**** BEGIN LOGGING AT Sun Oct 29 02:59:57 2006 Oct 29 05:02:36 03rwhitby * r4320 10optware/trunk/ (make/denyhosts.mk sources/denyhosts/rc.denyhosts): denyhosts: Updated rc script to kill the process according to the pid file Oct 29 05:11:23 03rwhitby * r4321 10optware/trunk/ (3 files in 2 dirs): denyhosts: Moved the pid file back to /var/run (to match other optware packages) Oct 29 05:39:49 03rwhitby * r4322 10optware/trunk/ (3 files in 2 dirs): dnsmasq: Now uses the pid file to kill the correct process Oct 29 05:54:09 03rwhitby * r4323 10optware/trunk/make/dnsmasq.mk: dnsmasq: Added CONFFFILES support so /opt/etc/dnsmasq.conf is not overwritten Oct 29 06:18:43 03rwhitby * r4324 10optware/trunk/ (3 files in 2 dirs): Oct 29 06:18:43 openssh: update start script to only kill the parent process, if it's been Oct 29 06:18:43 recorded in /var/run/sshd.pid - otherwise killall sshd. This allows us to Oct 29 06:18:43 override stock sshd on fsg-3, while still enabling upgrades on openssh without Oct 29 06:18:43 loosing the child connection. Oct 29 06:24:14 03bzhou * r4325 10optware/trunk/make/erlang.mk: erlang: share host build Oct 29 06:25:09 03bzhou * r4326 10optware/trunk/Makefile: added HOST_STAGING_* variables Oct 29 06:26:04 03bzhou * r4327 10optware/trunk/make/erl-yaws.mk: erl-yaws: use the new ERLANG_HOST_BUILD_DIR Oct 29 06:53:49 03rwhitby * r4328 10optware/trunk/sources/template/ (. postinst prerm rc.template): template: Added example startup, prerm and postinst scripts Oct 29 06:55:40 eno: we need a way to detect which optware host we are running on in the startup scripts (e.g. to do special things on an fsg-3, for instance, so the optware package can read settings from the stock configuration) Oct 29 06:55:46 eno: any ideas? Oct 29 06:59:20 morning all Oct 29 07:12:07 03rwhitby * r4329 10optware/trunk/ (make/bpalogin.mk sources/bpalogin/rc.bpalogin): bpalogin: fixed startup script Oct 29 08:44:25 03blaster8 * r534 10kernel/trunk/patches/2.6.18/ (5 files): Fix 2.6.18 clocksource support, thanks to Gordon F. for diagnosing and fixing this Oct 29 09:09:50 03blaster8 * r535 10kernel/trunk/patches/2.6.19/loft_defconfig: Add loft_defconfig - no reference to mtd-load patch included Oct 29 15:43:12 hi Oct 29 15:43:31 2morning Oct 29 15:43:35 s/2// Oct 29 15:43:35 NAiL meant: morning Oct 29 15:58:16 03g2-tbillman * r536 10kernel/trunk/patches/2.6.19/loft_defconfig: Changes for actual Loft testing. Oct 29 17:04:55 ~seen blaster8 Oct 29 17:05:00 blaster8 was last seen on IRC in channel #nslu2-linux, 22h 49m 47s ago, saying: 'good luck with this'. Oct 29 18:38:38 <[g2]> dumfrac ping Oct 29 18:39:05 [g2] pong Oct 29 18:39:16 <[g2]> dumfrac testing etch ? Oct 29 18:40:20 yeah - I'm trying to keep up with the 2.6.19 patches because Debian is running 2.6.18 and will probably be released with 2.6.18 Oct 29 18:41:05 <[g2]> ok, I've seen a little instability with CF Oct 29 18:41:13 oops - got to go and eat lunch Oct 29 18:41:23 <[g2]> cheers Oct 29 18:41:27 I haven't been following the CF stuff Oct 29 18:41:38 talk to you later Oct 29 18:47:35 [g2]: ping Oct 29 18:47:48 <[g2]> GPSFan pong Oct 29 18:48:13 howdy, do you know if anyone has had ethernet working on BE on the avila? Oct 29 18:48:48 <[g2]> GPSFan I haven't tried running BE with the new IXP drivers but I think they started BE Oct 29 18:48:57 <[g2]> GPSFan are you running the CF driver ? Oct 29 18:49:27 <[g2]> I've tested this AM for a couple hours and