**** BEGIN LOGGING AT Sat May 10 02:59:57 2008 May 10 09:12:38 rwhitby: hi May 10 11:09:11 booting debian with a 2.6.24 kernel on NAS100d, my USB is dead May 10 11:09:31 one of the host controller complains about PCI error and disables itself May 10 11:09:54 as a result, only bus 1 (which I guess is internal) is there May 10 11:10:21 does someone have seen this ? or better resolved this issue ? May 10 11:10:48 could this be related to the dma_free_coherent warning I see in syslog ? May 10 11:12:04 osmeest: is there anything connected to that USB port? May 10 11:12:13 yes May 10 11:12:21 an HP AiO printer May 10 11:12:39 hmm May 10 11:13:05 if you disconnect it do you also get the PCI error? May 10 11:13:15 that's one of my goals with the NAS is to have it as a printer server May 10 11:13:49 sure, but just to see if the error might be caused by the printer being connected... May 10 11:15:10 looking through the kern.log, yes, it does that without anything connected to it May 10 11:15:39 in fact I booted it without the printer, then connected it (hoping to see the detection) but nothing happened May 10 11:17:06 ok, that's interesting. Unfortunately I just know the NSLU2 where I though I had similar problems like you, but that's probably not the case then... May 10 11:18:26 uhci_hcd 0000:00:03.1: host system error, PCI problems? May 10 11:18:26 host controller halted, very bad! May 10 11:18:26 HC died; cleaning up May 10 11:19:04 uhci_hcd 0000:00:03.0: host system error, PCI problems? May 10 11:19:04 uhci_hcd 0000:00:03.0: host controller halted, very bad! May 10 11:19:04 uhci_hcd 0000:00:03.0: host controller halted, very bad! May 10 11:19:19 uhci_hcd 0000:00:03.0: HC died; cleaning up May 10 11:22:29 did it work with an older kernel version? May 10 11:22:57 I don't know... May 10 11:23:09 ah, ok May 10 11:23:42 with 2.6.18, the hard disk (pata_artop) does not exist May 10 11:24:00 so I don't have hard disk access (bad for Debian) May 10 11:25:57 RobNC: the >64MB kernel bug is independent of the distribution - it appears in all nslu2 firmware May 10 11:27:53 osmeest: I doubt anyone other than you has tested 2.6.24 on nas100d May 10 11:29:12 rwhitby: well, it's barely unusable May 10 11:29:36 rwhitby: DMA problems are pausing the system every now and then May 10 11:30:00 rwhitby: my ssh session appears like dead then revives automagically May 10 11:30:14 in 2.6.18, the HDD is using aec62xx ? May 10 11:30:31 (that's what I saw in the bootlog of the original firmware) May 10 11:30:34 does the same thing occur with the slugos/be 2.6.24 kernel and distro? May 10 11:30:50 I don't think so but can't say for sure... May 10 11:31:10 I'll make a try after lunch (it 1:30PM here in Belgium) May 10 11:31:44 rwhitby: which kernel would you advise me then ? the 2.6.18 from edgy ? or something else ? May 10 11:31:47 I've tested kernel boot for 2.6.24 on nas100d, but my development nas100d doesn't have a hard disk attached, so I have no idea whether the disk controller works May 10 11:31:57 my other nas100d in production runs Debian etch May 10 11:34:47 what I can say for sure is that I could debootstrap etch from SlugOS 4.8 without apparent problems May 10 11:35:51 (note that I used slugos/LE, not BE) May 10 11:36:06 can this have a bad influence ? May 10 11:41:06 no May 10 11:41:39 the kernels should be quite similar between slugos and debian - we share much the same patches May 10 12:18:01 OK... trying slugos/le in a minute then May 10 12:25:05 slugos uses 2.6.21 (I thought it was already at 2.6.24) May 10 12:25:28 and there is no error on USB May 10 12:28:38 slugos builds from latest monotone repository use 2.6.24.4 May 10 12:28:51 I'm using 4.8 beta May 10 12:28:55 (it's at 4.10-alpha at the moment) May 10 12:29:01 4.8-beta used 2.6.21, yes. May 10 12:29:48 DMA seems stable (yesterday it complained a lot while doing an fsck on the 