**** BEGIN LOGGING AT Tue Sep 18 02:59:59 2012
Sep 18 08:50:07 <ogra_> ppisati, ddi you include the fix for bug 746137 in your last upload ?
Sep 18 08:50:09 <ubot2> Launchpad bug 746137 in linux-ti-omap4 "Page allocation failure on Pandaboard and Beagle XM" [High,Confirmed] https://launchpad.net/bugs/746137
Sep 18 08:50:34 <ogra_> seems QA is seeing it recently on the bamboo-feeder http://pastebin.ubuntu.com/1212016/
Sep 18 09:15:14 <ppisati> ogra_: AFAIK there's no fix
Sep 18 09:15:58 <ogra_> err, cooloney worked one out weeks agai
Sep 18 09:16:00 <ogra_> *ago
Sep 18 09:16:20 <ppisati> then i'm not aware of it
Sep 18 09:16:30 <ppisati> i know there's a systcl to workaround it
Sep 18 09:16:39 <ppisati> but IMO all the fixes proposed so far
Sep 18 09:16:43 <ppisati> are not working
Sep 18 09:16:43 <ogra_> GFP_ATOMIC vs GFP_KERNEL i think
Sep 18 09:16:52 <ppisati> cooloney: ^
Sep 18 09:16:54 <cooloney> ogra_: i saw patches for Panda was merged by ppisati
Sep 18 09:17:07 <cooloney> the patch is a hack from Andy Green
Sep 18 09:17:11 <ogra_> oh, he merged it without knowing ?
Sep 18 09:17:16 <ogra_> haha
Sep 18 09:17:23 <ogra_> k, as long as its in ...
Sep 18 09:17:24 <cooloney> that's for panda, not for Beagle
Sep 18 09:17:29 <ogra_> ah
Sep 18 09:17:42 <ogra_> well, would be good to have it in beagle too
Sep 18 09:17:43 * ppisati is schizophrenic then
Sep 18 09:17:54 <ppisati> well guys
Sep 18 09:17:59 <ogra_> since we have no way to apply userspace hacks anymore like we did with preinstalled images
Sep 18 09:18:09 <ppisati> but if we still have that problem even with the supposed fix
Sep 18 09:18:13 <cooloney> ogra_: i tried a patch MingLei gave to me from upstream some maintainer.
Sep 18 09:18:15 <ppisati> it's clearly not working
Sep 18 09:18:22 <cooloney> ogra_: but it doesn't work.
Sep 18 09:18:23 <ogra_> ppisati, how would you know ?
Sep 18 09:18:38 <ogra_> your upload is in the archive but we have no new images with it yet
Sep 18 09:18:49 <ppisati> ogra_: but it's not there for sure
Sep 18 09:18:52 <ppisati> i mean
Sep 18 09:18:59 <ppisati> cooloney: which patch were you referring?
Sep 18 09:19:16 <ppisati> ogra_: in the last upload there's the video/unity flickering fix
Sep 18 09:19:17 <ogra_> hmpf, i had the impression that was long fixed
Sep 18 09:19:31 <ppisati> ogra_: and a rebase on origin/master (latest 3.5 stable)
Sep 18 09:19:32 <cooloney> ppisati: a patch not in public, it is for smsc95xx usb net driver
Sep 18 09:19:56 <ogra_> ah, if its driver specific that should work for beagle too then
Sep 18 09:20:00 <cooloney> ppisati: the patch from Andy is in your ti-omap4, so it should be workarounded
Sep 18 09:20:05 <ogra_> they use the same chip
Sep 18 09:20:14 <cooloney> but for beagle it is in master.
Sep 18 09:20:20 <ppisati> ogra_: iirc, conrad approached me weeks ago and told they have the same problem on builders
Sep 18 09:20:28 <cooloney> the hack will touch other flavors.
Sep 18 09:20:31 <ogra_> conrad ?
Sep 18 09:20:39 <ppisati> ogra_: adam
Sep 18 09:20:40 <ogra_> do you mean infinity ?