things look pretty good except I think there's an issue with the CF Oct 29 18:49:41 yes the cf driver seems to work ok, the npe firmware loads ok and the interfaces come up, but trash comes out the wire. Oct 29 18:49:51 <[g2]> I think there's race condition hiding or a timeout issue Oct 29 18:50:01 <[g2]> GPSFan are you booting to the CF ? Oct 29 18:50:14 ah, that could be hard to find without massive testing. Oct 29 18:50:27 <[g2]> well booting on the drive is enough Oct 29 18:50:49 I have a couple of times. with the 2.6.19-rc3 kernel on cf, and the rootfs on the cf too. Oct 29 18:51:07 'redboot likes hda1, the kernel, sda1 ;>) Oct 29 18:51:09 <[g2]> you've got a tiny rootfs right ? Oct 29 18:51:17 tiny indeed. Oct 29 18:51:40 <[g2]> right I'm testing with sarge and etch and a 500MBs rootfs Oct 29 18:51:55 what are you seeing? Oct 29 18:52:01 <[g2]> it works sometimes, but then the fs gets corrupted Oct 29 18:52:32 ewe... Oct 29 18:52:51 <[g2]> I've corrupted both a CF partition and Microdrive partition on two different boxes Oct 29 18:53:05 <[g2]> .18 with the old driver has been totally stable Oct 29 18:53:16 hmm, baybe a race or buffer flushing issue. Oct 29 18:53:22 <[g2]> oddly, SDA1 on an external drive see good too Oct 29 18:53:42 <[g2]> I haven't tried bonnie++ or anything Oct 29 18:54:06 [g2]: got problems? Oct 29 18:54:11 <[g2]> hey dew Oct 29 18:54:15 <[g2]> hey dwery Oct 29 18:54:37 <[g2]> dwery I was testing this AM for several hours Oct 29 18:54:49 <[g2]> things look pretty good, but not 100% Oct 29 18:54:54 dw hey.. Oct 29 18:55:15 mmm... did you got any dmesg related to corruption? Oct 29 18:56:05 dwery: if you didn't see my dmesg on the ethernet in the logs here it is: http://pastebin.ca/227042 Oct 29 18:56:13 GPSFan: I've got it, thanks Oct 29 18:56:25 strange that eth doesn't work. It was born BE Oct 29 18:56:44 you might want to try with another npe firmware Oct 29 18:57:14 <[g2]> lol "born BE" Oct 29 18:57:22 [g2]: may I have you dmesg lines identifying the CF and microdrive? Oct 29 18:57:42 yeah, I havn't a clue. the interfaces go up & down with cable insertion, and when I ping something the packet tx counter increases, but tcpdump on the target box says bad length (or somesuch) Oct 29 18:57:53 no errors in dmesg Oct 29 18:58:04 mmmm Oct 29 18:58:11 interesting. Oct 29 18:58:45 GPSFan: redboot version? Oct 29 18:58:49 2.02 Oct 29 18:59:03 <[g2]> dwery http://pastebin.ca/227925 Oct 29 18:59:05 I'd try with another firmware Oct 29 18:59:15 <[g2]> dwery 2.02 works fine for me Oct 29 18:59:16 you might download them from the intel website Oct 29 18:59:25 [g2]: ixp4xx net driver? Oct 29 18:59:38 [g2]: that's 2.6.18? Oct 29 18:59:47 maybe the net driver does not like the npe firmware in 2.02 redboot Oct 29 18:59:54 <[g2]> GPSFan .18 is fine, .19 is fine for the Intel Driver Oct 29 19:00:17 <[g2]> I've been runnning .18 alot with the new driver and RedBoot 2.02 Oct 29 19:00:23 [g2]: open source net driver works for you? Oct 29 19:00:38 <[g2]> dwery yeah I've been running a quite well Oct 29 19:00:48 <[g2]> s/a/it/ Oct 29 19:00:49 [g2] meant: dwery yeith I've been running a quite well Oct 29 19:00:56 <[g2]> doh! Oct 29 19:00:59 :) Oct 29 19:01:02 lol Oct 29 19:01:05 <[g2]> dwery yeah I've been running it quite well Oct 29 19:01:16 and you are running BE, right? Oct 29 19:01:19 <[g2]> LE Oct 29 19:01:26 BE Oct 29 19:01:40 <[g2]> nod, but it swaps the microcode Oct 29 19:01:46 [g2]: can