200GB partition, I try with slugos and it doesn't complain at all) May 10 12:32:47 can I use 2.6.22 from lenny in etch ? May 10 12:36:35 I have no idea. Best to ask that in #debian-arm on OFTC.net May 10 12:37:14 which module do you use in etch ? I can't find aec62xx.ko in lib/modules May 10 12:38:08 I'll have to check May 10 12:40:28 2.6.21-2-ixp4xx May 10 12:40:32 pata_artop May 10 12:40:47 Debian GNU/Linux 4.0 May 10 12:40:58 where did you get that kernel ? May 10 12:41:12 that's the same as from slug os but it's not on debian repos May 10 12:41:35 I may well have built it myself. May 10 12:41:56 I can put it somewhere if you want it May 10 12:42:09 could be useful :) May 10 12:42:29 that's probably why in the bootstrap page you were talking about the buildhost to copy the kernel from May 10 12:43:14 which page? May 10 12:43:38 Debian/Bootstrap May 10 12:43:53 near the end, you mention that a recent kernel should be installed May 10 12:43:58 and you scp it from buildhost May 10 12:44:37 the aec62xx driver from 2.6.18 is nowhere to be found in debian it seems May 10 12:45:36 right - the whole pata/sata subsystem had a major change between 2.6.18 and 2.6.21 May 10 12:46:51 yep... all related to libata (in the new form) and all hdds becoming sdX instead of hdX May 10 12:47:12 yep May 10 12:48:59 the fsck from slugos went without a hitch May 10 12:49:27 and usb is working May 10 12:49:51 I'll try the 2.6.22 from lenny (cannot break anything anyway) May 10 12:49:51 osmeest: 91f245a39fd7e999b46b4c7365835c09 linux-image-2.6.21-2-ixp4xx_2.6.21-6_arm.deb May 10 12:50:18 osmeest: let me know when you've got it so I can delete it from the temp location I /msg'd you May 10 12:51:30 if that kernel is mandatory for NAS100d (if not for other platforms), could be useful to drop in a public area that I can link to from the wiki page May 10 12:52:25 can't do that. I don't have the source readily available any more. May 10 12:52:52 hmm... kind of a dead end there... May 10 12:53:17 without this kernel, no one can really install debian on the nas then May 10 12:53:51 the latest slugos kernel, plus bootstrap, doesn't work? May 10 12:54:20 probably not because it's in the debian side that there is something missing May 10 12:54:38 how could it be something outside the kernel? May 10 12:54:39 for some reason, aec62xx has been removed from debian kernels in 2.6.18 May 10 12:54:56 it was probably never there May 10 12:55:05 nas100d has never been officially supported May 10 12:55:30 pityfully May 10 12:55:33 if 2.6.18 worked, I wouldn't have had to build my own 2.6.21 kernel May 10 12:55:48 "pityfully"? what does that mean? May 10 12:56:17 it sounds a bit ungrateful ... May 10 12:57:09 probably my bad english: I mean it's a pity that NAS100d is not supported officially May 10 12:57:22 (nas100d has never been officially supported by debian. it's supported by nslu2-linux for slugos) May 10 12:58:06 the only reason why it's not supported officially is cause there is probably less than 5 people in the world trying to run debian on a nas100d. May 10 12:58:19 I suppose this is not going to change as NAS100d is out of market May 10 12:58:30 not unless you change it :-) May 10 12:58:49 I won't be upgrading my nas100d any time soon - all it does is act as a bluetooth gateway. May 10 12:59:14 If I reinstalled it, I'd put openwrt on it instead. May 10 12:59:40 it's supported on the nas ? May 10 13:00:48 there is a configuration for it, and i've booted it on a nas100d without a hard disk. May 10 13:03:09 osmeest: BTW, I would have built that debian kernel using the nslu2-linux debian svn repository. May 10 13:03:22 http://svn.nslu2-linux.org/svnroot/debian/trunk/ May 10 13:03:43 what I can't guarantee is that it would still build from that today. May 10 13:04:04 that's why I can't put the binary anywhere public - it would be a gpl violation. May 10 