**** BEGIN LOGGING AT Wed Mar 28 02:59:56 2007 Mar 28 03:00:10 ok, dunno if that bin is correct for debian or not Mar 28 03:00:11 yes. It is loading tho, so if I ^C it, it shows 1.4.17 Mar 28 03:00:26 can you pastebin.ca the log of apex running? Mar 28 03:00:30 if LE or BE, would it just not work at all? Mar 28 03:00:56 yeah gimme a sec Mar 28 03:01:16 I'll do it from Redboot Mar 28 03:02:04 last line says "booting kernel from 0x00008000..." is that right? Mar 28 03:03:15 ok pastebin.ca/413080 Mar 28 03:04:18 looks like that apex bin is configured to be a first stage bootloader, not a second stage one. Mar 28 03:04:31 do you have the ability to rebuild the apex debian package? Mar 28 03:04:54 ah I see... I saw some instructions for it, but it will take me time to digest and get the env set up. Mar 28 03:05:00 can you Ctrl-C apex, and pastebin the output of 'version' ? Mar 28 03:05:26 we might be able to reconfigure it in apex Mar 28 03:05:33 yeah ok Mar 28 03:06:05 we also need the apex config file that is used in the current debian build. can you pastebin that too? Mar 28 03:06:13 (I don't have it handy) Mar 28 03:06:37 413082 now Mar 28 03:06:45 yeah will do that Mar 28 03:07:01 actually, I do have the config Mar 28 03:09:05 ah too late - it is now 413087 Mar 28 03:09:15 yeah, that apex binary is not the debian one Mar 28 03:09:33 weird, it is the one that is labelled for nslu2 Mar 28 03:09:34 (BTW, can you paste the full URL here, so I can just click on it?) Mar 28 03:09:51 nslu2 == debian, slugos, gentoo, .... Mar 28 03:10:13 ok sorry about that - will do that (URL post) Mar 28 03:10:28 http://pastebin.ca/413087 Mar 28 03:10:38 printenv output now please Mar 28 03:11:36 http://pastebin.ca/413091 Mar 28 03:12:16 change kernelsrc to "fis://Kernel" Mar 28 03:12:48 change ramdisksrc to "fis://Ramdisk" Mar 28 03:13:35 or just run the copy commands with those arguments manually Mar 28 03:14:01 hmm - that cmdline won't set up the RTC either. Mar 28 03:14:32 problem with apex is that the vars are not changeable for some reason. i.e., if I try "setenv kernelsrc 'fis://Ramdisk' " it Mar 28 03:14:36 it's probably better to just compile the debian package for 1.4.17 - it should compile out of the box (it has the correct debian/rules file in the source) Mar 28 03:14:41 will say env not editable Mar 28 03:15:09 what does "dump nor:0x7c000+15k" give as output? Mar 28 03:15:16 it says "Environment region is unwritable" Mar 28 03:15:27 (sorry that was not for your last request) Mar 28 03:15:47 looks like all Fs Mar 28 03:16:12 ah, the environment region is set up wrongly in that binary too (env => nor:128k+64k) - should be nor:0x7c000+15k Mar 28 03:16:31 and you can't change that with an environment variable :-) Mar 28 03:16:38 exactly... catch 22 Mar 28 03:16:51 rebuild apex source using dpkgbuild Mar 28 03:17:09 you know how to do that? Mar 28 03:17:23 ok prolly best to just install DI-RC2 (using original apex) and then compile it on NSLU2. Mar 28 03:17:37 I don't have native debian system (I use CentOS) Mar 28 03:18:03 do you have a master makefile environment set up? Mar 28 03:18:40 no, but I will, as soon as I reinstall DI-RC2 :-) Mar 28 03:18:51 no, I mean on your centos system Mar 28 03:19:23 no, but I could I guess... most instructions were for debian. I saw something about how to compile debianslug (very detailed) Mar 28 03:19:32 you could use our kernel svn repo makefile (http://svn.nslu2-linux.org/svnroot/kernel/trunk/Makefile) to build apex on the centos system, but you need a cross-compiler, and the easiest way to build the correct cross compiler is to set up the Master Makefile on the centos system Mar 28 03:19:56 ok, gotta do some wiki reading :-) Mar 28 03:20:00 (I compile all the nslu2-linux stuff on both debian and centos) Mar 28 