**** BEGIN LOGGING AT Sat Nov 11 02:59:57 2006 Nov 11 03:00:39 wait a moment i'll try apex swapped using devio on the slug Nov 11 03:01:02 ok Nov 11 03:03:29 nope it's not working... i'll try to build little again and swap it Nov 11 03:03:36 Hey all, thanks for that package name. However I'm getting 404 errors trying to grab it. Nov 11 03:04:10 and no... ican't upgrade to this version... but i'll se if it's available under slug Nov 11 03:04:40 it should available Nov 11 03:04:56 it'll take a few minutes Nov 11 03:05:03 Err http://mirrors.kernel.org etch/main nslu2-utils 0.10+r58-5 Nov 11 03:05:04 let me know if/when you are ready to continue the procedure with slugimage Nov 11 03:05:04 404 Not Found [IP: 204.152.191.7 80] Nov 11 03:05:18 tkimball: what package ? Nov 11 03:05:27 nslu2-utils Nov 11 03:05:35 That and slugimage really Nov 11 03:06:02 ah - are you using apt-get ? Nov 11 03:06:30 Been using aptitude. Nov 11 03:06:57 Lemme try again on apt-get... Nov 11 03:07:10 0.10+r58-5 is old - 10+r58-6 is the latest in testing - it should be trying to grab 0.10+r58-6 is you have your sources.list file pointing to testing Nov 11 03:07:38 tkimball: apt-get install nslu2-utils slugimage Nov 11 03:12:57 Hmm. I think I'm on the old (pre-apex) install. Lemme go thgrough the manual upgrade page. Nov 11 03:15:00 tkimball: it's a good thing to do ;) with rc1 if the kernel gets updated it flashes itself automatically to the slug... that's probably the best thing with this release Nov 11 03:17:06 Gah. The us.arm.mirror site is not reachable either. Nov 11 03:19:07 Hah, there we go. Nov 11 03:20:51 dumfrac: it'll take more than few minutes... i need to update from 11-04-2006... that's about 20mb of packages Nov 11 03:21:08 ok - i'll be here for a while Nov 11 03:21:49 but the version you said is available Nov 11 03:22:32 slugimage 0.10+r58-6 is the current version in testing Nov 11 03:23:04 also i'll try little binaries once again because i've got a little mixed up earlied and tried big 2 times :/ (i'm up over 48h so my thinking slowly goes away) Nov 11 03:24:23 48h ! my record is ~52h and I wasn't in good shape then :-) Nov 11 03:25:07 my record is over 80h !! 8] Nov 11 03:25:29 whoa ! Nov 11 03:25:35 and most of it in front of bds2006 Nov 11 03:25:44 delphi to be exact Nov 11 03:26:24 ok... the packages are now installing Nov 11 03:27:18 what is bds2006 ? Nov 11 03:27:34 borland developer studio 2006 Nov 11 03:29:38 oddly enough, I think that I spend a lot of quality time with Borland Turbo C++ (for dos) during those 50+ hours :-) Nov 11 03:29:52 I should probably get rid of these references to update/main Nov 11 03:30:42 dumfrac: nice =o] Nov 11 03:30:57 Any idea *why* I have references to updates/main in sources.list? Nov 11 03:31:21 Well, had. :) Nov 11 03:31:27 Szafran: nope - you're the sysadmin :-) Nov 11 03:32:14 Gah. Solaris admin here. I've not done anything that insane since the Y2K weekend. Nov 11 03:32:57 tkimball: I have deb http://security.debian.org/ etch/updates main in sources.list Nov 11 03:33:37 damn... sure hope to get the em extension working... it's gonna be a big power up for the little sluggy ;] Nov 11 03:33:49 oops: I mean tkimball: you're the sysadmin - sorry Szafran :-) Nov 11 03:33:50 *mem Nov 11 03:33:59 np Nov 11 03:34:30 msg GPSFan do you want svn repo access so you can commit your kernel patches? Nov 11 03:34:36 Hmm. I had that too, but also to mirrors.kernel.org (which was the original failure. Anyway, I moved over to the us.arm.mirror and am OK for now. Nov 11 03:35:30 I'll take a stable upgrade right now, and futz with the glitzy bits later. Nov 11 03:35:41 * tkimball waves at rwhitby . Nov 11 03:36:23 This is one thing that Solaris has over Linux, but as I understand the apt packager it gets easier. Nov 11 03:37:11 apt is a thing of beauty and joy forever ! Nov 11 03:37:12 rwhitby: not really, all I have is one, and it is almost identical to an existing one. thanks for the offer. Nov 11 03:37:31 dumfrac: apex little endian swapped works Nov 11 03:37:35 =o] Nov 11 03:37:39 COOL ! Nov 11 03:37:42 yeah Nov 11 03:37:59 now i need to build it with fatslug support Nov 11 03:38:54 AAh. Now *there's* an upgrade I'm holding on. My solder skills are not what they once were. :) Nov 11 03:38:54 remeber that if the apex binary isn't 4*n bytes, devio is not going to be able to swap it Nov 11 03:39:30 the devio command I gave up earlier Nov 11 03:39:31 * rwhitby waves back Nov 11 03:39:35 I did do the turbo change. Nicer. Nov 11 03:40:09 dumfrac: there is a new version of devio in the nslu2-linux svn which pads swaps Nov 11 03:40:29 I need to package it and get tbm to upload for me Nov 11 03:40:39 rwhitby: thanks - I was in the process of asking you exactly that question :-) Nov 11 03:43:37 Bugger. Another flash upgrade. *goes to get flashing notebook out of the car* Nov 11 03:44:16 rwhitby: what is the path to the new devio in the repository ? Nov 11 