Sep 18 09:20:43 <ppisati> yes
Sep 18 09:20:43 <cooloney> infinity: ^^
Sep 18 09:20:59 <ogra_> ah, conrad adam ... the evil twin of adam conrad *g*
Sep 18 09:21:06 <ppisati> :)
Sep 18 09:21:36 <ogra_> i dont think the builders run quantal, so thats not a surprise
Sep 18 09:22:25 <ogra_> and they dont use preinstalled images to set them up (which would get them the fix) but use ubuntu-core with manual config or some such
Sep 18 09:22:39 <ppisati> ogra_: yep
Sep 18 09:22:49 <ppisati> ogra_: what i meant was that we never really fixed it
Sep 18 09:22:59 <ogra_> s/get them the fix/get them the userspace fix/
Sep 18 09:23:01 <ppisati> ogra_: and if we have a supposed workaround in
Sep 18 09:23:07 <ppisati> ogra_: but we still experience it
Sep 18 09:23:29 <ogra_> well, i lost the opportunity to work around it in userspace ...
Sep 18 09:23:30 <ppisati> ogra_: then we should a) investigate the problem b) back out the workaround
Sep 18 09:23:41 <ogra_> and i havent seen it in ages here in my tests
Sep 18 09:24:01 <ogra_> so i personally had the impression it was long fixed
Sep 18 09:24:03 <infinity> ogra_: The buildds use d-i over PXE on precise, but they have the userspace hack in puppet.
Sep 18 09:24:12 <ogra_> until hggdh pinged me over night today
Sep 18 09:24:27 <ppisati> infinity: and do they suffer the problem?
Sep 18 09:24:33 <infinity> ppisati: No.
Sep 18 09:24:37 <ppisati> uhm
Sep 18 09:24:37 <ogra_> and funnily the saem thing is discussed on #beagle today too
Sep 18 09:24:54 <ogra_> ppisati, the userspace hack always worked
Sep 18 09:25:04 <infinity> ppisati: But any machine installing from an installer that doesn't put the userspace hack in place will have the issue.
Sep 18 09:25:12 <ppisati> the userspace stuff was the sysctl config, right?
Sep 18 09:25:16 <infinity> Right.
Sep 18 09:25:16 <ogra_> right
Sep 18 09:25:42 <ogra_> worst case we should just bump it at build time for certain arches
Sep 18 09:25:57 <ogra_> in the kernel code ... if thats possible
Sep 18 09:26:22 <ogra_> so we would just move the userspace hack into the kernel
Sep 18 09:26:28 <ppisati> i wonder if can pass it via cmdline
Sep 18 09:26:57 <ogra_> well you can definitely just create the sysctl.d file ... its a one liner
Sep 18 09:27:08 <ogra_> not susre sysctl values are read from cmdline
Sep 18 09:27:17 <ppisati> right
Sep 18 09:27:19 <infinity> If it should always start at a certain value, and it's not, passing it on the cmdline (or sysctl) is the wrong answer.
Sep 18 09:27:24 <infinity> The initial value should be correct.
Sep 18 09:27:30 <infinity> I think we've been over this a few dozen times.
Sep 18 09:27:34 <ppisati> but they are kernel variables
Sep 18 09:27:36 <ogra_> heh, yeah
Sep 18 09:27:57 <infinity> ppisati: Yes, it's a variable, but variables should have sane defaults.  This one clearly isn't sane for ARM.
Sep 18 09:30:00 <ppisati> vm.min_free_kbytes = 8192
Sep 18 09:30:23 <infinity> Anyhow, I should sleep.
Sep 18 09:30:23 <ppisati> uhm
Sep 18 09:30:33 <infinity> If I recall, apw was deeply involved in the last round of this conversation.
Sep 18 09:30:38 <cooloney> ppisati: i suggest we instroduce a config option in master branch like CONFIG_VM_MIN_FREE_KBYTES
Sep 18 09:30:41 <ppisati> sleep well
Sep 18 09:30:44 <infinity> And he doesn't seem to be in the channel.
Sep 18 09:30:51 <infinity> So you might want to take this to #-kernel.
Sep 18 09:30:57 <cooloney> which should be 8192 by default for x86, powerpc
Sep 18 09:31:05 <cooloney> and 32768 for ARM
Sep 18 09:31:19 <ppisati> vm.min_free_kbytes = 32768
Sep 18 09:31:23 <ppisati> on my Q/omap4
Sep 18 09:31:31 <cooloney> and porting Andy's hacking patch to master
Sep 18 09:31:57 <cooloney> ppisati: yeah, it should be 32768 for ARM since it was forced to be that in mm/page_alloc.c
Sep 18 09:32:12 <ppisati> so we are "already "good"
Sep 18 09:32:14 <ppisati> ?