you try the net driver in BE mode? Oct 29 19:02:03 <[g2]> dwery when I get around to testing BE Oct 29 19:02:18 <[g2]> I don't have the toolchain or rootfs setup Oct 29 19:02:22 I don't get any errors out of the microcode load. would things like packet counters work if the microcode ahd crashed? Oct 29 19:02:26 [g2]: uh. I thought you had Oct 29 19:02:36 <[g2]> dwery long time ago Oct 29 19:02:42 GPSFan: I suspect the driver is 99% ok Oct 29 19:02:55 dwery: I'm sure it is. Oct 29 19:03:02 with maybe a minor issue that prevents BE from working properly Oct 29 19:03:36 it would be interesting to have a tcpdump of a single IP packet, like a ping Oct 29 19:03:43 <[g2]> GPSFan dwery I've pinged Kaloz Oct 29 19:03:43 a ping with a known payload Oct 29 19:03:44 I looked through the code for any any endian references, but couldn't identify anthing obvious. Oct 29 19:04:32 can't remember the option, but I think there was one to se the payload. for the count the option is -c Oct 29 19:05:02 s/to se/to set/ Oct 29 19:05:02 dwery meant: can't remember the option, but I think there was one to set the payload. for the count the option is -c Oct 29 19:06:25 <[g2]> Kaloz appear! Oct 29 19:06:47 * [g2] concentrates Oct 29 19:06:57 /summon Kaloz Oct 29 19:07:05 howdy Oct 29 19:07:39 <[g2]> Kaloz welcome Oct 29 19:07:39 * NAiL suspects [g2] knew he was coming ;) Oct 29 19:07:39 dwery: I think this it. Oct 29 19:08:15 http://pastebin.ca/227938 Oct 29 19:08:15 :-D Oct 29 19:08:15 aha Oct 29 19:08:18 <[g2]> Kaloz fyi .19-rc3 and the new IXP open driver are working pretty well Oct 29 19:08:19 * NAiL suddenly knew where he recognized Kaloz from Oct 29 19:08:21 and nbd Oct 29 19:08:27 :) Oct 29 19:08:31 <[g2]> welcome nbd Oct 29 19:08:33 hiya there :) Oct 29 19:08:41 hi Oct 29 19:09:33 <[g2]> GPSFan Kaloz and nbd are openwrt guys Oct 29 19:09:38 still need to debrick my ixp unit, stupid openrg loader :) Oct 29 19:09:43 [g2]: good. any idea why doesn't 2.6.17 (w gcc 4.1.1 and binutils 2.17) boot on the avila? Oct 29 19:10:04 <[g2]> dwery wb Oct 29 19:10:06 gaim crashed :) Oct 29 19:10:13 ah, I have several wrt54g's !! Oct 29 19:10:25 [g2]: please introduce Kaloz and nbd to me Oct 29 19:10:35 <[g2]> Kaloz ppl have been runing 2.6.17/gcc 4.1.1 for quite some time Oct 29 19:10:52 <[g2]> dunno which the version binutils Oct 29 19:11:15 <[g2]> the 2.18 port for the Loft/Avila is all done for LE Oct 29 19:11:28 <[g2]> I'm testing .19-rc3 now Oct 29 19:11:37 <[g2]> dwery did CF/libata chagnes Oct 29 19:11:47 <[g2]> s/chagnes/changes/ Oct 29 19:11:48 [g2] meant: dwery did CF/libata changes Oct 29 19:11:53 GPSFan: can you send the tcpdump in binary form to me via email? Oct 29 19:11:59 [g2]: wondering if something is missing from mainline back then.. btw, am i wrong, or the avila is BE by default as well? Oct 29 19:12:17 <[g2]> yeah BE is the default Oct 29 19:12:41 dwery: uh, tell me what to set as params for tcpdump, and I'll get it done. Oct 29 19:12:44 <[g2]> we use a 2 instructin shim to switch the box to LE to run Debian Oct 29 19:13:20 GPSFan: tcpdump -w file Oct 29 19:13:34 <[g2]> Kaloz you'll also want to pickup FIS updates Oct 29 19:13:40 [g2]: yeah.. well, i won't switch it, as networking in general is BE :) we'll have iop support soon, tho (got an allnet 6500 last week) Oct 29 19:13:43 GPSFan: maybe you want to filter like "host " Oct 29 19:14:40 Anyone have a guess as to when kernel updates later than 473 will be added