03:20:16 maybe I can do it for you here - hang on,. Mar 28 03:20:22 ok, thanks! Mar 28 03:22:33 nope, apex 1.4.17 fails to build Mar 28 03:22:48 src/mach-ixp42x/initialize-sdram.c:68: error: 'CLEANALL_DCACHE' undeclared Mar 28 03:23:22 ah ok... weird... I installed 1.4.17 before, or so I thought, overtop redboot. But that's different than 2nd stage loader obviously Mar 28 03:24:08 I'm sure Marc can build it - you just need to get him to build the debian config. Or get Martin M. to rebuild the debian package for 1.4.17 Mar 28 03:24:55 ok, just ask him to build it for 2nd-stage loader? Mar 28 03:25:06 http://www.archivesat.com/NSLU2-Linux/thread2760935.htm Mar 28 03:26:41 thanks... I've been meaning to get cross-compiler system at sometime. Thanks for link, reading now. Mar 28 03:32:00 RobNC: /msg me an email address to send this apex binary Mar 28 03:32:43 (I got it to build) Mar 28 03:33:41 cool - great job. Mar 28 03:36:58 np - I needed to get that latest version building in our kernel svn repo anyway sooner or later Mar 28 03:37:22 great - I also want to test the fatslug (4-chips). Mar 28 03:39:00 apex probably can't test the memory, right? I mean perhaps a walking ones or something Mar 28 03:39:44 I thought there was a command in latest apex to do something like that Mar 28 03:39:55 did you get the email? Mar 28 03:40:24 yup... gonna try slugimage now Mar 28 03:50:57 yep it works - cool! Mar 28 03:51:33 it says "No SDRAM init when APEX executed from SDRAM" Mar 28 03:51:46 prolly that might be why the env vars were non-editable Mar 28 03:59:15 nope, env vars are separate Mar 28 03:59:24 can you pastebin the output now? Mar 28 03:59:43 sure, same as before, showing "printenv", "version", and? Mar 28 03:59:56 just those will do Mar 28 04:00:12 ok, I have to pull the power on it b/c it starts the D-I install Mar 28 04:01:24 03rwhitby * r805 10kernel/trunk/ (4 files in 2 dirs): Updated to apex 1.4.17 Mar 28 04:03:42 http://pastebin.ca/413126 Mar 28 04:04:12 see text near end: "after Rod Whitby rebuild of APEX 1.4.17" Mar 28 04:05:35 note the env => nor:0x7c000+15k (empty) Mar 28 04:05:46 that means you can edit the environment now Mar 28 04:06:02 cool... will have to try (that is great - so fatslug may work!) Mar 28 04:06:15 ok, that all looks correct. Mar 28 04:06:27 have fun! Mar 28 04:06:50 (and please document all this, cause you're the first to do it) Mar 28 04:07:03 yes you are right! cool!! Mar 28 04:07:32 thanks again, so this apex will be in the newest debian build (after new DI-beta3 perhaps) Mar 28 04:08:08 I will add to the fatslug docs, but should prolly get better way to have DI-rc2 updated with newer apex? Mar 28 04:34:30 rwhitby: cool - it recognizes the full 256MiB... cool!! Mar 28 04:35:14 RobNC: you'll need to get Martin M. to update D-I - I am not involved in that Mar 28 04:35:37 I think I gotta get someone to make and send me a fatslug ... Mar 28 04:35:46 ok no problem... at least now no need to do custom builds (for those like me that are hardware eng not s/w) Mar 28 04:36:20 rwhitby heh I have some extra chips if you're good at soldering. I have metcal stuff at work, with microscope. I wouldn't have attempted it w/o those two. Mar 28 04:37:04 RobNC: how much would you charge me to make one up with 256MiB? I'd pay for the slug, the work, and the shipping to australia ... Mar 28 04:37:51 I'll do it for free, since you helped me get this working. Mar 28 04:38:05 I mean free labor :-) Mar 28 04:38:11 awesome. Mar 28 04:38:28 how much are slugs locally to you, and do you accept paypal? Mar 28 04:38:57 sure, dunno how much. I think TigerDirect (local to us) has them - lemme check. I don't have paypal but wife does Mar 28 04:40:25 it says tigerdirect has them in stock but only 1 (a few miles from work). that must be a