. [01:41] fabbione: Can you take a look and tell me where you need pauses added? Since I don't have that setup, all I'd be doing is guessing. [01:42] jbailey: no ping, caps lock doesn't work. [01:42] jbailey: There's no way of knowing for sure how long USB will take to populate [01:42] jbailey: scripts/functions.. i added a sleep 5 before and after calling scsi-$something.. [01:42] mjg59: What's the right solution? [01:42] Mithrandir: 2.6.12? [01:42] at the end of functions [01:42] mjg59: yes. [01:43] jbailey: just a sec and i will be more precise [01:43] Mithrandir: PM is generally broken on our 2.6.12. Wait for the one fabbione is currently building. [01:43] mjg59: ok [01:43] jbailey: If the root device hasn't appeared and looks like it might be USB, spend a while spinning? [01:43] fabbione: If you look in /usr/share/initramfs-tools/scripts/functions [01:43] jbailey: yes.. it's in there.. but just one sec. i am powering it on [01:44] fabbione: TThe load_modules function is the one you want. Currently I have: [01:44] # Give the USB bus a moment to catch up [01:44] sleep 2 [01:44] jbailey: it's a bit after that [01:44] if you can wait a minute that's booting.. [01:44] Sure. =) [01:45] done [01:45] ide_boot_events [01:45] scsi_boot_events [01:45] } [01:45] i did add sleep 5 before scsi_boot_events [01:45] and for sake 5 after [01:45] And that still wasn't enough? [01:45] that gives enough time for the disk to appear in /sys [01:45] Oh ouch. =( [01:46] i did only one test adding 5 at the same time [01:46] otherwise scsi and usb-storage are probed [01:46] but not sd_mod [01:46] So you shouldn't need the 5 after, I think. [01:46] mostlikely not.. [01:46] Is there no way to tell that the bus is still scanning? [01:46] but since i didn't read all the code [01:47] dunno.. [01:47] i will need to investigate that [01:47] but that's not todays stuff [01:47] Ayup [01:51] Erp, the concordia i386 chroot doesn't have kernel-wedge and stuff installed? [01:51] yes it does [01:51] linux32 dchroot -c breezy-i386 [01:51] Doh. [01:51] I'm in hoary-i386 [01:51] my bad [01:51] jbailey: if you can give me a few more minutes.. i have almost done with concordia [01:52] and i can hand over to you a kernel source [01:52] Cool. I'm just going to make sure the patch applies cleanly to what I've got right now. [01:54] jbailey: wait.. what you got there is ages behind [01:54] you are really going to waste your time [01:55] i so much have the feeling that i am going to regret the next patch i will apply... [02:01] mjg59: do you have the SATA GO FUCK YOUR SELF SLEEP AND RESUME patch for 2.6.12? [02:01] because the one i got yesterday doesn't apply [02:02] and i don't plan to rediff all that stuff [02:02] i know.. i know i am going to regret to add this patch... === Seveas [n=seveas@seveas.demon.nl] has joined #ubuntu-kernel [02:12] jbailey: i am doing the last test build for today.. and i can give you the source. [02:13] it's tested only ONE flavour on amd64 [02:13] fabbione: Nope. I'll generate you one now. [02:13] it means that it might not build on other arches [02:14] fabbione: Lovely. [02:14] Where do I find it? [02:14] jbailey: as soon as i am done i will tell you and it will be on concordia [02:15] fabbione: Do you know about how long? [02:15] jbailey: 10 minutes? [02:15] Cool. [02:16] fabbione: Mailed [02:18] mjg59: building... [02:19] fabbione: Cool [02:21] * Add experimental sata suspend remove support (it clearly cannot be worst [02:21] than it is now): [02:21] - Add patch external-experimental-sata_suspend-resume.dpatch. [02:21] michele: you got your patch.. now.. next time i pass by italy, you will offer me a beer [02:23] fabbione, let's make them two :) [02:23] michele: where do you live in italy? === zul [n=chuck@CPE0006258ec6c2-CM000a73655d0e.cpe.net.cable.rogers.com] has joined #ubuntu-kernel [02:26] heylo [02:27] fabbione, treviso [02:28] where are you from? [02:28] ah [02:28] michele: i am from Rome, but i have relatices in Venice.. [02:28] hey zul [02:28] relatvies even [02:28] i pass by that area quite often [02:29] ok, let me know next time :) (micampe@micampe.it) [02:29] and well'have a beer or something [02:30] michele: yeah.. it won't be before Xmas [02:30] ok, let's hope I'm not in france by then :) [02:31] what?? NO BEERS? === fabbione removes the sata patch... [02:31] jbailey: building the latest diff.gz for you right now... [02:31] ahah [02:32] fabbione, I'll make sure you'll have your beers before leaving [02:32] eheh [02:32] anyway... when will that kernel be packaged and distributed? [02:32] no later than monday [02:33] ok. colony 3 looks decent, I might try to install it [02:36] fabbione: sata patch sucked? [02:36] zul: yeah.. [02:36] i know i will regret it for the rest of my life [02:36] figured [02:36] but this is my last kernel upload.. or at least i hope so [02:37] so BenC will hunt me down to death at the next Ubuntu conference [02:37] I'll help him [02:37] jbailey: uploading now to concordia [02:37] nah i will still be around till breezy release [02:37] and probably after for the cluster stuff === jbailey blinks [02:38] fabbione: You off kernel stuff completely? [02:38] jbailey: almost :) [02:38] fabbione: you leaving kernel for userspace ? [02:38] it's not my decision, but i agree with who made it.. [02:38] the insanometer is out of scale here [02:39] fabbione, will you regret even after my delightful beers? [02:40] home-brewed just for you! [02:40] michele: the dark side of the force phear you must === fabbione takes a long break [02:47] fabbione: so ill tell benc to where to get my stuff whee.. [02:51] zul: sure.. [02:51] i did merge your last stuff yesterday... [02:51] it's in for the next upload [02:51] but i remember you did a lot of patches while i was in holidays [02:51] you didn't port them to 7.11, did you? [03:03] fabbione: no they were changes to the ide subsystem [03:03] hm ok... [03:04] they were backported from .13 so im not sure if we want them there or not [03:04] hmmm [03:05] i would have to see them again... [03:05] probably it's not a good idea to merge too much from head [03:19] yeah i was thinking that [03:20] ill send you them tonight..it was just two patches though [03:20] both from fedora that got merged into head [03:25] both from alan cox [03:26] ok [03:26] i will look at them tomorrow morning [03:26] i have Danish class later [03:33] okie dokie === Seveaz [n=seveas@seveas.demon.nl] has joined #ubuntu-kernel [04:14] how do I revert a changed file in baz? [04:15] BenC: hey.. [04:15] did you already commit the file? [04:15] hey [04:15] no [04:15] need to revert to get your changes [04:15] are there more changes or only that one? [04:15] just that one [04:15] baz undo [04:16] ok [04:16] note that it will create a ,,LONGDIRECTORYNAME that you can safely remove [04:16] BenC: i did fix all the ACPI stuff [04:16] it builds everywhere.. [04:16] we need to re-test build, save the ABI files and we are ready to upload.. [04:17] what time is it now where you live? [04:17] 10am [04:17] do you think you can come tomorrow at 8am? [04:17] yeah [04:18] so we can do the upload stuff together [04:18] 10:20am I mean [04:18] and i will drive your trough the baz dance [04:18] yeah.. [04:18] sure thing [04:18] and how to start the new release [04:18] so after that you can take over.. [04:18] or almost :) [04:19] we are waiting for a patch from jbailey [04:19] to attach DSTS to initramfs [04:19] that's the latest missing code for 7.11 in what was my list [04:19] if you have more, go ahead and add it today [04:20] tomorrow i will do the build orgy everywhere [04:20] fabbione: woot! [04:20] if you have time to test the kernel, that would be very good too... [04:21] lamont-away: http://people.ubuntu.com/~fabbione/OCFS2-1.1.0.png [04:21] now gimme the util-linux crack :) [04:22] anyway i am off for the next 4 hours and 30.. probably more [04:22] danish class will start soon :) [04:22] and i will pass by again [04:22] lamont-away: in case, can you help BenC with baz while i am not around? === fabbione goes off line [04:24] later guys [04:24] lamont-away: you also want to check that hppa still applies properly to 7.11 [04:24] i had no time to look at it [04:27] fabbione: uploading util-linux, then I have a task to do here, and off to the office. [04:27] will do my best to help BenC if I'm here. [04:28] lamont-away: maybe you can help me get my hppa installed :) [04:29] that's easier. [04:30] http://people.ubuntu.com/~lamont/ubuntu-hppa/tree is a hoary archive for hppa [04:30] alternatively, you start with a sarge install, and then dist-upgrade it to breezy with a few well placed --reinstall's as needed. [04:30] still no d-i for hppa/ubuntu [04:31] I have woody on my hppa, so maybe that wil work [04:31] should - dist-upgrade to hoary, then dist-upgrade