Sep 18 09:32:28 <infinity> Oh, kay, so we HAVE worked around it in the kernel?
Sep 18 09:32:30 <cooloney> yeah, but beagle is omap3, which is master kernel
Sep 18 09:32:30 <ogra_> on omap4, but not on omap (which uses the same driver)
Sep 18 09:32:40 <cooloney> not on your branch,
Sep 18 09:32:44 <infinity> Ahh, not in master.  Check.
Sep 18 09:32:48 <ppisati> wait
Sep 18 09:32:52 <cooloney> we need put such similar patch in master
Sep 18 09:33:00 <infinity> Couldn't you just wrap it in #ifdef __ARM__ and commit to master?
Sep 18 09:33:02 <ppisati> ogra_: didn't you say we hit this problem on omap4 tonight?
Sep 18 09:33:22 <cooloney> if you put the same patch in master, it will set it to 32768 for all the flavors
Sep 18 09:33:27 <cooloney> like x86 and powerpc
Sep 18 09:33:32 <infinity> I guess highbank and armadaxp would pick it up too, then, but that's not world-ending.
Sep 18 09:33:38 <ogra_> ppisati, i was notified by hggdh in #ubuntu-bugs tonight with the pastebin i gave above ... havent talked to him yet
Sep 18 09:33:44 <infinity> cooloney: Hence why I said wrap it in an ifdef.
Sep 18 09:33:45 <ppisati> ogra_: oh
Sep 18 09:33:53 <infinity> cooloney: Though, obviously a CONFIG_ is better.
Sep 18 09:34:02 <ogra_> (since he is asleep like every sane person on that american continent atm)
Sep 18 09:34:04 <ppisati> so, even with the fix
Sep 18 09:34:08 <ppisati> we hit the problem
Sep 18 09:34:09 <infinity> cooloney: And the CONFIG_ change would be upstreamable even.
Sep 18 09:34:26 <cooloney> infinity: yeah, that should be ok for just local hacking, heh
Sep 18 09:34:32 <ogra_> ppisati, well, lets see what exactly he tested there
Sep 18 09:34:37 <ppisati> ok
Sep 18 09:34:38 <ogra_> i need to talk to him first
Sep 18 09:34:47 <cooloney> infinity: i don't think we need such hacking for highbank and amadaxp
Sep 18 09:34:48 <ppisati> ok
Sep 18 09:34:50 <ppisati> wait a sec
Sep 18 09:34:54 <ogra_> in any case people in #beagle are seeing it and have discussed it today
Sep 18 09:35:00 <cooloney> because they don't use usb ethernet controller
Sep 18 09:35:01 <infinity> cooloney: Right, hence why a CONFIG_ would be better.
Sep 18 09:35:02 <ppisati> qait wait
Sep 18 09:35:06 <ppisati> i just noticed one thing
Sep 18 09:35:08 <cooloney> only beagle
Sep 18 09:35:52 <cooloney> infinity: CONFIG_ can be set as different value for different flavor
Sep 18 09:35:56 <infinity> ogra_: That pastebin is an ancient kernel.
Sep 18 09:36:00 <ppisati> tight
Sep 18 09:36:03 <ppisati> rigjt
Sep 18 09:36:05 <ppisati> ah
Sep 18 09:36:09 <ppisati> it's a 3.4.x
Sep 18 09:36:09 <ogra_> i have heard many reports that people on intel have seen it too btw
Sep 18 09:36:11 <infinity> cooloney: Yes, I know. :P
Sep 18 09:36:14 <ppisati> we have the fix in 3.5
Sep 18 09:36:22 <ogra_> it isnt arch specific but USB related
Sep 18 09:36:37 <cooloney> or just use #ifdef USB_NET SMSC95XX
Sep 18 09:36:43 <cooloney> like this kind of hacking
Sep 18 09:37:41 <cooloney> ppisati: what's kind of fix in 3.5?