13:05:33 I see May 10 13:07:34 I never build anything manually - it's always built from a makefile in a public repo. May 10 13:08:02 so there's a good chance it would work if you can replicate the environment I had when I built it (which was probably debian etch or testing at that time) May 10 13:09:25 where did your cross tool chain come from ? slugos-image build ? or some packaged tool chain ? May 10 13:09:43 http://svn.nslu2-linux.org/svnroot/debian/trunk/toolchain/ May 10 13:10:12 everything is in the Makefiles May 10 13:10:31 cool :) May 10 13:11:49 you just need to remember which repo, which makefile :) May 10 13:12:37 about flash-kernel, I've seen that there are new versions in the pipeline (1.7 and 1.8 with 2 new platforms supported) May 10 13:12:50 osas: http://trac.nslu2-linux.org/ is the place to start May 10 13:12:54 (for which repo) May 10 13:13:05 adding the NAS would not be too complicated May 10 13:13:57 but the other platform do not seem to rely on a secondary bootloader (it adds some prefix to the kernel - I suppose for the endianness - or is it for sercomm ?) May 10 13:14:21 osmeest: nas100d has crc enabled in redboot - which makes it a bit harder to flash partitions from userspace May 10 13:14:39 prefix is for endianness and machine id May 10 13:15:13 the nas is really different from the others then :-/ May 10 13:15:22 Look in http://svn.nslu2-linux.org/svnroot/kernel/trunk/Makefile to see the prefixes. May 10 13:15:43 Not that different - if you flash the kernel and rootfs from the redboot console there is no problem. May 10 13:16:02 indeed (that's what I've done until now) May 10 13:16:36 ah... I've a BUG in dma_free_coherent with slugos as well May 10 13:16:43 was just longer to come May 10 13:39:21 with your 2.6.21 kernel, everything is OK (USB works fine and HDD seems OK as well) May 10 13:40:13 problem solved :-) May 10 13:47:31 yep, thx a lot May 10 13:47:44 I'll try to rebuild the kernel from the repo May 10 13:48:01 if it works, I'll update the wiki page about it May 10 14:30:13 anyone available to try and help me troubleshoot a problem? May 10 14:33:20 Is the problem "world peace" ? May 10 14:33:32 (or the lack thereof) May 10 14:34:21 (In IRC, it's customary to just state the problem first, rather than asking if there is anyone here to help you on a problem that they don't know the details about yet) May 10 14:34:48 I figure if we can knock this one out I can fix world peace on sunday - first I want my data back May 10 14:36:10 My harddive is no longer be recognized as formated. I'm running V2.3R63-uNSLUng-6.8-beta with a drive only on the disk2 port. May 10 14:36:32 Ah, did you post this to the mailing list today? May 10 14:36:39 yep May 10 14:36:46 My condolences. May 10 14:37:03 But seriously, does your Linux box not see anything in /proc/partitions when you plug in the drive? May 10 14:39:44 doh May 10 14:40:18 Does your Linux box not see anything in /proc/partitions when you plug in the drive? May 10 14:41:00 First caveat is I'm not a real strong linux user. I installed the drive into my box yesterday and restarted ubuntu. The drive didn't show up at all. I'm wondering if I need to manually mount it. I tried attaching it with just a USB cable and physically installing it in my PC. May 10 14:42:13 when you say "didn't show up" do you mean on the GUI, or didn't show up in dmesg or /proc/partitions? May 10 14:42:21 I was somewhat hopeful last night when I ran across this "http://www.nslu2-linux.org/wiki/HowTo/TroubleshootADriveNotFormattedCondition" May 10 14:42:35 Didn't show up on the GUI May 10 14:43:31 try it again, and see if it appears in dmesg output or in 'cat /proc/partitions' May 10 14:43:59 if it does, you have some hope. if it doesn't then you start to be seriously concerned. May 10 14:44:26 if it does show up, then immediate dd it to another drive or to a file on a bigger