03:53:48 http://svn.nslu2-linux.org/svnroot/devio I think Nov 11 03:54:08 Hmm - nope. Nov 11 03:54:15 I'm thinking of upslug' Nov 11 03:54:33 I haven't moved devio, cause I need to check with jbowler first. Nov 11 03:54:42 it's at nslu.sf.net's cvs repo Nov 11 03:55:01 actually, it's at devio.sf.net Nov 11 03:55:35 I've been too busy with the nslu2 kernel and apex support and stuff that I haven't packaged up the latest version Nov 11 03:58:35 rwhitby: is it http://devio.cvs.sourceforge.net/devio/devio/src/ ? The last check in was 5 months ago and the comment is: jbowler devio 1.3 (not complete), changes xp to allow byte swapping of lengths which are... Nov 11 04:00:57 dumfrac: thanks for your help... i'll continue the battle when i get up Nov 11 04:01:01 bye everyone Nov 11 04:10:45 dumfrac: that's the version. Nov 11 04:11:03 thanks rwhitby Nov 11 04:44:10 03bzhou * r4435 10optware/trunk/ (Makefile make/py-ctypes.mk): added py-ctypes, promoted for ds101g only Nov 11 05:30:44 greetings friends - Anyone use rtorrent? I'm trying to use it but I get "could not open port for listening" errors Nov 11 05:35:36 I think something's misconfigured, because although I'm connected to a bunch of peers, there's no data transfer Nov 11 05:56:12 I tried py-bittorrent but got errors - testing ctorrent now Nov 11 06:00:53 Anyone else have success with bittorrent clients on the nslu2? Nov 11 06:31:38 03bzhou * r4436 10optware/trunk/ (Makefile make/quickie.mk): added and promoted quickie Nov 11 06:47:06 03bzhou * r4437 10optware/trunk/ (Makefile make/mpack.mk): added & promoted mpack Nov 11 06:59:24 03bzhou * r4438 10optware/trunk/make/py-roundup.mk: py-roundup: upstream upgrade to 1.3.1 Nov 11 07:01:57 03bzhou * r4439 10optware/trunk/make/git-core.mk: git-core: upstream upgrade to 1.4.3.4 Nov 11 07:02:23 03bzhou * r4440 10optware/trunk/make/cogito.mk: cogito: upstream upgrade to 1.8.1 Nov 11 09:56:12 * rwhitby runs upslug2 for the first time since the project started Nov 11 10:41:30 morning Nov 11 14:23:34 Hi. Can someone please help me figure out why my hdd keeps spinning up and down all the time? I would like to keep it spinned down... Nov 11 14:23:46 I've moved /var/log and /dev to ramdisk Nov 11 14:24:08 And I only have three cronjobs run every day (for setting the time) Nov 11 14:24:24 which services? Nov 11 14:26:35 I don't know... Nov 11 14:26:38 How would I check? Nov 11 14:26:44 This is unslung btw Nov 11 14:28:48 this is what ps aux gave me if that's any help: http://www.rafb.net/paste/results/Vy4F6k92.html Nov 11 14:32:32 do you need those upnpd functions? just a side notation Nov 11 14:33:09 not really :) Nov 11 14:33:21 and you know how to disable it? Nov 11 14:33:38 no :) Nov 11 14:33:42 I'll have a look in the wiki Nov 11 14:33:45 from web admin Nov 11 14:34:34 System->untick Enable UPnP Support Nov 11 14:35:05 got it Nov 11 14:35:59 What is CheckDiskFull and WatchDog in crontab? Nov 11 14:36:23 the first should be pretty obvious Nov 11 14:36:40 Yeah, but why do I need it? Nov 11 14:36:49 What does it do if it thinks the disk is full? Nov 11 14:36:50 if your disk space usage gets to 99 and crontab runs it - beep beep beep Nov 11 14:37:09 I've disabled it Nov 11 14:37:20 very annoying during night time Nov 11 14:38:29 hehe, yeah :) Nov 11 14:41:08 What about WatchDog? Nov 11 14:43:33 What does that do? Nov 11 15:37:43 hm... Why after flashing APEX the OS isn't getting data from DHCP ?? (debian-installer OS) Nov 11 15:48:04 sorry got disconnected... hm... Why after flashing APEX the OS isn't getting data from DHCP ?? (debian-installer OS) ... also my router doesn't even see the slug on LAN (just like the slug doesn't have the MAC adress no more) ... after flashing RedBoot back everything is ok... can someone help with this ?? Nov 11 15:48:34 do i need to set MAC adress for APEX in some way ?? Nov 11 16:28:37 can anyone help with the APEX + slugs MAC adress problem ?? Nov 11 16:33:48 Szafran: hi - are you using the 2.6.17 kernel version ? Nov 11 16:34:33 I'm checking whether this has the MAC address patch that tbm applied (I can't remember if it applied to 2.6.17) Nov 11 16:40:20 03bzhou * r4441 10optware/trunk/make/imagemagick.mk: imagemagick: upgrade to 6.3.0 Nov 11 16:41:04 Szafran: ah - 2.6.17 uses the Intel ethernet driver - is your MAC address being set correctly i.e. does it agree with the MAC address printed on the bottom of the slug casing ? Nov 11 16:54:46 2.6.17-2 to be exact Nov 11 16:55:15 and earlier redboot was setting the mac adress Nov 11 16:55:22 now it's gone and apex is there Nov 11 16:55:31 and no mac adress :/ Nov 11 16:56:26 ofcourse it works if i set the send dhcp-client-identifier in /etc/dhclient.conf Nov 11 16:56:45 but was hoping for a change in the apex to make it work Nov 11 16:59:48 dumfrac: right now my slug runs using 64mb, but when i put apex with 128mb support than i'm getting these errors: BUG: warning at