[11:27] <Mithrandir> BenC: did you drop the planned kernel upload?
[11:39] <poningru> hi anyone alive?
[11:39] <TheMuso> c
[11:40] <poningru> is there a possibility of getting this module uploaded http://www.promise.com/support/download/download2_eng.asp?productId=139&category=driver&os=3&go=GO#
[11:40] <poningru> TheMuso: I am having the hardest time getting it built
[11:40] <poningru> cause they have a weird ass makefile
[11:41] <TheMuso> poningru: Sorry dude, didn't mean to do that. KVM and speakup like doing funny things with each other
[11:41] <poningru> blargh?
[11:43] <mjg59> Why is sata_promise.ko not adequate?
[11:46] <mjg59> poningru: Given that there's an in-kernel driver for the hardware already, it's unlikely that we'll include it. Why do you want it?
[11:48] <poningru> mjg59: that module is apparently causing troubles
[11:48] <mjg59> poningru: Which one, and where?
[11:48] <poningru> sorry sata_promise.ko
[11:49] <mjg59> Got a bug number?
[11:49] <poningru> nope about to file one ;)
[11:49] <mjg59> Ok, based on reading that driver, there's no real chance of us including it
[11:49] <poningru> hmm k
[11:49] <mjg59> It seems to hook straight into the scsi layer
[11:50] <mjg59> Rather than using libata
[11:50] <poningru> hmm ic
[11:50] <poningru> http://lwn.net/Articles/122289/ thats the patch that created support for this device back in 2.6.11
[11:50] <poningru> atleast in this hardware config sata_promise.ko is failing
[11:51] <poningru> so I have to test out couple of other hardware configs to make sure
[11:51] <poningru> the big prob is this box is an old ass dell optiplexx
[11:51] <mjg59> In what way is it failing?
[11:52] <poningru> just wont boot even if just the pci card is plugged in and the hdd is plugged directly into the mobo's ide controller
[11:53] <mjg59> Won't boot in what way?
[11:54] <poningru> yeah let me copy the logs hold on
[11:54] <mjg59> Thanks
[12:00] <poningru> err do you know which logs I should look under?
[12:00] <poningru> /var/log/dmesg* hasnt turned up anything
[12:01] <poningru> mjg59^^
[12:02] <mjg59> poningru: If the system is alive when it stops booting, /var/log/kern.log is probably the best bet
[12:02] <mjg59> Otherwise you're going to have to copy them down by hand
[12:03] <poningru> frack
[12:03] <poningru> ok let me do that then
[12:15] <poningru> mjg59: ok well the last thing is this: udevd-event[1772] :run-program:'/sbin/modprobe' abnormal exit
[12:16] <poningru> right before this there is a huge stack trace
[12:16] <poningru> mentioning everything from libata and sata_promise to mem pointers
[12:17] <poningru> hmm you know what hold on let me see if I can put it on my camera phone
[12:28] <poningru> mjg59: well it drops me down to the busybox cli after a while
[12:28] <poningru> is there anything I can run in this that will show anything useful?
[12:42] <poningru> mjg59: http://img220.imageshack.us/img220/9712/0130070621wp0.jpg
[12:43] <poningru> http://img219.imageshack.us/img219/1164/0130070622rb0.jpg
[12:44] <poningru> let me know if you want to see more screencaps
[12:49] <mjg59> poningru: Which version of Ubuntu is this?
[12:49] <poningru> feisty
[12:49] <mjg59> Ok
[12:49] <poningru> grabbed the cdimage like 2 days ago
[12:50] <mjg59> I suspect that it's related to https://launchpad.net/ubuntu/+source/linux-source-2.6.20/+bug/78224
[12:51] <poningru> ah hmm
[12:51] <poningru> well mine is a bit different
[12:51] <mjg59> If you could boot without the splash and quiet options, and add vga=ask then choose the 80x50 mode, you ought to be able to get more of it
[12:51] <poningru> ooh ok
[12:51] <mjg59> Alternatively, if it drops you to busybox, does shift+pageup work?
[12:51] <poningru> doh
[12:51] <poningru> let me try both
[12:52] <poningru> hold on
[12:54] <poningru> the vga option and shift+pageup worked
[12:54] <poningru> let me grab some better screenies then
[12:55] <poningru> yeah nm its almost word for word what that bug shows
[12:55] <poningru> except ofcourse the other hardware modules
[12:56] <mjg59> The important bit is what the "EIP is at" line says
[12:59] <poningru> EIP is at pdc_port_start+0x96/0xc0 [sata promise] 
[01:00] <mjg59> Ok
[01:01] <poningru> btw this bug is in sarge, etch and edgy 
[01:01] <mjg59> Heh
[01:01] <mjg59> So probably upstream
[01:01] <poningru> yeah
[01:02] <poningru> I was going to try a vanilla but never got around to it
[01:02] <mjg59> Interesting. It's not entirely clear where it's failing.
[01:02] <mjg59> Anyway. Can you follow up to that bug, stick the "EIP is at" line there as well, and mention the precise kernel version you have installed?
[01:02] <poningru> yep
[01:03] <poningru> btw thanks for helping with this
[01:03] <poningru> I was confoozled for a while with this
[01:03] <mjg59> No problem
[01:03] <mjg59> We'l try to get that fixed
[01:03] <poningru> thanks :)
[01:42] <poningru> done 
[02:27] <zul> hiho
[02:32] <Mithrandir> zul: did you seem my prod about xen-3.0 FTBFS-ing?
[02:34] <zul> when?
[02:35] <zul> the xentop issue in -4?
[02:35] <zul> yeah fixed in -5 :)
[02:42] <Mithrandir> good
[02:52] <BenC> Mithrandir: can I still upload kernel now?