partition May 10 14:45:04 are you going to be around for a while? I've never used dmesg or cat /proc/partitions May 10 14:45:15 I'm a serious newbie at this stuff May 10 14:45:39 It's after midnight here, but anyone else here should be able to help you with dmesg and /proc/partitions May 10 14:46:01 lunchbox2: plug in your drive May 10 14:46:11 then type dmesg | tail -20 May 10 14:46:27 you should see some text indicating that a new device has been inserted May 10 14:47:06 would you recommend I first try just using the USB cable or go just put it in the box to start with May 10 14:47:42 put it in the box directly May 10 14:47:48 less to go wrong May 10 14:47:52 k May 10 14:48:21 I have a little work to do then - I shall return....with questions I'm sure. Thanks guys. Heres to hoping the data gods are feeling generous May 10 14:48:54 BBIAB May 10 15:37:05 still around osmeest? May 10 15:39:35 yep May 10 15:39:55 to recap, you try to access a USB HDD on your PC, right ? May 10 15:40:08 or is it on an NSLU that you try ? May 10 15:40:26 yep - installed the drive to my pc and am trying to recover the data directly May 10 15:40:47 and your PC is running Ubuntu, right ? May 10 15:40:50 so this is messed up - i'm not seeing it in the dmesg - or I don't know what I should be looking at May 10 15:40:52 correct May 10 15:41:02 but it shows up in GParted as an unallocated drive May 10 15:41:25 can you send the last 10 lines of dmesg ? May 10 15:41:30 dmesg | tail -10 May 10 15:42:18 [ 1403.486315] ide: failed opcode was: unknown May 10 15:42:19 [ 1403.486340] hdd: drive not ready for command May 10 15:42:21 [ 1403.534179] ide1: reset: success May 10 15:42:23 [ 1403.534212] hdd: status error: status=0x00 { } May 10 15:42:24 [ 1403.534214] ide: failed opcode was: unknown May 10 15:42:26 [ 1403.534219] end_request: I/O error, dev hdd, sector 0 May 10 15:42:27 [ 1403.534232] hdd: drive not ready for command May 10 15:42:29 [ 1403.534806] hdd: status error: status=0x00 { } May 10 15:42:30 [ 1403.534808] ide: failed opcode was: 0xea May 10 15:42:32 [ 1403.534819] hdd: drive not ready for command May 10 15:44:09 you took the HDD out of a HDD USB box and plugged it straight in the IDE controller ? May 10 15:44:22 yes May 10 15:44:41 are you sure your cabling is OK ? May 10 15:44:52 in particular, check master/slave jumpers May 10 15:45:01 I did change it from a master to a slave May 10 15:45:52 its on the same IDE cable as the ubuntu drive which is a master May 10 15:48:15 why didn't you leave the HDD in its uSB box ? May 10 15:50:04 before I left if I should put it in the box or leave it in the USB form - rwhitby recommended I put it in the box and save myself trouble - I can put it back in if thats going to be easier May 10 15:50:24 I'll light myself on fire if it means I get the data back May 10 15:50:48 I can't promise anything :) May 10 15:51:07 but putting it in the USB box should avoid all the IDE cabling problems May 10 15:51:32 ok - i'll do that - I'll be back in a few :) don't go anywhere please! hehe May 10 15:51:51 I'll do my best but be short :) May 10 16:00:55 [ 74.674146] cx2388x v4l2 driver version 0.0.6 loaded May 10 16:00:57 [ 74.688379] lirc_i2c: chip 0x10020 found @ 0x71 (Hauppauge PVR150) May 10 16:00:59 [ 74.688551] lirc_dev: lirc_register_plugin: sample_rate: 10 May 10 16:01:00 [ 95.338389] NET: Registered protocol family 17 May 10 16:01:02 [ 119.127251] zd1211rw 2-3:1.0: error ioread32(CR_REG1): -11 May 10 16:01:03 [ 119.142480] ADDRCONF(NETDEV_UP): eth1: link is not ready May 10 16:01:05 [ 119.894317] ieee80211_crypt: registered algorithm 'WEP' May 10 16:01:06 [ 119.944603] SoftMAC: Open Authentication completed with 00:0f:66:c2:e0:96 May 10 16:01:08 [ 120.092230] ADDRCONF(NETDEV_CHANGE): eth1: link becomes ready May 10 16:01:09 [ 126.203308] eth1: no IPv6 routers prese May 10 16:03:03 hmmm nothong about