[01:33] <sn9> 2.6.15-13 segfaults on boot. anybody interested in details?
[01:35] <cjb> I'm not one of the developers, but I haven't heard of a bug like that; I think you should file a bug report.  Did previous kernels work?
[01:36] <sn9> it's mostly a breezy system with a few packages from dapper. 2.6.15-9 worked considerably better than the breezy kernel
[01:38] <sn9> someone evidently borked the keywest i2c driver after that
[01:38] <sn9> that's what segfaults
[01:38] <cjb> .. or it's your mix or kernel/userland.
[01:39] <cjb> s/or/of/
[01:39] <sn9> i have the dapper udev
[01:41] <sn9> the system still boots like nothing's wrong, but things that rely on i2c are borked, like alsa
[01:42] <sn9> dmesg is what revealed the segfault
[03:32] <mxpxpod> BenC: ping
[03:54] <fabbione> BenC: no it was firewire and another one that i can't remember
[04:05] <psusi> fabbione, ahh, you're back from vacation I see?  enjoyed it I hope?  before you left you asked why I reassigned the dmraid bug to you instead of adam conrad?  he said he had no idea why it was assigned to him and that I should switch it to you
[04:09] <fabbione> psusi: ok, in anycase since the package is in universe you can just ask MOTU to look/upload it
[04:09] <fabbione> there is no need of me doing it
[04:09] <psusi> ok....  right now i'm working on properly udebifying it and integrating it into partman... 
[04:10] <fabbione> i am really of no use for it as it is
[04:10] <fabbione> i still didn't get to look at the kernel problem i am having
[04:10] <psusi> kernel problem eh?
[04:10] <fabbione> = my system doesn't see the disks
[04:11] <psusi> you aren't using both sata and pata are you?
[04:11] <fabbione> i only have pata
[04:11] <psusi> it seems current kernels don't like mixing the two
[04:11] <psusi> ohh
[04:11] <fabbione> nothing like that
[04:11] <fabbione> the controller module is loaded
[04:11] <psusi> wow, so it just doesn't see the disks at all eh?
[04:11] <fabbione> but it sees no disks
[04:11] <psusi> which controller is this?
[04:11] <psusi> promise?
[04:12] <fabbione> 0000:00:08.0 RAID bus controller: Promise Technology, Inc. PDC20378 (FastTrak 378/SATA 378) (rev 02)
[04:12] <psusi> I believe the via module has to program the controller to run in non raid mode for the disks to show up normally, then it hands off to libata... maybe the promise module doesn't do that part?
[04:12] <psusi> do the disks show up if you tell the bios to use non raid mode?
[04:13] <psusi> ok... so that is the controller that supports both pata and sata... but you only have pata drives plugged in?
[04:13] <fabbione> i didn't check anything yet
[04:13] <fabbione> -ENOTIME
[04:13] <fabbione> and -ELOWINTEREST
[04:13] <psusi> I've seen a lot of problem reports about that controller and pata in recent kernels, but I had thought it was from trying to use both at once
[04:13] <psusi> hehe
[04:15] <psusi> earlier on in the bug history, you mentioned that it should eventually be assigned to installer-integration I think it was?   is it time for that now?
[04:16] <fabbione> psusi: not without udeb and main inclusion report
[04:16] <fabbione> you will need to test it first
[04:16] <fabbione> + you are doing the installer integration
[04:17] <fabbione> nobody knows better than you what you are doing
[04:17] <psusi> ok.... I'll keep working on that then.... I think the udeb is ready... just need to modify partman to use it correctly from the setup cd.... espresso install from livecd should be easy though once that's working
[04:17] <psusi> I guess I should start working on a main inclusion report then too eh?
[04:18] <fabbione> psusi: the main inclusion report is a wiki page.. get teh code for partman working first
[04:18] <psusi> ok....
[04:18] <psusi> yea, I've seen the page
[05:16] <dilinger> dapper + cryptdisks (dm-crypt) + 3ware + raid5 (swraid) + 700gb data is giving me such headaches :/
[05:17] <dilinger> getting random unreproducible read errors
[05:18] <psusi> poke at it with smartctl?
[05:18] <dilinger> yea, already did
[05:18] <dilinger> no smart errors, lm-sensors think everything is fine
[05:19] <psusi> was getting some odd access denied errors in my daily backups on a server at work the last few days... had smartctl make it do it's long internal test... had a few completely unreadable sectors, and a TON of recoverable errors...
[05:19] <fabbione> dilinger: meh dude....
[05:20] <dilinger> psusi: also, rechecking is valid
[05:20] <dilinger> psusi: ie, flac -t foo.flac will fail the first time, succeed the second time
[05:20] <psusi> ohh, that's nice
[05:20] <psusi> anything interesting in dmesg?