Sep 18 09:44:49 <ppisati> 86b83747713545234c28e5347c6f0e5efb652332
Sep 18 09:44:53 <ppisati> HACK force min_free_kbytes set by proc to min 32K to workaround smsc problems
Sep 18 09:44:59 <ppisati> Signed-off-by: Andy Green <andy.green@linaro.org>
Sep 18 09:45:22 <ppisati> actually even this one
Sep 18 09:45:26 <ppisati> 89b16e837180b215b667affdd4227220827f8bb4
Sep 18 09:45:30 <ppisati> HACK force min_free_kbytes to 32K to workaround smsc problems
Sep 18 09:45:37 <ppisati> both are needed
Sep 18 09:45:51 <ppisati> the second one is a safe belt
Sep 18 09:45:56 <ppisati> the first one is the real one
Sep 18 09:53:08 <ogra_> ppisati, did you put the LP bug number for the flickering in the changelog ? (i didnt see it in teh pull request mail)
Sep 18 09:53:19 <ogra_> Laney, ^^^
Sep 18 09:54:40 <cooloney> ppisati: i didn't see that patch in the real code. can you see that code in the mm/page_alloc.c?
Sep 18 09:54:54 <cooloney> ppisati: although i can git show that patches you posted
Sep 18 09:58:43 <ppisati> ogra_: i didn't know there was a bug opened for it
Sep 18 09:59:41 <ogra_> bug 1045491
Sep 18 09:59:43 <ubot2> Launchpad bug 1045491 in linux-ti-omap4 "Moving mouse messes up the desktop" [High,Confirmed] https://launchpad.net/bugs/1045491
Sep 18 09:59:48 <ppisati> ah ok
Sep 18 09:59:49 <ppisati> ok
Sep 18 09:59:54 <ppisati> i'll close it manually
Sep 18 10:00:04 <ogra_> well, let me take care
Sep 18 10:00:16 <ogra_> i'll close it if it has been tested with the images :)
Sep 18 10:02:48 <ogra_> (if i dont, kate will hunt me down anyway)
Sep 18 11:34:36 <ppisati> ogra_: where is the template for preEnv.txt? or, how do i make it look as i like? or how do i make flash-kernel NOT replace its content at every run?
Sep 18 11:35:00 <ogra_> ppisati, /etc/default/flash-kernel
Sep 18 11:35:24 <ppisati> ogra_: ack
Sep 18 12:09:19 <sveinse> As a part of a custom/proprietary first-boot system, I need to run a series of disk partitioning operations prior to mounting the real root. What would be preferred: a dash script with the required support tools (sfdisk, awk, grep, etc.) or to pull in python into initramfs?
Sep 18 12:10:11 <ogra_> sveinse, have a look at the jasper-initramfs code
Sep 18 12:11:08 <sveinse> ogra_, voila. Thanks
Sep 18 12:11:33 <ogra_> specifically at the jasper_growroot script
Sep 18 12:11:41 * sveinse discoveres that the wheel has already been invented
Sep 18 12:12:06 <ogra_> and obsoleted too :)
Sep 18 12:12:20 <sveinse> I'm not there yet :P
Sep 18 12:39:21 <sveinse> Any of you people the author(s) of jasper-initramfs?
Sep 18 12:39:27 <ogra_> yes
Sep 18 12:39:41 <sveinse> Why do you reboot after resizing root?
Sep 18 12:39:42 <ogra_> any of me are :)
Sep 18 12:40:08 <ogra_> because i'm a coward :)
Sep 18 12:40:25 <ogra_> and want to make sure it actually worked before moving on with the install
Sep 18 12:40:49 <ogra_> its not actually necessary if you are sure its all fine
Sep 18 12:41:10 <ogra_> (i.e. if you dont have people with random SD cards as media)
Sep 18 12:41:57 <sveinse> its just that there any isn't more you can do with a reboot compared to just continue. But I get the point
Sep 18 12:42:32 <sveinse> Except that you know that the bootloader is still intact after altering the parttables
Sep 18 12:42:58 <ogra_> you do test if the UUID and partitioning is still  right and if the bootloader still does the right thing on a fresh boot
Sep 18 12:43:09 <sveinse> yup
Sep 18 12:43:47 <ogra_> jasper_growroot also isnt the only script there ... jasper modifies other things too and sets them up
Sep 18 12:43:50 <sveinse> i also noticed a loop around resize2fs. Doesn't it do enough in one iteration?