arch/arm/mm/consistent.c:363/dma_free_coherent() - any idea how to get rid of them ?? Nov 11 17:01:19 Szafran: what do mean redboot is gone ? Are you using apex as a first stage bootloader ? Regardless, I thought that the Intel ethernet driver set the MAC address (but I could be wrong) - do you still have the structure in flash specified at http://www.nslu2-linux.org/wiki/Info/BootFlash Nov 11 17:02:05 Szafran: sorry, I've never tried compiling the kernel for a FatSlug Nov 11 17:03:13 if you have replaced redboot, could you also have overwitten the MAC address in flash at 0x5003FFB0 Nov 11 17:05:42 this is redboots adress space Nov 11 17:05:52 Szafran: you have a 128 mb slug??? Nov 11 17:06:12 if i don't clear this (second) block of redboot befor flashing apex than it wont work Nov 11 17:06:16 yes 128mb Nov 11 17:06:19 4 chips Nov 11 17:06:23 * dwery wants one too! Nov 11 17:06:29 =o] Nov 11 17:07:50 i have one from about 3 months now and just today i was able to use more than 32mb Nov 11 17:07:50 Szafran: 0x5003FFB0 is where the MAC address for the slug is stored AFAIK Nov 11 17:07:50 i know Nov 11 17:07:50 it's the end of the fis redboot partition Nov 11 17:07:50 Szafran: you can dump that partiton and see if it's still there Nov 11 17:08:00 so, if you have overwritten this, then how will the slug know the MAC address Nov 11 17:09:27 Szafran: was it difficult to solder those 4 chips? Nov 11 17:09:33 damn now it's there... so currently i don't know what helped... or the adress in fis or the changes in /etc/dhclient.conf Nov 11 17:09:41 need to check that... brb Nov 11 17:10:48 Szafran: try using the kernel from the nslu2 cvs Nov 11 17:11:01 bbl Nov 11 17:12:33 dwery: yeah... i've read on the wiki that it works better with fatslugs Nov 11 17:13:50 dumfrac: yeah it works because the fis entry Nov 11 17:15:10 dwery: will this cvs kernel work for debian-installer ?? Nov 11 17:15:51 Szafran: dunno. probably yes. maybe it's better to first install and then changing the kernel Nov 11 17:18:39 i have this debian on my slug Nov 11 17:19:14 and currently i'm setting options for recompiling the 2.6.17 kernel (with options for 128mb) Nov 11 17:22:24 I'd suggest .19 Nov 11 17:22:29 Szafran: you could also try using the Debian kernel in unstable - have a look at http://article.gmane.org/gmane.comp.misc.nslu2.linux/16144 Nov 11 17:22:31 bbl Nov 11 17:23:35 first i'll try with .17 and see what happens - unfortunatelly i'll know that in about 8h because of the compile time on the slug Nov 11 17:25:01 you would have to either do apt-get source linux-2.6/unstable or directly download the .dsc, .diff.gz, and .tar.gz files and then do dpkg-source -x linux-2.6_2.6.18-5.dsc Nov 11 17:26:26 dunno if i'm experienced enough to do it Nov 11 17:27:46 Szafran: it's pretty easy - if you can fatten your slug and build a kernel for it, you _will_ be able to figure it out :-) Nov 11 17:29:32 Szafran: also, if you build a cross-compilation toolchain, you won't have to compile the kernel on the slug - it is much faster :-) check out http://article.gmane.org/gmane.comp.misc.nslu2.linux/16144 (same link as above) Nov 11 17:30:18 assuming you have a Debian box (not the slug) handy Nov 11 17:30:51 ok... i'll try... but now i'll get the slug to compile .17 kernel with 128mb support... and while he does that i'll try to build a crosscompilation debian system under my vmware Nov 11 17:31:28 sounds good Nov 11 17:31:51 i've downloaded debian-testing-i386-netinst.iso yesterday... will it sufficient for this ?? Nov 11 17:31:58 *be Nov 11 17:34:14 not sure - I suspect not because I think that the install is still using 2.6.17 Nov 11 17:35:15 s/install/installer Nov 11 18:02:40 Szafran: you don't need to buil one, just apt-get install gcc-4.0-arm Nov 11 18:03:03 and then make ARCH=arm CROSS_COMPILE=arm-linux- bzImage Nov 11 18:04:00 and Nov 11 18:04:00 and this will give me an deb package ?? Nov 11 18:04:02 devio '<<'arch/arm/boot/zImage >/var/lib/tftpboot/zImage.nslu2.sw 'wb 0xe3a01c02,4' 'wb 0xe3811055,4' 'wb 0xee110f10,4' 'wb 0xe3c00080,4' 'wb 0xee010f10,4' 'xp $,4' Nov 11 18:04:08 no, it will give you a kernel :) Nov 11 18:04:52 you;ll have to download the kernel sources from kernel.org and apply the patches in the nslu2 svn repository Nov 11 18:04:55 it's pretty easy. Nov 11 18:05:42 now i'm confused :/ Nov 11 18:05:57 it's easy, really. start by downloading the kernel Nov 11 18:05:58 i think i'll stick to dumfrac's way Nov 11 18:06:29 the .19-rc5 is quite more advanced than any stock kernel Nov 11 18:07:00 that way i'll get a deb package which i just need to install on the slug and than it'll flash itself Nov 11 18:08:17 oh right. Nov 11 18:08:17 that;'s true. I'm used to test kernels by tftp Nov 11 18:08:32 it'll also be easier for me to compile sthing else than the kernel... i think Nov 11 18:09:40 i need to go or my fiance will kill me in a moment =o] ... bbl Nov 11 18:10:03 :) Nov 11 20:41:12 hi Nov 11 20:45:53 03bzhou * r4442 10optware/trunk/make/py-ctypes.mk: py-ctypes: upstream upgrade to 1.0.1 Nov 11 20:54:39 i keep on getting bus errors when i try to "dpkg-reconfigure locales" on a freshly installed debian Nov 11 20:56:48 03oleo * r4443 10optware/trunk/ (Makefile make/esniper.mk): optware: esniper initial import Nov 11 20:57:06 there seems to be matching scio errors in /var/log/messages Nov 11 20:59:40 ooh it's sdb... maybe my flash swap is worn out... gonna try to setup a swap file instead Nov 11 21:02:11 you wouldnt happen to know why I am getting Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) in the openslug that i just compiled? Nov 11 21:02:23 and by you I mean anyone listening Nov 11 21:03:11 meklort: because openslug head is currently broken Nov 11 21:03:22 the redboot parsing appears to have been broken somehow Nov 11 21:03:22 ok Nov 11 21:04:08 is there anyway that i can fix it (such as tell it the block device in numbes)? Nov 11 21:12:04 ok, the problem was the swap on flash drive. now the locales are generated Nov 11 21:31:02 03bzhou * r4444 10optware/trunk/ (Makefile make/wput.mk): added & promoted wput Nov 11 21:34:53 meklort: nope - we're looking at the problem now Nov 11 21:35:22 (the FIS directory is not getting parsed correctly - you don't want to go further until that is diagnosed and fixed) Nov 11 21:41:49 quick question: does apex support loading the kernel from an usb-disk? the apex-wiki isn't clear about that. Nov 11 21:43:43 rwhitby - we've completed our "port" of openslug to the iomega storcenter! Nov 11 21:43:57 HopsNBarley: who's "we" ? Nov 11 21:44:16 EvilDevil: don't think so Nov 11 21:44:16 my partner and i. two guys and a dog (-; Nov 11 21:44:21 HopsNBarley: cool Nov 11 21:44:28 are you building from OE ? Nov 11 21:44:32 thanks rwhitby Nov 11 21:44:43 yeah, just wanted to say thanks for laying such great ground work. yeah, OE kernel and all. Nov 11 21:45:14 the storcenter is nice, twice the ram and twice the bogomips of the slug. Nov 11 21:45:56 one minor problem remains and i'm hoping you can point me in the right direction. Nov 11 21:46:08 we only have kernel 2.6.12 working, so we are stuck in devfs land. Nov 11 21:46:28 where is the right place to pivot that from /initrd/dev to /dev? Nov 11 21:46:29 HopsNBarley: do you have OE commit ? Can you add it as a new target? Nov 11 21:47:04 i haven't got a key yet, but we are happy to help the OE community. i've got a cross-compiling apache too. Nov 11 21:47:39 bbiab Nov 11 21:48:36 HopsNBarley: congrats! Nov 11 21:49:01 if you need any help with the kernel, drop me a line. we should be easily able to get 2.6.19-rc5 running. Nov 11 21:49:36 consider this droppping a line! i've got vendor patches to 2.6.12.6. i tried these in 2.6.13, but got a kernel that didn't do anything. Nov 11 21:50:20 dwery, i should say, i tried hand applying these to 2.6.13, and got them in with a bit of work... Nov 11 21:50:49 dwery, but, my kernel kungfu is weak, so... Nov 11 21:50:56 are those patches gpl? which kind of processor there's on the appliance? Nov 11 21:51:14 yeah, gpl patches, 8241 (ppc603e core) Nov 11 21:51:54 well, I'm new to the ppc... but I can give a look at the patches if you want Nov 11 21:52:13 HopsNBarley: cool work guys (and pet the dog as well) :-) Nov 11 21:52:54 we are planning on putting a image up soon. are you guys interested? Nov 11 21:53:06 HopsNBarley: what kind of hardware is in the storcenter? Nov 11 21:53:25 I've got a new ppc nas that I'm going to get optware on ;) Nov 11 21:53:37 it's also running 2.6.12.6 IIRC Nov 11 21:53:53 (in the vendor firmware) Nov 11 21:54:21 its a freescale 8241,which is ppc 603e core @200Mhz. 64M ram, gig-E Nov 11 21:54:56 my nas got gig-E, 64M ram and a ppc603e core. Dunno clockspeed. Might be 175Mhz. Nov 11 21:55:09 Not exactly sure which freescale it is :-P Nov 11 21:55:28 NAiL, whats the vendor? Nov 11 21:56:14 It's a QNap TS-101. Nov 11 21:56:41 it looks like we'll have to learn ppc :) Nov 11 21:56:45 rather nice feature set software-wise :) Nov 11 21:57:38 HopsNBarley: The norwegian distributor sent me one so I could get optware running and hopefully also slugos without qnap getting loads of RMA Nov 11 21:58:01 Although, "flash_eraseall" in /sbin isn't a "Good Thing" for less RMA :-P Nov 11 21:58:26 ah, the qnap. Nov 11 21:58:45 NAiL, i'd love to work with you on that. Nov 11 21:58:55 it's nice. I like that it's fanless :) Nov 11 21:59:10 the storcenter has a fan, but you can control it (-; Nov 11 21:59:33 ah, same as the n2100/fsg-3 that I'm working on too ;-) Nov 11 21:59:50 the n2100 and TS-101 has a realtek gig-E interface Nov 11 22:00:12 storcenter has the 8169 Nov 11 22:00:35 yeah, it's the same Nov 11 22:00:39 which RTC? Nov 11 22:00:52 rs5? Nov 11 22:01:45 lemme check... Nov 11 22:02:03 dwery, how can i get those patches to you? Nov 11 22:02:09 Does PPC use MACH_ID like arm does? Nov 11 22:02:21 so a common kernel could be built to work