to the main repositories? Oct 29 19:14:45 <[g2]> Kaloz that's fine. We're pretty endian/libc agnostic Oct 29 19:15:05 joshin: HEAD builds svn rev 534 now Oct 29 19:15:14 <[g2]> heh... a gentoo guys speaks :) Oct 29 19:15:17 Oh, that must be new. Oct 29 19:15:24 joshin: yeah, earlier today Oct 29 19:15:28 Cool! Oct 29 19:15:36 Now if only I could connect. :p Oct 29 19:15:48 Yeah, Gentoo Forever!!!! Oct 29 19:15:50 :) Oct 29 19:16:08 <[g2]> joshin are you running the box BE or LE ? Oct 29 19:16:35 * [g2] would think you'd want to run BE Oct 29 19:16:54 Both my Slugs are BE and running 473 kernels. One is Gentoo bootstrapped over OpenSlug and the other is OpenSlug only Oct 29 19:17:08 * NAiL hopes the LE network performance will get a boost using NPE-A Oct 29 19:17:21 * joshin is using his Gentoo slug running irssi to type to all of you Oct 29 19:17:48 <[g2]> joshin which kernel/IXP driver are you currently running Oct 29 19:18:08 * [g2] guesses 2.6.18/open IXP driver Oct 29 19:18:30 Linux slug 2.6.18slugos #1 PREEMPT Wed Oct 18 07:22:58 MST 2006 armv5teb XScale-IXP42x Family rev 1 (v5b) Linksys NSLU2 GNU/Linux Oct 29 19:18:53 <[g2]> joshin that's built with gcc 4.1.1 right ? Oct 29 19:19:14 IXP4XX MAC driver 0.2.0 Oct 29 19:19:30 The kernel and all drivers are built with the defaults as per the master makefile Oct 29 19:19:43 so the open ixp driver seems to work fine in BE... ? Oct 29 19:19:44 My Gentoo chroot/pivot-root is all built with 4.1.1 and glibc2.4. Oct 29 19:19:50 cat /proc/version will let you know Oct 29 19:19:51 <[g2]> I'm pretty sure that's 4.1.1 but dunno the binutils ver Oct 29 19:20:13 I had some ipv6 related oddnesses with glibc2.5 that I haven't figured out. (May have been my fault) Oct 29 19:20:18 Checking on versions now... Oct 29 19:21:00 <[g2]> hey that /proc/version trick is pretty handy :) Oct 29 19:21:13 gcc is 4.1.1 , binutils looks to be 2.16 Oct 29 19:21:36 my gentoo chroot/pivot-root uses binutils 2.17 Oct 29 19:22:45 * joshin makes an offering to the MTN gods to allow my connection Oct 29 19:23:16 <[g2]> lol Oct 29 19:23:50 [g2]: you might try adding a small delay in ixp4xx_mmio_data_xfer Oct 29 19:24:22 [g2]: just after *data->cs0_cfg &= ~(0x01); Oct 29 19:24:27 dwery: email on the way, tcpdump didnt find anything with the filter on, so I just captured a few packets the ethernet mac is in the body of the e-mail message Oct 29 19:24:30 <[g2]> dwery ok. root delay doesn't seem to effect it Oct 29 19:24:35 and another before *data->cs0_cfg |= 0x01; Oct 29 19:24:48 the old ide driver had udelay(100) Oct 29 19:25:05 that will affect performance :( Oct 29 19:25:10 Any hints on this: usb 1-1.2: reset high speed USB device using ehci_hcd and address 4 Oct 29 19:25:13 Oct 29 19:25:26 Its my microdrive. Oct 29 19:25:42 joshin: I'am having troubles with usb too, but I'am running .19-rc3 Oct 29 19:26:02 It does recover. I'm not sure if it is a kernel issue or if I should track down an AC adapter to make the hub/reader powered. Oct 29 19:26:02 <[g2]> joshin I was seeing some of those on 2.6.18 with OHCI under heavy load Oct 29 19:26:14 Gentoo by default makes it heavy load. :) Oct 29 19:26:17 <[g2]> but only on 1 drive Oct 29 19:26:44 <[g2]> joshin yeah building Gentoo is a good CPU and Disk test Oct 29 19:26:55 <[g2]> a developer's patience test too Oct 29 19:27:01 <[g2]> :) Oct 29 19:27:15 Thank goodness for distcc. My 3 A64's do the compiling which speeds things up dramatically. Oct 