Sep 18 12:44:52 <sveinse> (not that I've deciphered the sed and char logic in there)
Sep 18 12:44:56 <ogra_> thats for plymouth
Sep 18 12:45:06 <sveinse> ah. of course
Sep 18 12:45:30 <ogra_> resize2fs runs through several iterations, the loop just picks up the output and dumps it to plymouth or the text console
Sep 18 12:45:57 <ogra_> its not resize2fs itself thats gets looped over ... only the output
Sep 18 12:46:08 <sveinse> you mean resize2fs does NOT run in loop, yes, I see that now
Sep 18 12:46:15 <ogra_> right
Sep 18 12:46:24 <ogra_> the while sits after the pipe
Sep 18 12:48:45 <sveinse> and I also found some nice sysctl settings for having root on sd-cards. Are there any more goodies like this I should be aware of?
Sep 18 12:49:38 <lilstevie> I need to start using that :p
Sep 18 12:50:05 <lilstevie> I often get users of my transformer images complaining about it just stopping when resize2fs is running
Sep 18 12:51:07 <ogra_> sveinse, the zram-conf package is helpful if your kernel supports it (speeds up swapping a lot)
Sep 18 12:51:42 <ogra_> and here on my ac100 i just experiment with a journal-less ext4 ... thats classes faster than any FS i have tried yet
Sep 18 12:52:03 <ogra_> though with that you should force an fsck on every boot
Sep 18 12:53:38 <ogra_> ogra@horus:~$ cat /etc/security/limits.d/90-stack.conf
Sep 18 12:53:39 <ogra_> * soft stack 256
Sep 18 12:53:52 <ogra_> thats also pretty helpful when being low on ram
Sep 18 12:54:37 <lilstevie> what is required for zram anyway
Sep 18 12:54:40 <ogra_> (default is 8k per thread stack, makes your apps use less ram (and since its soft they can just ask for more if needed))
Sep 18 12:54:58 <ogra_> lilstevie, that you enable it in your kernel config
Sep 18 12:54:59 <lilstevie> I have no idea if I included the right bits
Sep 18 12:55:02 <lilstevie> :p
Sep 18 12:55:21 <ogra_> i think there is only one option (or probably two)
Sep 18 12:55:36 <lilstevie> config_cramfs?
Sep 18 12:56:18 <lilstevie> I guess I should just try it lol
Sep 18 12:56:34 <ogra_> no, thats a filesystem
Sep 18 12:56:46 <lilstevie> ah right
Sep 18 12:56:57 <ogra_> search for zram
Sep 18 12:57:10 <ogra_> (hit / in menuconfig)
Sep 18 12:57:22 <lilstevie> yeah I know how to search :p
Sep 18 12:58:07 <lilstevie> ok yeah CONFIG_ZRAM
Sep 18 12:58:11 <lilstevie> and it is enabled
Sep 18 12:58:36 <ogra_> great
Sep 18 13:00:55 <lilstevie> I have double the ram of the ac100, but IMO 1GB is far from enough
Sep 18 13:02:43 <ogra_> yeah, imho zram should become a default as staging swap
Sep 18 13:03:02 <ogra_> so that you first swap into compressed ram before actually writing to disk
Sep 18 13:03:26 <lilstevie> yeah
Sep 18 13:03:41 <lilstevie> I will have to give it a go, see how much of a performance increase it gets
Sep 18 13:05:21 <ogra_> you will only note it when swapping
Sep 18 13:05:40 <lilstevie> so the moment I open chromium then :p
Sep 18 13:06:35 <ogra_> yeah, firefox uses lots less ram in its recent versions ...