on the storcenter and ts101? Nov 11 22:03:05 i dunno. since they are both 8241, i'd think the kernel would work on either. Nov 11 22:03:09 HopsNBarley: do you have an url for them? Nov 11 22:03:18 dwery, working on it (-; Nov 11 22:03:30 even the manufacturer one would be ok :) Nov 11 22:05:00 heh... googling for powerpc MACH_ID, the third hit is [g2] ;) Nov 11 22:05:03 03:52.50, [g2], when you start the kernel, it's sets up the MACH_ID, and then switches Nov 11 22:05:22 btw, which storcenter is the one you've got? Nov 11 22:07:52 can't find the manufacturer link - as i recall they come on a cd with the product Nov 11 22:08:15 dwery, we've got both the single and dual drive units. Nov 11 22:08:40 the storcenter that I can find on the iomega page is a Pentium 4 unit Nov 11 22:09:36 you have to click on "hard drive" -these mysteriously are labeled network hard drive and not nas Nov 11 22:09:48 could i pastebin them to you? Nov 11 22:10:56 ah,too large for pastbin... Nov 11 22:11:25 NAiL, what does [g2] mean? and what's MACH_ID? Nov 11 22:11:44 HopsNBarley: It's kinda irrelevant ;) Nov 11 22:11:53 NAiL, i'm curious! Nov 11 22:11:58 [g2] is a person, MACH_ID is a machine. Nov 11 22:12:00 Anyway, the thing with MACH_ID is that a kernel can be built for several devices that has different setups Nov 11 22:12:00 :-) Nov 11 22:12:21 NAiL: I think it (MACH_ID) is unique to ARM. Nov 11 22:12:26 dang :-\ Nov 11 22:12:47 Isn't there a similar thing on PPC? Nov 11 22:13:24 the mechanism is great ;) Nov 11 22:15:59 NAiL: is that a SATA disk in the QSnap? Nov 11 22:15:59 <[g2]> NAiL on PPC there was PREP and CHRP Nov 11 22:15:59 likewise: yes Nov 11 22:16:02 <[g2]> brb Nov 11 22:16:04 HopsNBarley: is that a SATA disk in the Iomega Storecentre? Nov 11 22:16:26 [g2]: could you elaborate? :) Nov 11 22:16:51 NAiL: Do you know if the SATA controller is in the 603e, or which controller chip is in? Nov 11 22:17:01 NAiL: PREP and CHRP are the modern BIOSes for PowerPC Nov 11 22:17:12 oh Nov 11 22:17:20 likewise: hang on and I'll check Nov 11 22:17:38 <[g2]> NAiL for a long time a go the PPC to may platforms were PREP and CHRP (Common hardware reference platform) one was IBM platform the other was apple Nov 11 22:17:45 <[g2]> this was probably 10 years ago Nov 11 22:17:55 good old times... Nov 11 22:17:56 NAiL: don't break anything, now be careful with that screwdriver. Nov 11 22:18:12 <[g2]> so there was a whole lot less diversity as compared to the ARM camp Nov 11 22:18:24 <[g2]> hey dwery Nov 11 22:18:28 [g2]: sniff, and the BeBox... and ... Nov 11 22:18:43 likewise, sorry - no. twin ide. Nov 11 22:18:43 <[g2]> that's an oldy the BeBox Nov 11 22:18:43 the amiga ppc... Nov 11 22:19:05 <[g2]> dwery do you get country stations over there ? Nov 11 22:19:15 <[g2]> country music Nov 11 22:19:18 I skipped the ppc when finally giving up on the Amiga. I couldn't get rid of my 68040 A4000 though... Nov 11 22:19:24 likewise: I can't open it right now. dmesg doesn't say anything useful though... It's using libata and that's it :-P Nov 11 22:19:37 likewise: I still have a 060 amiga and at least 9 other amigas Nov 11 22:19:37 NAiL: will google some. Nov 11 22:19:46 [g2]: country music? Nov 11 22:19:52 * [g2] did the 860 and a 603 chips bsp Nov 11 22:20:09 <[g2]> dwery like "hee haw" Nov 11 22:20:14 [g2]: BSD for what OS/company? Nov 11 22:20:25 s/BSD/BSP Nov 11 22:20:26 <[g2]> likewise BSP Nov 11 22:20:34 <[g2]> board support package Nov 11 22:20:47 [g2]: I know the music but don't think I have a station at handy :) Nov 11 22:20:51 <[g2]> bootloader etc... Nov 11 22:20:56 [g2]: yup, saw the typo. I meant BSP for what OS? Nov 11 22:21:38 <[g2]> dwery there's a line from a current county song "Good Ol' days" and it goes something like "lets make some good days tonight" Nov 11 22:21:52 <[g2]> likewise our own OS of course :) Nov 11 22:21:54 eheh Nov 11 22:22:14 :-) Nov 11 22:22:29 * dwery want a PS3 with Cell processor Nov 11 22:22:44 <[g2]> it was a C++ OS and the debugger was taught about the task contexts Nov 11 22:22:44 I ported eCos to the PowerQUICC II, I believe, wasn't that the 860? Nov 11 22:23:14 <[g2]> likewise I think the 860 was the PowerQUICC Nov 11 22:23:48 <[g2]> was it the 8460 the PowerQUICC II Nov 11 22:24:00 [g2]: Ah yes, the PowerQUICC 1. I think I once moved a PowerPC eCos over to it. Not touched any PPC since. Nov 11 22:24:18 <[g2]> we did a lot with the 68360s and I did some stuff with the 860 and 603s Nov 11 22:24:33 I think I still know more m68k and ppc assembly then I understand ARM assembly now :-) Nov 11 22:24:43 s/then/than Nov 11 22:25:24 <[g2]> we has a bit** of an issue with the rev B to rev C stepping for the quicc engines Nov 11 22:26:05 <[g2]> after about 6-8 weeks moto fess'ed up they didn't recompile the microcode and thing were relocated between revs .... doh! Nov 11 22:26:16 yuck. Nov 11 22:26:38 [g2]: I am looking at using the newer PPCs/PQUICCs again for