mistake. I'll check tomorrow if you want. Mar 28 04:41:25 shipping to australia by usps global express is usually about $15 if it's just the slug (no box, no power supply, etc) Mar 28 04:42:02 gotcha, dunno about export stuff tho. I mean customs, etc. Mar 28 04:42:13 no problems there. done it before lots. Mar 28 04:42:27 ka6sox usually sends me stuff Mar 28 04:42:39 * RobNC has never done outside USA Mar 28 04:43:12 I'll get the details of how you do it from Tom and let you know. I'm pretty sure it's dead easy. Mar 28 04:43:32 gotcha. It's sad b/c I took the memory off boards that were to be scrapped. the company takes our boards away and we don't pay them, they don't pay us. Mar 28 04:43:35 the slug by itself is nice and light, which is why it doesn't cost much Mar 28 04:44:00 you can do serial port mod on it? Mar 28 04:44:17 presumably yes Mar 28 04:44:23 sure, but if you've got a four pin header spare and are soldering already ... ;) Mar 28 04:45:02 heh on mine, I actually put the 3202 upside down on top of the memory. Mar 28 04:45:10 superglued Mar 28 04:45:27 ah, I just use usb<->3.3V cables (I have at least one spare here) Mar 28 04:45:53 I got some 16MB flash too, wanted to try that but opted for the safer approach of just keeping 8MB flash on-board. Mar 28 04:46:15 ah ok usb cables with the prolific chipset makes it easier Mar 28 04:46:19 oooh - 16MB flash would be v nice. Mar 28 04:46:41 * mwester just wants to find a source for the original 8MB flash chip! Mar 28 04:46:48 oh intel right? Mar 28 04:47:10 yeah, but no longer being made - or stocked. Mar 28 04:47:41 I must say that you are right - they're used on products of ours that are discontinued. I had to dig deep for 16MB flash, projects 3yrs old. Mar 28 04:48:25 so for 16MB flash, I assume that the wiki instructions are right - regarding removing R64 and then it can be accessed Mar 28 04:49:12 RobNC: dunno if anyone has done that. Mar 28 04:49:17 which wiki page? Mar 28 04:49:19 I've not tried, and there might be some software issues with that as well (I wonder if the "alias" for the flash would still work) Mar 28 04:49:26 did we ever get the pci bounce buffer stuff sorted for > 64MB RAM? Mar 28 04:50:17 http://www.nslu2-linux.org/wiki/Info/FlashExpansion Mar 28 04:50:37 jacques: dunno Mar 28 04:51:27 RobNC: all the authors of that page are reputable, so I would expect it to be correct. Mar 28 04:51:48 yeah and it seemed correct. Plus, it'd be a pain to re-organize for >16MB. Mar 28 04:52:48 rwhitby if it is possible to binary copy the 8MB image onto a 16MB flash, then I can do that. Mar 28 04:52:49 ok, I gotta go interview someone now - talk to you later RobNC. Let me know how much and where to paypal by email. Mar 28 04:52:57 ok thanks will do. Mar 28 04:53:49 mwester wish there was another source for Intel flash but it seems that everytime we tried a 2nd source, it wasn't compatible with Intel (i.e., AMD or STMicro) Mar 28 04:56:58 I think there's a moral there. Mar 28 05:00:58 yeah so much for monopoly Mar 28 05:13:05 nod Mar 28 05:18:00 03bzhou * r5835 10optware/trunk/ (4 files in 3 dirs): adduser & busybox: 1.1.3 -> 1.5.0 Mar 28 09:21:15 03oleo 07uclibc-unstable * r5836 10optware/ (8 files in 3 dirs): unstable buildroot: upgraade to 4.1.2 and latest SVN Mar 28 09:22:46 03oleo 07uclibc-unstable * r5837 10optware/make/mediatomb.mk: unstable mediatomb: add some types and explicit thread config **** BEGIN LOGGING AT Wed Mar 28 10:36:56 2007 Mar 28 15:51:20 03bzhou * r5838 10optware/trunk/make/lftp.mk: lftp: 3.5.9 -> 3.5.10 Mar 28 16:51:02 03oleo 07uclibc-unstable * r5839 10optware/ (make/screen.mk make/vsftpd.mk sources/vsftpd/sysutil.c.patch): unstable: fix vsftpd, screen Mar 28 19:54:21 03bzhou * r5840 10optware/trunk/make/py-simplejson.mk: py-simplejson: 1.7 -> 1.7.1 