[02:52] <BenC> fell asleep last night while builds were finishing
[02:53] <BenC> but it's ready
[02:53] <kylem> BenC, i think feisty is frozen atm.
[02:53] <BenC> kylem: that's why I'm asking the freezer for an exception :)
[02:55] <kylem> hehe
[02:55] <Mithrandir> BenC: ok, go ahead.
[03:12] <BenC> Mithrandir: bombs away
[03:16] <Mithrandir> thanks
[04:58] <mjg59> BenC: Why do we build prism2_cs?
[04:58] <BenC> mjg59: For some people, it works better than the alternative
[04:59] <mjg59> BenC: Here, it doesn't work at all
[04:59] <mjg59> Despite me having an entirely normal prism 2 card...
[05:00] <BenC> does hostap_cs work?
[05:00] <mjg59> Yup
[05:04] <BenC> is there conflicting PCI id's in them? If so, I can disable them in prism2_cs
[05:04] <mjg59> It's not PCI
[05:04] <BenC> or you can just wait on the driver-manager to select the one you want :)
[05:04] <BenC> I thought PCMCIA used PCI id's of sorts
[05:04] <mjg59> But yes, there are going to be conflicting IDs - prism2_cs and hostap_cs should drive the same hardware
[05:04] <BenC> I thougth I cleaned out the ones in prism2 that hostap had
[05:04] <mjg59> Similar, but PCMCIA is effectively ISA
[05:06] <mjg59> BenC: I suspect it's matching on some subdevice thing
[05:07] <mjg59> PCMCIA_DEVICE_PROD_ID123("D",  "Link DWL-650 11Mbps WLAN Card",  "Versi\
[05:07] <mjg59> on 01.02", 0x71b18589, 0xb6f1b0ab, 0x4b74baa0), // D-Link DWL-650 11Mbps 802.11\
[05:07] <mjg59> b WLAN Card
[05:07] <mjg59> That's binding from prism2_cs
[05:08] <mjg59> Are you sure there were people that found prism2_cs to work better? prism2_pci I could believe.
[05:08] <mjg59> Oh, it's loaded orinoco_cs as well
[05:08] <mjg59> This is something of a mess...
[05:09] <BenC> Yeah, for most of edgy, I had prism2_cs disabled because it wouldn't compile in .17, and sufficient numbers of folks complained about hostap_cs not working to get me to fix it
[05:09] <mjg59> Got a bug number?
[05:09] <BenC> let me check the change log
[05:09] <mjg59> Actually, it's using orinoco_cs
[05:12] <BenC> so it's not prism2_cs, or it's not hostap_cs?
[05:13] <BenC> I found what I think was the bug, but it mainly mentions prism2_usb
[05:13] <BenC> at least it was the bug that prompted me to fix things
[05:13] <mjg59> By default, prism2_cs binds and nothing works
[05:13] <mjg59> Yeah, prism2_usb is a reasonable choice
[05:13] <mjg59> I don't think prism2_cs is
[05:14] <BenC> wasn' the first mention of it (I rejected any bugs dealing with it before that)
[05:32] <fabbione> kylem: 81782 is an easy one.. tho i don't think we will rebuild d-i or cdimages for it
[05:33] <fabbione> kylem: might be still good to get the kernel in case cjwatson decides to upload a new d-i (at least we can get netboot/netinstall to work)
[05:35] <Mithrandir> we'll need a new d-i once the new kernel is in, yes.
[05:38] <fabbione> Mithrandir: yeah but that's edgy...
[05:46] <BenC> mjg59: Disabled prism2_cs in git
[05:46] <mjg59> BenC: Ta
[05:47] <Mithrandir> BenC: lrm accepted
[05:48] <BenC> Mithrandir: I'll wait to make sure linux-image and lrm build ok before uploading linux-meta
[05:49] <Mithrandir> BenC: thanks.  linux-image will need NEW processing and I'm leaving for the movies in about an hour, so if they're done before I'm back, either get another archive admin to process them or wait for me to do it (at about 21 UTC, I suspect)
[05:49] <BenC> Mithrandir: ok, thanks
[05:49] <BenC> Mithrandir: how's the weather there since we left?
[05:50] <Mithrandir> BenC: slightly warmer and a bit more snow.  So it's nice.
[05:50] <BenC> Miss it already...Oslo was a nice place
[05:51] <Mithrandir> it's a lovely place. :-)
[06:38] <zul> BenC: yeah kernel-package is crap
[09:12] <pinskian> hi
[10:56] <Mithrandir> kylem: you don't happen to know anything about the LSI 1068 RAID controller in Dapper, do you?  It's supposedly supported, but not with hardware raid activated.
[10:56] <Mithrandir> it's also supposedly fixed in 2.6.16.
[10:57] <kylem> it sounds familiar.
[10:59] <Mithrandir> I don't have a machine yet, but I'd be very interested in getting it fixed in dapper, if that'd not too big of a backport.
[11:00] <kylem> i'm struggling to remember what the problem is. does it use the cciss driver?
[11:01] <Mithrandir> https://launchpad.net/ubuntu/+source/linux-source-2.6.15/+bug/37452 is the same bug, it seems
[11:02] <Mithrandir> so yes, it seems like you've committed a fix for it.
[11:03] <kylem> woo. go me. :)
[11:03] <kylem> oh. no, it doesn't work right yet. 
[11:03] <kylem> i don't have hw to test, and no one has identified which patches fix it.
[11:03] <Mithrandir> I'll get hardware to test it on in a couple of weeks, then
[11:03] <kylem> so i guessed, and wrong.
[11:04] <Mithrandir> well, as long as there exists _some_ fix for it, I'll be able to find out which.