new products. Nov 11 22:27:16 the 834x's are really nice. great $/mips ratio. Nov 11 22:27:26 <[g2]> likewise I'd guess if you did a lot of vector floating point it might be ok Nov 11 22:28:31 <[g2]> but really it depends on the apps Nov 11 22:31:43 So we need another name for SlugOS now, if it's going to be ported to PPC/NAS too? Nov 11 22:32:17 <[g2]> rwhitby I thought we needed a distro name a while ago Nov 11 22:32:30 I intend running slugos on my n2100, fsg-3 and ts-101. Getting debian to run is secondary ;-) Nov 11 22:32:31 <[g2]> rwhitby btw are you following the openmoko stuff ? Nov 11 22:32:54 [g2]: yep Nov 11 22:33:18 <[g2]> rwhitby what do you think about common ARM packages ? Nov 11 22:33:34 HopsNBarley: I just hope the MIPS/WATT ratio (or instruction/J) ratio is high. Nov 11 22:33:45 there were alot of little things that had to be tweaked to get slugos on a new platform. Nov 11 22:34:18 likewise, yeah, it is. 'bout double the ARM, according to my hardware guy. Nov 11 22:34:35 HopsNBarley: Double the MIPS for each Watt??! Really? Nov 11 22:34:59 <[g2]> HopsNBarley compared to what ? Nov 11 22:35:09 likewise, i'll double check that. the ixp425 Nov 11 22:35:10 [g2]: I have no idea what "common ARM packages" means Nov 11 22:35:55 HopsNBarley: the cpu in the storcenter has an FPU, right? Nov 11 22:36:04 NAiL, yeah. Nov 11 22:36:11 ah, brilliant Nov 11 22:36:15 speaking of PowerQUICC's, I architected the new PowerQUICC III chip while I was at Freescale. (MPC8568) Nov 11 22:36:21 HopsNBarley: Do you mean the 834x or 854x? Nov 11 22:36:26 <[g2]> rwhitby openmoko will be building packages for the s3c2410 iirc (ARM920T) Nov 11 22:36:43 <[g2]> they'd run on the slug Nov 11 22:36:58 likewise, the storcenter is the 8241, one more rev back, sadly. Nov 11 22:37:30 likewise, oh, my previous comment, we are trying to keep price down, so 834x, x=3, 7, 9 Nov 11 22:37:59 rwhitby: nice, architecting it at what level or component? Nov 11 22:38:11 likewise: SoC architecture Nov 11 22:38:43 (I was chief architect at Freescale Australia design centre before they shut it down) Nov 11 22:38:53 rwhitby, the PQ II and III family are really nice! Nov 11 22:39:00 HopsNBarley: what's the model number on the StorCenter? Nov 11 22:39:13 * rwhitby is looking to see how much it is in AUD$ Nov 11 22:39:17 rwhitby: I didn't know that. How did you end up compiling anything for ARM? :-) Nov 11 22:39:33 likewise: I had to keep it a secret. Nov 11 22:39:46 rwhitby: when was it shut down? Nov 11 22:39:52 Jan this year Nov 11 22:40:08 MPC8568 has just been announced Nov 11 22:40:20 rwhitby, i've got a dhd250-n3 Nov 11 22:40:50 rwhitby: and where is design now? Nov 11 22:41:01 likewise: can't say - NDA and all that Nov 11 22:41:13 (but it's no longer in Australia :-) Nov 11 22:41:16 dwery, ??? Nov 11 22:41:31 rwhitby: but it still *is* anywhere? Nov 11 22:41:37 likewise: yes Nov 11 22:42:01 rwhitby: you still at freescale? Nov 11 22:42:10 likewise: nope, they shut us down Nov 11 22:42:26 I now run (with 4 others, and 30 staff) our own company Nov 11 22:42:28 www.astc-design.com Nov 11 22:43:04 rwhitby: my girls' nephew just started his own company in Australia. He has a plumbing company though :-) Nov 11 22:43:24 Not quite 90nm scale I guess. Nov 11 22:43:30 hmm Nov 11 22:44:18 HopsNBarley: see /msg Nov 11 22:44:24 * NAiL notes that the CPU in the qnap ts101 is "from the Freescale PowerQUICC family" Nov 11 22:44:35 NAiL: yep Nov 11 22:44:53 HopsNBarley: the rtc seems a ds1337 Nov 11 22:45:49 interrupt less i2c Nov 11 22:45:55 * rwhitby has to go to a kid's bday party now. will look at logs later Nov 11 22:45:56 interrupt-less Nov 11 22:46:09 hmm Nov 11 22:46:22 via ide controller Nov 11 22:46:26 * NAiL notes that the article at tomshardware doesn't have the same hardware list as me Nov 11 22:46:57 the TS-101 jbuzbee tested had an Intel ethernet chip Nov 11 22:47:10 NAiL: and yours? Nov 11 22:47:20 realtek 8169 Nov 11 22:47:32 NAiL: the Intel LX971 or so? Nov 11 22:47:45 Intel 82540EM. Nov 11 22:51:11 HopsNBarley: it should be easy to forward port the patch Nov 11 22:52:44 <[g2]> dwery is 39-mtd load broke for the Loft in the lastest svn ? Nov 11 22:52:51 ;HopsNBarley: regarding TOD, it seems it's only there to set freq = 25000000 Nov 11 22:53:07 [g2]: it's not broke, I think the it's missing the platform device bits Nov 11 22:53:14 let me check Nov 11 22:53:33 HopsNBarley: regarding TOD, it seems it's only there to set freq = 25000000 Nov 11 22:54:05 [g2]: I thought a patch for avila mtd setup was added in the last 24/48 hours Nov 11 22:55:49 <[g2]> dwery I don't see it Nov 11 22:55:57 dwery, how about the 1337 patch? Nov 11 22:56:40 HopsNBarley: you are using the 8343 for a new design? Nov 11 22:56:54 likewise, yeah. Nov 11 22:57:17 HopsNBarley: consumer device, industry control, ...? Nov 11 22:57:20 we are using the storcenter for now as a "close enough" cheap reference platform. Nov 11 