29 19:27:28 Now if only I could figure out how to do the same with OE. Oct 29 19:27:38 GPSFan: interesting, you're runninv IPv0 :D Oct 29 19:27:47 <[g2]> dwery I'll have to come back to testing the libata changes in a bit Oct 29 19:28:01 dwery: that's interesting. Oct 29 19:28:35 it really seems to be a byte swap issue Oct 29 19:37:23 GPSFan: given the level of alteration in the packet, I don't think anymore is a simple byte swap problem Oct 29 19:38:37 you might want to try with another firmware or maybe use the other interface Oct 29 19:44:51 dwery: tried the other interface, it didn't work either, I sent you a tcp dupm of its packets too. Oct 29 19:45:15 Stupid MTN gods didn't like my sacrifice. Still refusing to allow me to connect. Oct 29 19:50:12 GPSFan: :( Oct 29 19:51:17 GPSFan: do you have another fw? Oct 29 19:51:22 dwery: wouldn't there be an error message from the npe if it didn't like the downloaded firmware ie. corruption/checksum error etc? Oct 29 19:51:29 the fw is ok Oct 29 19:51:42 but probably the driver doesn't like it Oct 29 19:51:56 dwery: no I don't I'll have to dl it from Intel.. gotta run out for a while. thanks for all the help. Oct 29 19:51:59 in terms of minor changes in the api Oct 29 19:52:04 or something like that.. I'am guessing Oct 29 19:52:11 we would need its author Oct 29 19:52:18 ok, see ya. Oct 29 19:58:12 [g2]: you wanted to speak about something, didn't you? :) i'm a bit sleepy :P Oct 29 19:59:13 <[g2]> Kaloz I wanted to let you know several ppl are running and testing the latest in BE land Oct 29 19:59:31 <[g2]> specfically, GPSFan is beginning testing on 2.6.19-rc3 Oct 29 19:59:40 <[g2]> and joshin has Gentoo going Oct 29 20:00:06 <[g2]> Kaloz that's all Oct 29 20:00:47 <[g2]> I'll be getting to testing BE in a little while, but right now mostly LE Oct 29 20:01:15 okay :) Oct 29 20:02:56 will the CF driver go into mainline? Oct 29 20:03:49 Kaloz: yes, if we find the corruption bug Oct 29 20:04:16 <[g2]> Kaloz and have it's tested for both endians Oct 29 20:06:18 okay, sounds good Oct 29 20:06:46 [g2]: theorically, we should check the IXP4XX specs regarding the setup time of the exp bus line we are twiggling in the cf driver Oct 29 20:06:52 but I'am lazy :) Oct 29 20:08:16 <[g2]> dwery you're not writing directly to the EXP bus you're interfacing via the memory controller right ? Oct 29 20:08:16 no Oct 29 20:08:16 the bit is twiggled directly Oct 29 20:08:16 data goes thru the mem controller Oct 29 20:09:51 <[g2]> Ok so you are setting everything up via bit twiddling then pulling the data via memory accesses ? Oct 29 20:10:03 yes. like the old ide driver. Oct 29 20:11:08 btw, my loft nfs boots in 15 secs :) Oct 29 20:12:27 <[g2]> :) Oct 29 20:15:41 I've checked the ixp driver, there should be no endianness issues Oct 29 20:15:47 I can only think of different microcodes Oct 29 20:16:46 <[g2]> well everyone but us runs BE Oct 29 20:17:42 <[g2]> dwery I've got some more test data Oct 29 20:18:03 everyone but us uses different microcode :) Oct 29 20:18:26 <[g2]> that's pretty easy to switch Oct 29 20:18:45 yep. we need a BE user to try Oct 29 20:18:49 more test data? Oct 29 20:19:45 <[g2]> Ok the image booting to sda1 on external driver and on hda1 with the old GW ide LE patch is booting fine on 2.6.19-rc3 Oct 29 20:20:25 <[g2]> the libata driver hangs during "Init", but think it's because data isn't getting written to the disk properly Oct 29 20:20:43 <[g2]> but that needs more investigation Oct 29 20:20:51 you're using ide and libata together? Oct 29 20:21:04 <[g2]> no. One or the other Oct 29 20:21:13 ok Oct 29 20:21:27 <[g2]> I tried the libata stuff first Oct 29 20:21:34 it hangs with any message? Oct 29 20:22:17 <[g2]> Init hangs, then on reboot the rootfs is corrupted Oct 29 20:22:36 <[g2]> pretty badly Oct 29 20:22:47 <[g2]> like fsck doesn't fix it Oct 29 20:23:09 <[g2]> it repairs it, but the fs isn't bootable anymore Oct 29 20:25:16 quite strange. Oct 29 20:25:58 mine works quite realiably Oct 29 20:27:53 03pawsa * r4330 10optware/trunk/ (3 files in 2 dirs): sendmail: fix queue location in default config file Oct 29 20:30:25 [g2]: any integrity test I can run? Oct 29 20:35:09 <[g2]> dwery I think we can setup some tests Oct 29 20:35:13 [g2]: I just copied a 30MB file Oct 29 20:35:28 and the md5sum is NOT correct Oct 29 20:35:42 <[g2]> heh... that's a good test Oct 29 20:35:45 <[g2]> :) Oct 29 20:35:51 also Oct 29 20:35:53 ata1: no sense translation for status: 0x51 Oct 29 20:35:53 ata1: translated ATA stat/err 0x51/00 to SCSI SK/ASC/ASCQ 0xb/00/00 Oct 29 20:35:53 ata1: status=0x51 { DriveReady SeekComplete Error } Oct 29 20:35:53 ata1: no sense translation for status: 0x51 Oct 29 20:35:53 ata1: translated ATA stat/err 0x51/00 to SCSI SK/ASC/ASCQ 0xb/00/00 Oct 29 20:35:54 ata1: status=0x51 { DriveReady SeekComplete Error } Oct 29 20:35:59 <[g2]> "if it bleeds, we can kill it" :) Oct 29 20:36:07 trying to add some delay Oct 29 20:53:30 03azummo * r537 10kernel/trunk/patches/2.6.19/96-pata-ixp4xx.patch: Fixed data integrity bug. Oct 29 20:56:06 does anyone know if you can use a usb hub with debianslug? Oct 29 20:57:17 barnseenio: I think so. Oct 29 20:58:16 k, i'll have another sniff on d'web, and maybe mailing list Oct 29 20:59:42 03pawsa * r4331 10optware/trunk/sources/sendmail/config.patch: sendmail: one more default queue location fixed Oct 29 20:59:53 <[g2]> dwery THX for the fix Oct 29 21:10:42 [g2]: it seems the second udelay is required too Oct 29 21:11:01 <[g2]> dwery more errors ? Oct 29 21:11:07 yes Oct 29 21:11:19 <[g2]> is only the first in the repo ? Oct 29 21:11:48 yes Oct 29 21:12:17 * [g2] stops build Oct 29 21:16:07 03azummo * r538 10kernel/trunk/patches/2.6.19/96-pata-ixp4xx.patch: Hopefully fixed the data integrity bug Oct 29 21:17:39 <[g2]> new build started Oct 29 21:19:03 I'am doing more integrity checks Oct 29 21:21:26 <[g2]> super Oct 29 21:22:07 <[g2]> if I can boot I'll run bonnie++ Oct 29 21:23:07 bonnie also does integrity? Oct 29 21:24:43 apt-get install bonnie++ :) Oct 29 21:26:17 <[g2]> dwery do you have a microdrive ? Oct 29 21:26:25 nope :( Oct 29 21:26:47 <[g2]> use and old flash :) Oct 29 21:26:54 <[g2]> s/and/an/ Oct 29 21:26:54 [g2] meant: use an old flash :) Oct 29 21:27:14 yes.. I will see If I can fine a cheap one on ebay Oct 29 21:27:19 s/fine/find/ Oct 29 21:27:19 dwery meant: yes.. I will see If I can find a cheap one on ebay Oct 29 21:30:02 speaking of which, regular write cycles to /var and /tmp are supposed to hurt compact flash media, how about that on a slug running off e.g. a 1gig usb thumb drive? Oct 29 21:32:09 [g2]: im trying with delay 10. integrity seems ok, but bonnie++ triggers ata errors Oct 29 21:32:35 <[g2]> dwery could be more than one problem :) Oct 29 21:32:51 <[g2]> a) data integrity Oct 29 21:33:01 <[g2]> b) libata stress Oct 29 21:33:16 I'll let you try with delay 100 :) Oct 