usb... May 10 16:03:11 that's not good May 10 16:03:20 :( May 10 16:04:08 I can't say more from here May 10 16:04:49 thats.......... bad :( May 10 16:05:42 [ 253.251024] end_request: I/O error, dev fd0, sector 0 May 10 16:05:44 [ 253.251030] Buffer I/O error on device fd0, logical block 0 May 10 16:05:46 [ 258.783853] end_request: I/O error, dev fd0, sector 0 May 10 16:05:47 [ 258.783859] Buffer I/O error on device fd0, logical block 0 May 10 16:05:49 [ 264.316303] end_request: I/O error, dev fd0, sector 0 May 10 16:05:50 [ 264.316309] Buffer I/O error on device fd0, logical block 0 May 10 16:05:52 [ 269.863653] end_request: I/O error, dev fd0, sector 0 May 10 16:05:53 [ 269.863659] Buffer I/O error on device fd0, logical block 0 May 10 16:05:55 [ 275.418008] end_request: I/O error, dev fd0, sector 0 May 10 16:05:56 [ 275.418014] Buffer I/O error on device fd0, logical block 0 May 10 16:06:35 fd0 is floppy, nothing to do with the hdd May 10 16:06:57 try dmesg | less May 10 16:07:16 go to the end, go backward looking for usb stuff May 10 16:08:21 will it be labeled as USB or have another designation like fd0 May 10 16:09:05 you should see usb stuff and then sd? (? = letter) when it has found partitions on it May 10 16:09:55 I can tell you that it's device /dev/sbd - its being recognized in Gparted as unallocated with no paritions May 10 16:10:33 dev/sdb May 10 16:12:02 oops hold on - i hadn't hit return after that last command so it didn't scroll through everything May 10 16:12:06 [ 46.036268] usb-storage: device scan complete May 10 16:12:07 [ 46.037393] scsi 4:0:0:0: Direct-Access WDC WD12 00JB-22GVA0 PQ: 0 ANSI: 0 May 10 16:12:09 [ 46.039493] sd 4:0:0:0: [sdb] 234441648 512-byte hardware sectors (120034 MB) May 10 16:12:10 [ 46.042115] sd 4:0:0:0: [sdb] Write Protect is off May 10 16:12:12 [ 46.042117] sd 4:0:0:0: [sdb] Mode Sense: 33 00 00 00 May 10 16:12:13 [ 46.042119] sd 4:0:0:0: [sdb] Assuming drive cache: write through May 10 16:12:15 [ 46.044238] sd 4:0:0:0: [sdb] 234441648 512-byte hardware sectors (120034 MB) May 10 16:12:16 [ 46.047111] sd 4:0:0:0: [sdb] Write Protect is off May 10 16:12:18 [ 46.047114] sd 4:0:0:0: [sdb] Mode Sense: 33 00 00 00 May 10 16:12:19 [ 46.047115] sd 4:0:0:0: [sdb] Assuming drive cache: write through May 10 16:12:21 [ 46.047119] sdb: unknown partition table May 10 16:12:59 the last one cofirms what you saw in gparted May 10 16:13:23 was the drive shocked while running ? May 10 16:14:04 was the drive formatted on an nslu ? May 10 16:14:05 "No such file or directory" - it's possible we had a storm come through a few days ago and the power flickered - it's plugged into a surge protector May 10 16:14:17 correct - it's been on there for like 3 yrs May 10 16:15:01 i don't know if the partition table is pc compatible on the nslu May 10 16:19:21 quick question - just upgraded from NSLU2 to ND Pro - only to find there is no libstdc++ or vtun kernel module ... so back on IRC - just wanted to know why the optware SXXoptware script references an rc.optware file in /opt/etc where none exists. Also the /opt/etc/init.d scripts dont seem to be run on reboot - am I missing somehting? May 10 16:20:48 Is rwhitby still the lead maintainer for optware these days? May 10 16:21:51 03bzhou * r8296 10optware/trunk/make/calc.mk: calc: 2.12.3.0 -> 2.12.3.3 May 10 16:23:25 copperbeech: i'm optware feed manager, but i don't have a nd-pro May 10 16:23:29 Hum... May 10 16:23:46 which feed is ND Pro using? May 10 16:26:05 ah ok, ND Pro is using optware/vt4 feed May 10 16:26:25 IIRC, vt4 toolchain sucks, no libstdc++ May 10 16:27:01 http://www.openfsg.com/forum/viewtopic.php?f=18&t=5789 May 10 18:08:00 03bzhou * r8297 10optware/trunk/make/ (libtorrent.mk rtorrent.mk): libtorrent: 0.12.1 -> 0.12.2; rtorrent: 0.8.1 -> 0.8.2 May 10 21:32:49 hi, anybody using transmission? **** ENDING LOGGING AT Sun May 11 02:59:56 2008