Mar 28 19:55:08 has anyone tried software RAID-0 across disks attached to the Slug? Mar 28 19:56:10 if we're talking one disk per port, I would think we could get some performance benefit out of it Mar 28 19:56:24 hardly Mar 28 19:56:36 the transfer speed from the usb chip is dog slow Mar 28 19:58:57 hillct, we've done that. it does work. Mar 28 20:01:27 HopsNBarley: doe it offer any benefit at all? Mar 28 20:01:55 other than that of logical partitions Mar 28 20:02:05 only if you care about your data (-; Mar 28 20:02:18 heh Mar 28 20:02:25 Care? Mar 28 20:02:31 About Data? Mar 28 20:02:34 Ha!!!! Mar 28 20:02:57 * hillct laughs at poor pathetic non-redundant data! Mar 28 20:03:02 any drive can fail at pretty much any time, all joking aside. Mar 28 20:03:11 yeah Mar 28 20:03:13 sometime you get a "warning" Mar 28 20:13:56 * mwester thinks a cron-based rsync task is a better slug solution than RAID. Mar 28 21:01:21 03bzhou * r5841 10optware/trunk/make/erlang.mk: erlang: R11B3 -> R11B4 Mar 28 21:33:39 03oleo 07uclibc-unstable * r5842 10optware/ (make/uclibc-opt.mk platforms/toolchain-ddwrt.mk): unstable uclibc: libraries change Mar 28 21:35:43 hi Mar 28 22:15:06 who can upload a picture to the wiki? Mar 28 22:16:28 jelle: email it to me Mar 28 22:22:05 I wrote some things I found out about the internal usb bus, that I didn't find in the wiki, so I wrote a page on it. Maybe someone can check if I don't lie about things? http://www.nslu2-linux.org/wiki/Info/USBSlotsAndBuses Mar 28 22:28:15 NAiL: there seems to be something messed up with the slugos-bag/cross feed. While they're listed in the pacakges file, the openpbx.org-fax and openpbx.org rev2 packages are missing, while openpbx.org-sounds and openpbx.org-ogi rev2 are found. Is there something wrong? Mar 28 22:28:48 are other recently updated packages affected similarly? Mar 28 22:29:30 I will have to check tomorrow Mar 28 22:30:21 oh, when I say not found, I mean the URLs specified in packages.tar.gz appear to be properly formatted but they result in a 404 Mar 28 22:30:22 ok Mar 28 22:48:54 hillct: remind me in about 2 hours time and I can look at it Mar 28 22:49:56 rwhitby: the 3.10-beta feeds are rebuilt and synced, but I think there's some magic missing Mar 28 22:57:11 I was also seeing some DNS issues which are presumably unrelated, where on occasion, there was no response from whatever nameservers are serving the domain Mar 28 23:50:11 03bzhou * r5843 10optware/trunk/Makefile: obsoleted erl-escript, escript is back in erlang distribution starting R11B4 Mar 29 01:26:51 the feed seems to be operating properly with respect to the openpbx rev2 packages Mar 29 01:27:08 now anyway Mar 29 01:27:25 duno if that was the result of some manual change or an automated rebuild Mar 29 01:27:35 thanks Mar 29 01:30:33 I guess sox never made it into the feed - just sox-native Mar 29 01:52:03 Any APEX experts here - please let me know. Problem with APEX and FatSlug - memory detected fine but access extremely slow Mar 29 01:53:46 hey RobNC Mar 29 01:54:12 can you pastebin where apex detects the memory? Mar 29 01:56:42 hey rwhitby. yeah it detects fine. Will paste URL gimme sec Mar 29 01:56:55 actually I updated the wiki with the output Mar 29 01:57:00 gimme sec to find it Mar 29 01:57:58 http://www.nslu2-linux.org/wiki/Debian/FatSlug - see about 1/3 down the page. Mar 29 02:02:05 ok, that looks fine. what's the deal with slow access then? Mar 29 02:03:33 I dunno, doesn't make sense. When I mean slow, I mean like at least 10-20x slower access. I couldn't do the D-I with 256MB. I reverted the env back, and then installed D-I. Now that I have it installed, I'm rsync'ing to my linux server. I could try now changing the apex env options and boot now, but it bothered me that the whole installer seemed to work