29 21:33:53 <[g2]> laptop still building Oct 29 21:34:22 <[g2]> wow... plugs in CPU fan Oct 29 21:35:11 <[g2]> cpu temp 82C Oct 29 21:35:54 [g2]: how much should a new 2GB microdrive cost? Oct 29 21:36:10 <[g2]> < $100 Oct 29 21:36:18 <[g2]> may $125 Oct 29 21:36:30 I think I've found one at 79EUR at a supplier Oct 29 21:36:32 <[g2]> the 4Gs were $150 6 months ago Oct 29 21:36:37 and several others on ebay Oct 29 21:37:16 maybe I'll buy one .. Oct 29 21:37:54 <[g2]> depends on what your are planning to do Oct 29 21:37:59 <[g2]> you' Oct 29 21:38:13 playing with the loft :) Oct 29 21:39:08 <[g2]> well same cost gets you a 120G USB 2.0 drive that runs 15X faster on the Loft Oct 29 21:39:28 <[g2]> 60x capacity, 15X speed same cost Oct 29 21:39:37 but wouldn't allow me to test the driver :) Oct 29 21:39:52 03blaster8 * r539 10kernel/trunk/patches/2.6.18/11-arm-generic-time.patch: Remove timeofday.h noise from 11-arm-generic-time.patch Oct 29 21:39:54 <[g2]> right it all depends on your goals Oct 29 21:43:36 * [g2] hugs dwery http://pastebin.ca/228153 Oct 29 21:43:45 <[g2]> woot! Oct 29 21:45:17 <[g2]> bonnie running Oct 29 21:45:23 <[g2]> ++ Oct 29 21:45:39 great! Oct 29 21:47:31 dumfrac: ping Oct 29 21:47:35 ~seen dumfrac Oct 29 21:47:52 dumfrac is currently on #nslu2-linux (1d 5h 51m 12s). Has said a total of 6 messages. Is idling for 3h 6m 15s, last said: 'talk to you later'. Oct 29 21:53:20 [g2]: how's bonny doing? Oct 29 21:53:25 <[g2]> dwery running away Oct 29 21:53:37 :) Oct 29 21:55:02 dwery: data-corruption bug? Oct 29 22:03:16 <[g2]> dwery it's on to the rewriting stage Oct 29 22:03:24 good Oct 29 22:03:46 maybe I might add a module parameter for that udelay value Oct 29 22:04:20 <[g2]> dwery I'm gonna be afk for a bit Oct 29 22:04:44 <[g2]> want me to post the results in here or e-mail when I check back in later ? Oct 29 22:06:07 email please Oct 29 22:06:14 I'll go to bed in a while Oct 29 22:06:32 <[g2]> dwery nod. THX for all the help and fixes Oct 29 22:06:54 thanks for your tests Oct 29 22:07:02 <[g2]> sweet dreams if I don't see you before bed Oct 29 22:07:48 <[g2]> dwery np. always a pleasure Oct 29 22:08:12 <[g2]> still running Oct 29 22:17:01 03blaster8 * r540 10kernel/trunk/ (12 files in 3 dirs): Cleanup to patch numbers Oct 29 22:17:37 dwery: ping Oct 29 22:17:44 blaster8: pong Oct 29 22:17:45 Time to run svn update - I've changed some of the patch numbers Oct 29 22:17:52 ok Oct 29 22:23:57 Ahahahaha, the MTN gods finally allowed me to connect! Oct 29 22:37:04 damn... jtag is really slow from a 125 mhz device with only ppdev available Oct 29 23:03:35 <[g2]> nbd which arch ? Oct 29 23:04:28 [g2]: mipsel, wl-500g Oct 29 23:04:36 but i switched to using my amd64 box now Oct 29 23:04:37 much faster :) Oct 29 23:05:47 <[g2]> heh Oct 29 23:10:37 i hope the redboot image boots on this board Oct 29 23:10:45 i don't want to see the rgloader crap again. ever. Oct 29 23:38:19 03rwhitby * r4332 10optware/trunk/Makefile: Makefile: demoted bpalogin Oct 30 00:36:42 [g2]: ping.. how did your bonnie tests go? Oct 30 02:41:51 03rwhitby * r4333 10optware/trunk/make/libftdi.mk: libftdi: Fixed the section field. Oct 30 02:59:04 [g2]: found a problem with the realtime clock. "hwclock --show" returns: # hwclock --show Oct 30 02:59:05 hwclock: ioctl() to /dev/rtc to turn on update interrupts failed unexpectedly, errno=25: Inappropriate ioctl for device. **** ENDING LOGGING AT Mon Oct 30 02:59:57 2006