to breezy (dunno if you can skip...) === michele [n=michele@host147-90.pool8540.interbusiness.it] has left #ubuntu-kernel ["*click*"] [05:36] can someone else look at bug 3117 and tell me if that looks like an acpi problem, perhaps something with cpu frequency changes for power saving? [05:58] That's an unusual one. [05:58] Hrm, I have a machine that does on-demand frequency scaling, and I've never seen the clock skew like that. [05:58] Normally excessive work results in missed timer interrupts and a slower clock [05:58] In fact, most of us do. [05:59] I'm willing to bet mjg59 has a few dozen. :) [05:59] "I'm seeing this behavior on a brand new Dell so I don't think it's due to crappy [05:59] hardware." ha ha ha [05:59] Oh - if it's related to spread spectrum stuff, then it probably is broken hardware [06:00] The idea there is to change all the clocks slightly to reduce RF interference [06:00] Which tends to break stuff [06:01] Spread spectrum is such utter crack. [06:02] When my phone rings, all the monitors in the house shake, but clearly I should be concerned about the tiny amount of RF interference from my RAM. [06:06] mjg59: so is there a workaround other than telling people to disable spread spectrum? [06:07] BenC: Not that I know of. It's not something that's ever hit me. [06:12] damn [06:12] even google searching just says "disable spread spectrum" [06:12] even the NTP site says to do it === doko_ [n=doko___@dsl-084-059-064-008.arcor-ip.net] has joined #ubuntu-kernel === lamont [n=lamont@15.238.5.97] has joined #ubuntu-kernel [07:51] BenC: ping? [07:52] mjg59: Hey, something that might interest you... [07:52] jbailey: Mm? [07:52] mjg59: The way I'm doing the DSDT replacement, I suspect that it's happening early enough to repalce other tables if you need. [07:53] jbailey: Ah [07:53] Hmm. In principle that's useful, but there's very few cases where people have broken other tables [07:53] 'k, no worries. [07:54] jbailey: You've seen the problems some people seem to be having with booting on Colony 3? [07:54] If it comes up during LaptopTesting, though, I suspect it's an option. [07:54] Excellent [07:54] mjg59: Blank consoles and whatnot? [07:54] jbailey: Failing to mount the root filesystem and dropping to a shell [07:54] Hmm, generally on ubuntu-devel@ or in the channel? [07:54] I've been tring to follow the list and haven't seen much there. [07:55] The list [07:55] Ah, small pile of messages from this morning. [07:55] I'll read them [07:56] The lilo issue I know about and think I have a fix for. [07:56] Dunno about the people specifying a device and it not being there. Probably just a missing driver. [07:57] Anyone here use VmWare? [08:00] jbailey: Oh, yes. I had an issue with a missing driver [08:00] It didn't pull in my IDE module [08:00] I suspect that the logic for grabbing device drivers is faulty [08:00] mjg59: The list that's included is a flat list right now. [08:01] jbailey: Ah! [08:01] I'm going to play with autogenerating it. [08:01] jbailey: Well, it's incomplete :) [08:01] Which module are you missing, though, in the meantime? [08:01] atiixp [08:01] Looks like various SATA and SCSI ones may be missing [08:01] Quite possibly. [08:06] BenC: Whenever you're back, the initramfs DSDT patch is in my homedir on concordia and is called drivers-acpi-osl_attach-dsdt-to-initrd.dpatch [08:07] BenC: It just replaces the current version of that patch. [08:07] BenC: I have tested it booting with and without it replacing the DSDT from the DSDT.aml file in the initramfs [09:40] ah finally good news [09:42] hmmm? [09:43] the patch from jb [09:43] ah [09:53] well im heading home [09:53] c ya later [09:54] later [10:34] fabbione: as you may note, new util-linux for you [10:34] rather, -6ubuntu1 is in the archvie [10:37] lamont: thanks... [10:37] i saw that.. [10:54] fabbione: I haven't had time to play wiht 7.11 on hppa [10:55] i will try a test build tomorrow [10:55] thansk [11:57] jbailey: cool, thanks [11:58] BenC: Do you know when you're targetting an upload of it? [11:59] fab wants to get the current 7.11 uploaded, and then turn thngs over to me tomorrow, so I suspect soon after that [12:00] Cool, so I don't need to stay up all night tonight watching for people complaining of things breaking. =) [12:01] nope :) [12:01] got the patch [12:02] any certain order that this has to be applied in relation to the other acpi patches?