Sep 18 13:26:03 <sveinse> What's the deal with CHS geometry on omap3 and bootloader? I haven't been able to find any other combination other than H=255, S=63 that works. But I see H=128, S=32 is used in jasper
Sep 18 13:26:17 <sveinse> so I guess that also works
Sep 18 13:26:44 <ogra_> the ROM needs to find the first stage loader at a certain point on a vfat to boot
Sep 18 13:27:06 <sveinse> (I have to admit I'm confused about CHS vs LBA and how they interact)
Sep 18 13:27:07 <ogra_> thats omap specific
Sep 18 13:28:05 <ogra_> i doubt you need CHS on tegra
Sep 18 13:30:10 <lilstevie> ogra_, tegra scans until it finds the bct
Sep 18 13:30:38 <ogra_> right
Sep 18 13:30:45 <sveinse> from what I understand, there partition table contains both CHS and LBA fields. And Linux uses only LBA, right.
Sep 18 13:31:30 <sveinse> So given that you're able to boot, when sfdisk complains about partitions not on cylinder boundaries is just a warning, but it does not affect anything
Sep 18 13:49:30 <sveinse> ogra_, A proposal: You dont strictly need to use fdisk to determine the size of the disk. sfdisk can do that as well. So you can spare a binary in initramfs
Sep 18 13:49:54 <ogra_> sveinse, well, jasper is pretty much dead beef
Sep 18 13:50:05 <sveinse> ok, just saying
Sep 18 13:50:23 <ogra_> we only keep it in the archive in case we will actually spin preinstalleds once again
Sep 18 13:50:51 <ogra_> over time it will vanish from the arhcive though, i wont put any work into it
Sep 18 13:51:05 <ogra_> and yeah, i know that sfdisk can do this
Sep 18 14:40:28 * ogra_ looks at bug 1044709 and wonders what got infinity the honor to be the new unity slave
Sep 18 14:40:29 <ubot2> Launchpad bug 1044709 in unity/6.0 "unity-6.4.0 from quantal-proposed crashed with SIGSEGV on omap4" [Critical,Fix committed] https://launchpad.net/bugs/1044709
Sep 18 14:41:35 <ogra_> oh you actually commited that to their tree
Sep 18 20:40:23 <jimerickson> omap4+armhf: it works. enough said.
Sep 18 20:44:33 <GrueMaster> jimerickson: Erm, when didn't it?
Sep 18 20:54:56 <jimerickson> was have some flashing screen problems with the last kernel. but this latest update has resolved it.
Sep 18 21:03:49 <GrueMaster> Ah.
Sep 18 21:14:43 <Deformati> Anyone here know about performance counters?
Sep 18 21:46:21 <GrueMaster> I do, sort of.
Sep 18 21:46:29 <GrueMaster> At least from the Pentium Pro days.
Sep 18 22:54:41 <Deformati> GrueMaster, For some reason on omap4 devices the performance counters are disabled and there doesn't seem to be documentation for enabling them.
Sep 18 22:54:52 <Deformati> So I suppose pentium pro isn't relevant.
Sep 18 22:55:11 <GrueMaster> Yea, propably not.  :P
Sep 18 22:55:48 <GrueMaster> You might ask on #pandaboard though.  They can probably help, if it is harder than simply rebuilding the kernel.
Sep 18 23:00:02 <GrueMaster> It should be fairly straight-forward though.  Looks like it can be done by downloading the kernel source package for the kernel you are using, then modify the config to enable perf counters.
Sep 18 23:00:06 <GrueMaster> And rebuild.
Sep 18 23:00:49 <GrueMaster> Documentation is at http://help.ubuntu.com/community/Kernel/Compile
Sep 18 23:27:43 <Deformati> GrueMaster, I tried that.
Sep 18 23:28:28 <Deformati> There is a bug report on launchpad.
Sep 18 23:28:31 <GrueMaster> And the new custom kernel didn't work?
Sep 18 23:28:43 <Deformati> The existing kernel has it enabled.
Sep 18 23:28:48 <GrueMaster> ah, ok.
Sep 18 23:29:06 <Deformati> But warmcat says on the launchpad bug report that it is intentionally disabled, i cannot find what they disabled or where.
Sep 18 23:29:14 <Deformati> I just get errors every time I try to use the PMU.
Sep 18 23:32:20 <GrueMaster> Sounds like possibly a #linaro or #pandaboard issue.
Sep 18 23:32:31 <Deformati> Ok.
Sep 18 23:33:09 <GrueMaster> I seem to remember some discussion on this a while back (like 11.10 timeframe).
**** ENDING LOGGING AT Wed Sep 19 02:59:58 2012