[05:20] <dilinger> but it's about 3 flac files out of about 700GB
[05:20] <dilinger> nope
[05:20] <dilinger> kernel doesn't see any errors at all
[05:21] <dilinger> i've only done smart short tests, though
[05:21] <psusi> my via controller gets timeouts after I resume from suspend... it seems it needs longer than the kernel is willing to wait to wake back up
[05:21] <dilinger> i guess i can schedule some long tests
[05:21] <psusi> increased the timeout duration fixed it
[05:21] <psusi> do the long one, yea... short doesn't seem to do much
[05:21] <dilinger> this would be a lot easier if i could reproduce it easier.. as it stands, it takes about 2 days to run the full test
[05:22] <psusi> also look at smartctl -A... see if it reports a number of recoverable errors
[05:22] <psusi> why does it take 2 days?
[05:22] <psusi> takes me ~30 mins on my 10,000 rpm drives here... took ~85 mins on the big 7200 rpm drives on the server at work
[05:22] <dilinger> 'cause it's 700gb of data :)
[05:22] <psusi> spread on multiple drives
[05:23] <psusi> all the drives can be doing their tests at once ;)
[05:42] <hub> hi
[05:42] <hub> I have a huge kernel regression
[05:42] <hub>  /etc/init.d/udev fails on startup
[05:42] <hub> udevplug seems to be failing
[05:43] <hub> any clue to help debugging?
[05:43] <hub> (runnind dapper)
[08:13] <fabbione> hub: #ubuntu-boot and the fact that udev fails is not necessarely a kernel fault
[08:13] <fabbione> garbage_: can you please fix your connection? you are bouncing every 5/10 minutes
[08:40] <oceandead> I'm sorry if I'm in the wrong place but I wasn't sure where to go with this (I didn't want to send email to a list - fearing this topic has already been brought up). Also, I haven't been able to participate in laptop-testing since Breezy Colony3 (will again be testing soon), so I'm also not sure if it's already been included or not. 
[08:40] <oceandead> Anyway, I was wondering what the status of inclusion of rtl8180-sa2400 wifi chipset drivers might be.
[08:44] <oceandead> I checked all the mailing list archives I could find but saw nothing about it. Anyway, the sourceforge page can be found here - http://rtl8180-sa2400.sourceforge.net - They (rtl8180-0.21) compiled fine for me under Hoary. I had problems compiling the CVS version under Breezy but luckily a user who successfully compiled them created a .deb package of the compiled drivers for i386. 
[08:47] <fabbione> oceandead: if it is not in dapper, open a bug in malone
[08:47] <oceandead> fabbione,  will do 
[08:47] <oceandead> thanks 
[08:47] <oceandead> and thank you all for all of your hard work 
[08:49] <oceandead> be well! 
[02:09] <hub> crap. somebody replied to me and I didn't get it
[02:09] <hub> hi all btw
 hub: #ubuntu-boot and the fact that udev fails is not necessarely a kernel fault
[02:09] <hub> fabbione: okay, thanks
[03:07] <mxpxpod> BenC: how's it coming?
[03:20] <zul> morning
[06:28] <zul> http://www.thesmokinggun.com/archive/0123061mugpair1.html
[06:31] <fabbione> BenC: ping?
[06:31] <fabbione> hey zul
[06:31] <BenC> fabbione: pong
[06:32] <fabbione> BenC: i did test the new amd64-k8 on UP
[06:32] <fabbione> i still have performance problems when i am doing disk I/O
[06:32] <zul> hey fabbione 
[06:32] <fabbione> it's easy to reproduce when running dselect
[06:32] <BenC> as compared to pre-SMP versions?
[06:32] <zul> BenC: i got some patches for you later..
[06:32] <BenC> zul: ok
[06:33] <fabbione> BenC: yeps
[06:33] <fabbione> what tests can i do before i leave tomorrow for uk?
[06:33] <BenC> fabbione: can you recompiled -13 with CONFIG_SMP=n, and give me some numbers?
[06:33] <fabbione> what numbers do you want?
[06:33] <BenC> like bonnie or something
[06:33] <fabbione> i can see a lot of interrupts
[06:34] <fabbione> really an excess compared to .12
[06:34] <fabbione> or .15 pre-smp
[06:34] <fabbione> and the sata_promise doesn't see my disks, but i had no time to bisec that 
[06:34] <BenC> are they pata disks, or real sata disks?
[06:35] <BenC> current git will recognize the pata disks now (tested now that I have a promise sata controller)
[06:35] <fabbione> pata disk on fasttrack that has a sata raid
[06:35] <fabbione> it's a regression from .12
[06:35] <fabbione> ah ok
[06:35] <BenC> should work with current git
[06:35] <fabbione> will check that with the next upload
[06:35] <fabbione> 0000:00:08.0 RAID bus controller: Promise Technology, Inc. PDC20378 (FastTrak 378/SATA 378) (rev 02)
[06:35] <fabbione> this onwe
[06:35] <zul> brb
[06:36] <fabbione> BenC: what was the option to disable the SMP2UP and let the kernel run without it?
[06:37] <BenC> smp-alt-disable
[06:37] <fabbione> ok
[06:37] <fabbione> i will see if i can manage today
[06:38] <fabbione> otherwise it will have to wait when we are back from uk
[06:42] <zul> bleah..
[07:18] <mxpxpod> BenC: how goes the suspend
[07:39] <mxpxpod> BenC: ping