22:57:23 consumer. Nov 11 22:57:59 HopsNBarley: Yeah, those devices act as great ref platforms, lotsa community effort, often much better than working with a vendor based board Nov 11 22:58:23 * [g2] waits for you guys to start supporting the community :) Nov 11 22:58:27 rwhitby: the cpu in the ts101 gives 175 bogomips on a 2.6.12.3 kernel. Any guesses on mhz? Nov 11 22:58:44 * rwhitby is not here Nov 11 22:58:50 oh Nov 11 22:58:52 :) Nov 11 22:59:26 anyone else? :) Nov 11 22:59:39 likewise, we tried the vendor first)-; Nov 11 22:59:42 <[g2]> 175 Nov 11 23:00:02 <[g2]> NAiL cat /proc/cpuinfo ? Nov 11 23:00:34 NAiL, my storcenter gives me 133 bogo's at 133Mhz. Nov 11 23:00:52 so, you've probably got the 175 (-; Nov 11 23:01:05 ok, trying to figure out exactly what CPU it is Nov 11 23:01:16 it's an 82xx, apparently at 175Mhz Nov 11 23:01:23 my nas has 2378.95 bogoMIPS ;) Nov 11 23:01:39 but mine cost $80 ! Nov 11 23:01:40 according to tomshardware, it's a PowerQUICC II, but I can't find a 185Mhz model on that list. Nov 11 23:01:52 s/185/175& Nov 11 23:02:16 i bought mine for 75 Euros Nov 11 23:02:36 * NAiL got the TS-101 for EUR0 Nov 11 23:02:46 EvilDevil, wow! Nov 11 23:02:49 what is it? Nov 11 23:03:21 likewise, what are you designing? Nov 11 23:03:37 HopsNBarley: In general: video broadcast equipment Nov 11 23:03:53 [g2]: cpuinfo doesn't give much more info than Sandpoint SPS, 175 bogomips, revision 1.4 Nov 11 23:04:44 HopsNBarley: looking for a follow-up for the ixp425 (which we don't like to much except for its 533Mhz) Nov 11 23:05:03 it *was* a linux based thin client. Fujitsu Siemens Scovery xS (celeron1200, 256mb, 128mb compactflash on a CF to ide adapter, 27watts power consumption at 100% cpu) Nov 11 23:05:39 <[g2]> dwery wb Nov 11 23:05:58 nice. the storcenter is, as i recall, 12W with disk seeking. Nov 11 23:05:58 <[g2]> dwery btw the pata-ixp4xx patch works well and looks good Nov 11 23:06:14 [g2]: great. it will be merged upstream soon. Nov 11 23:06:34 HopsNBarley: Anyway, I'm not going to try to get a custom up-to-date kernel running on the ts101 until I've gotten optware running on the bugger Nov 11 23:06:55 <[g2]> dwery so there's no mtd patch for loading the code from RedBoot Nov 11 23:06:56 GPSFan: ping Nov 11 23:07:18 [g2]: there's the global code. then each platform needs a small bit in the serup code Nov 11 23:08:36 NAiL, okay. well, stay in touch. i do have a package DB... Nov 11 23:09:42 dwery: pong.. Nov 11 23:10:01 GPSFan: you managed to write that avila microcode loading patch, right? Nov 11 23:10:22 package DB? Nov 11 23:11:07 dwery: yep here it is: http://pastebin.ca/246195 Nov 11 23:11:51 GPSFan: ty Nov 11 23:12:03 [g2]: can you test it on the loft and eventually commit to the cvs? Nov 11 23:12:07 <[g2]> GPSFan can you e-mail that to me Nov 11 23:12:09 s/cvs/svn/ Nov 11 23:12:09 dwery meant: [g2]: can you test it on the loft and eventually commit to the svn? Nov 11 23:12:26 <[g2]> dwery yeah that's what I was going to do Nov 11 23:12:30 dwery: All the current patches except 53,78,99 apply properly to 2.6.19-rc5 Nov 11 23:12:30 NAiL, the openslug packages as built for the ppc603e. that might be useful for you. Nov 11 23:12:38 [g2]: sure. Nov 11 23:12:56 <[g2]> GPSFan are you at head on the repo ? Nov 11 23:13:18 53 applied fine on my tree Nov 11 23:13:53 <[g2]> 53 clean here too Nov 11 23:14:58 <[g2]> 78 and 99 commented out in the repo series file Nov 11 23:15:30 [g2]: this was 567 Nov 11 23:15:53 <[g2]> GPSFan I'm on 567 Nov 11 23:16:07 <[g2]> GPSFan thx for the patch btw Nov 11 23:18:11 HopsNBarley: got an email address that you could share with me? I may have a question or two to follow-up on the 834x... Nov 11 23:18:32 <[g2]> who added the apex build btw rwhitby ? Nov 11 23:18:35 HopsNBarley: the 8343 runs at 2 Watts for 400MHz. Not really bad. Nov 11 23:19:04 [g2]: me - the defconfigs need work Nov 11 23:19:12 * rwhitby is really gone now Nov 11 23:19:19 cu Nov 11 23:19:52 [g2]: this was a couple of days ago, I got a clean tarball of rc5 and did an svn update of my older patches directory, then applied the patches in order. since it was nslu2 related i didn't worry. but thought it odd. Nov 11 23:20:02 HopsNBarley: probably not. The TS-101 uses uClibc Nov 11 23:21:04 NAiL: which bootloader on your nas? Nov 11 23:21:16 u-boot Nov 11 23:21:37 same Nov 11 23:22:18 ty Nov 11 23:22:41 (IRC) Nov 11 23:22:41 (aaaugh Nov 11 23:22:41 the wireless network in my house sucks bad Nov 11 23:22:51 NAiL, darn that luck. Nov 11 23:23:30 any you guys know devfs and early boot sequence? Nov 11 23:23:58 i'm pivoting from /initrd (flash) to the disk, of course, leaving devfs mounted on /initrd/dev, where it's not really helping. Nov 11 23:24:10 can I just do a mount --move /initrd/dev /dev ??? Nov 11 23:24:40 that, or use --bind, I would try both Nov 11 23:25:46 before or after the pivot? Nov 11 23:26:26 [g2]: I also built another toolchain, I started with