significantly slower. Maybe because it wasn't in low-mem mode? Mar 29 02:04:24 so just DI was slower, or is normal stuff slower too? Mar 29 02:04:35 i.e. can you do any benchmarks ? Mar 29 02:05:17 it seemed that everything else was slower too, like partitioning the HDD. Do you know of any benchmarking programs I can run in 32MB mode, and again in 256MB mode? I wonder if something dumb like the IXP420 cache is being disabled and never re-enabled or something else Mar 29 02:06:19 I saw that the new APEX disables the cache (i think I read) so that sdram can be detected properly Mar 29 02:07:28 yeah they called it "cache lock-down", so that SDRAM can be properly detected. Mar 29 02:10:03 yeah I was just about to say L1 cache Mar 29 02:10:21 I've sent an email to Marc Singer - hopefully he sees it and is able to join us here. Mar 29 02:12:11 ok, I have emailed him bunches of time in the past but maybe he has gmail set to go to spam bait. I am wondering if the cache is re-enabled after SDRAM config changed. Mar 29 02:13:48 if I had to make a guess, I'd say it performed like PCs in the past - with the old HX chipset. You remember - go above 256MB or something, and then the thing acts very slow due to high mem not being cached and windows running out of high mem first. Mar 29 02:14:51 sounds plausible Mar 29 02:15:11 does setting mem=64M in kernel command line help? Mar 29 02:17:21 I can try that if necessary. I was using apex to get me into D-I RC2 so not sure if I could access kernel command line, unless it's through the "kernelsrc" parameter Mar 29 02:18:02 hmm - wonder if this the dmabounce thingo? Mar 29 02:18:48 ;-) Mar 29 02:18:58 RobNC: it would be helpful if you could set yourself up to build from our kernel svn repo Mar 29 02:19:22 then we can use that kernel (which does accept cmdline args from apex) to do some testing. Mar 29 02:19:28 yeah I will... but work projects are driving me crazy - get home at 9PM and stuff. not much brain juice left :-) Mar 29 02:20:01 I'm more a h/w engineer but I have done linux kernel compiles and program compiles, so it shouldn't be too hard Mar 29 02:20:11 ok, we can feed you kernels anyway. Mar 29 02:21:01 ya, and best of all, I have JTAG *BigGrin* Mar 29 02:21:02 can you get the latest SlugOS/LE alpha from http://www.hentges.net/tmp/snapshots/SlugOS/ ? Mar 29 02:21:24 ok, is that gonna affect debian-install ? Mar 29 02:21:33 yes Mar 29 02:21:47 it's going to replace the firmware in flash, but not affect anything on your disk. Mar 29 02:22:05 if you're in userland now, you can backup your firmware in flash so you can upslug2 it back later Mar 29 02:22:30 ah ok gotcha, so instead of running 2.6.18-4-ixp4xx kernel Mar 29 02:22:48 we have a tool that backs up flash? cool Mar 29 02:22:54 instead of running Debian altogether. Mar 29 02:23:00 jacques: yeah, it's called 'cat' Mar 29 02:23:14 cat /dev/mtdblock* > backup.bin Mar 29 02:23:31 backup.bin is immediately useable by upslug2. Mar 29 02:23:45 I have my previous D-I RC1 backed up (with functioning Redboot). But once you've used JTAG and get familiar, it's not too "scarey" anymore Mar 29 02:23:58 none of this will touch redboot Mar 29 02:24:03 so no jtag required Mar 29 02:24:43 are you in http://www.hentges.net/tmp/snapshots/SlugOS/accepted/4.3-Beta/20070310/LittleEndian/ yet? Mar 29 02:24:51 familiar? that's a handheld distro Mar 29 02:25:12 I think RobNC meant get familiar with using jtag Mar 29 02:25:14 gotcha, that's a plus obviously. If you guys want, I can run this now. Mar 29 02:25:31 sure, I have 30 mins free. Mar 29 02:26:01 ah heh Mar 29 02:26:11 ok will try now Mar 29 02:27:07 you need to read the licese linked off http://www.hentges.net/tmp/snapshots/SlugOS/ and type the password Mar 29 02:27:14 s/licese/license/ Mar 29 02:27:15 rwhitby meant: you need to read the license linked off http://www.hentges.net/tmp/snapshots/SlugOS/ and type the password Mar 29 02:27:31 then get the file http://www.hentges.net/tmp/snapshots/SlugOS/accepted/4.3-Beta/20070310/LittleEndian/slugosle-4.3-beta-nslu2.bin Mar 29 02:27:41 unpack that with slugimage Mar 29 02:28:12 I will email you a new apex to use Mar 29 02:30:01 ok sorry info overload Mar 29 02:30:12 ok getting that one Mar 29 02:30:45 let me know when you've flashed the new image built from the apex I just sent you and the other files from that slugosle 8MB image. Mar 29 02:32:16 rwhitby you mean slugimage it too right Mar 29 02:32:22 yep Mar 29 02:32:56 unpack the file you get from hentges.net, pack it back up with the apex I sent you 5 minutes ago Mar 29 02:33:16 oh, I need to give you the right slugimage command for that Mar 29 02:33:39 ok gotcha... my email etc is on PC I am on now, but PC with serial port, etc. is another machine ( Mar 29 02:34:51 ok backing up flash now Mar 29 02:35:07 gonna do the packing in a sec Mar 29 02:35:11 I'm prototyping the steps here now Mar 29 02:36:53 RobNC: you need "-r Flashdisk:Flashdisk" for the slugimage Mar 29 02:37:07 (the ramdisk.gz is dummy for slugosle) Mar 29 02:38:04 ok backing up flash via sftp right now Mar 29 02:38:56 slugimage -u -i slugosle-4.3-beta-nslu2.bin Mar 29 02:38:57 do I need to get the ramdisk image from that site Mar 29 02:39:06 no. just the 8MB .bin file Mar 29 02:39:10 I meant zimage Mar 29 02:39:13 ok Mar 29 02:40:11 ok gonna do slugimage commad now Mar 29 02:41:15 slugimage -p -o slugos-newapex.bin -b RedBoot -s SysConf -L Mar 29 02:41:15 apex-slugos-nslu2-arm-1.4.17.bin -k vmlinuz -r Flashdisk:Flashdisk -t Trailer Mar 29 02:42:13 don't I slugimage -u first right Mar 29 02:42:26 yep - I pasted that above Mar 29 02:43:14 yup I see that.. I need to use a real IRC client Mar 29 02:46:01 RobNC: actually, you need to download the http://www.hentges.net/tmp/snapshots/SlugOS/accepted/4.3-Beta/20070310/LittleEndian/zImage-ixp4xxle file and use that instead of the vmlinuz you unpacked. Mar 29 02:46:47 (cause the vmlinuz in the .bin file has the arm-kernel-shim which hardcodes the cmdline) Mar 29 02:46:53 ok thought there'd be a need for that - gonna do that now Mar 29 02:48:16 grrr firefox keeps putting in the .htm extension. Mar 29 02:48:23 in that file (with no extension) Mar 29 02:49:31 make sure you get a file that's about 1MB Mar 29 02:49:52 thing I'm worried about is txt instead of bin mode for xfer. Mar 29 02:50:33 1040272 bytes Mar 29 02:50:55 md5sum is 79b5b305f7b247b77df57f409a02e13b Mar 29 02:51:04 length is correct Mar 29 02:51:19 I'm flashing that now Mar 29 02:51:34 ok checking md5sum now too Mar 29 02:51:45 yep its ok Mar 29 02:52:11 that replaces vmlinuz right Mar 29 02:52:16 yep Mar 29 02:52:34 k Mar 29 02:52:35 ok, that booted Mar 29 02:52:56 flash that image, and make sure it boots normally Mar 29 02:53:21 then reboot and interrupt apex Mar 29 02:53:32 ok gonna try now on mine - gotta update via tftp (b/c using work laptop w/o network access to my linux box) Mar 29 02:55:20 ok, I was able to change the apex cmdline and set mem=16M and boot with 16MB Mar 29 02:55:33 what parameter do you change for that? Mar 29 02:55:51 add "mem=XXM" to the end of the cmdline environment variable Mar 29 02:56:17 setenv cmdline "root=/dev/mtdblock4 rootfstype=jffs2 console=ttyS0,115200 Mar 29 02:56:18 mem=16M" Mar 29 02:56:47 jacques: what should we use in slugos for a simple mem access benchmark? Mar 29 02:57:23 gotcha - brb Mar 29 02:59:13 ok my cmdline for apex is: cmdline *= console=ttyS0,115200 rtc-x1205.probe=0,0x6f noirqdebug Mar 29 02:59:55 that is unmodified (from original apex from D-I RC2 - or was it from rwhitby?) **** ENDING LOGGING AT Thu Mar 29 02:59:56 2007