crosstool-0.42, and adjusted some of the options, got a gcc-3.4.5-glibc-2.3.5-binutils-2.16.1 for armeb. it built 2.6.19-rc5 avila kernel, madwifi drivers & tools, the glibc is compatible with the DebianSlug distro. Nov 11 23:27:28 after. Nov 11 23:27:56 I'm hitting my bed. Feeling a whee bit tired here. Nov 11 23:28:24 cya Nov 11 23:30:04 <[g2]> GPSFan just double check all the toolchain patches from OE Nov 11 23:31:44 <[g2]> GPSFan looks like the patch works Nov 11 23:32:27 [g2]: I tested it last night, there is not much to it dwery hinted at what it needed. Nov 11 23:33:17 <[g2]> GPSFan nod. Call me crazy but I like to test stuff :) Nov 11 23:33:28 <[g2]> "Trust -- but verify" Nov 11 23:34:10 [g2]: absolutely! Nov 11 23:34:42 that's why I'm puzzled about why 53 applies for you & dwery and not for me. Nov 11 23:35:19 [g2]: is there a central place to look for the OE toolchain patches. I'm not too keen on installing OE. Nov 11 23:35:21 GPSFan: you applied all the patches in the series? Nov 11 23:35:39 <[g2]> GPSFan in all the .bb files Nov 11 23:35:58 dwery: yes, in numerical order, and it barfed on 53. Nov 11 23:36:06 very strange Nov 11 23:36:14 GPSFan: do you use quilt? Nov 11 23:36:14 03oleo * r4445 10optware/trunk/make/esniper.mk: esniper: add missing depends Nov 11 23:36:22 dwery: If I get a chance later I'll repeat the process. Nov 11 23:36:32 and pastebin the errors. Nov 11 23:36:42 dwery: no Nov 11 23:37:02 it's much easier if you use quilt Nov 11 23:37:39 cd to your source tree, mkdir patches, put patches inside and the quilt push -a Nov 11 23:37:58 03g2-tbillman * r568 10kernel/trunk/patches/2.6.19/ (43-avila-mtd-microcode.patch loft_defconfig series): Add GPSFan's avila mtd-micro code patch. Update series file to include and minor cmdline tweak on loft_defconfig Nov 11 23:38:42 dwery: I'm sure there a lot of things that are easier if I were to use XX____XX (fill in blank with your favorite utility) ;>) Nov 11 23:39:14 <[g2]> GPSFan you'll want to play with quilt for an hour or two Nov 11 23:39:25 quilt is brilliant :) Nov 11 23:40:06 quilt is used by all the major kernel hackers Nov 11 23:40:06 dwery: I'll look at quilt, I have a small script that will apply all patches in a directory in numerical order. same for the shim at the start of the zImage code. Nov 11 23:40:08 :) Nov 11 23:40:36 I hope the series really is in numerical order :-D Nov 11 23:41:35 dwery: I read somewhere that the patches were supposed to be able to be applied in any order. forgot where. Nov 11 23:41:48 well, not exaclt in any. Nov 11 23:41:56 they are supposed to work correctly in numerical order Nov 11 23:42:04 but the actual order is specified in the series file Nov 11 23:43:32 dwery: ah that may be the issue, I'll check it out. Nov 11 23:43:57 anyway, it seems the numeric order is correct Nov 11 23:48:33 dwery: just installed quilt & gquilt, looks nice so far. Nov 11 23:49:10 quilt can be very complex, but at minimum you will use Nov 11 23:49:14 quilt push and quilt pop Nov 11 23:49:31 once you are located in the proper patch, you can quilt edit any source file Nov 11 23:49:39 that's all. :) Nov 11 23:49:50 ;>)) Nov 11 23:52:15 * [g2] predicts another convert :) Nov 11 23:53:04 <[g2]> dwery so how close do you think .19 is to going gold ? Nov 11 23:54:08 [g2]: I'd say quite soon. rc5 seems quite stable. Nov 11 23:59:56 is the ixp_npe ethernet driver already included in .19? Nov 12 00:01:30 <[g2]> GPSFan so other than the EEPROM mac hack, there's nothing missing from the repo rightnow correct ? Nov 12 00:09:24 [g2]: I can't think of anything else atm. Nov 12 00:10:01 <[g2]> GPSFan I've got a patch for that but the nslu2 guys aren't crazy about it Nov 12 00:10:38 * [g2] is testing on r 568 now Nov 12 00:10:48 <[g2]> but it worked before Nov 12 00:11:13 [g2]: yeah, I've read the irc logs. could you e-mail me a copy, I'd like to see how it works. bbl dinner... ;>) Nov 12 00:11:29 <[g2]> GPSFan sure Nov 12 00:11:34 <[g2]> have a great dinner Nov 12 00:14:23 <[g2]> GPSFan I'll have to freshen it up a bit for .19-rc5 as there's a conflict with 43-avila microcode patch Nov 12 00:25:09 <[g2]> GPSFan a funny thing happened on the way to the repo.... turns out you patch applies at the end of the series list, but if applied in the middle breaks another patch Nov 12 00:40:29 [g2]: how interesting, what patch does it break? Nov 12 00:42:21 <[g2]> the mtd pathc Nov 12 00:43:17 [g2]: the patch was made form the diff of the non-working, but patched source, and the working version. so it must rely on something subsequent to 43 making the patching possible. Nov 12 00:45:43 [g2]: probably 97. it's patching avila-setup.c, and probably puts something in that my 43 patch needs. we could just re-number it to 98... Nov 12 00:46:16 <[g2]> possibly it works at the end Nov 12 00:47:47 yeah, thinking back, I should have numbered it 98. sincd I did not do the diff at the proper point in the patching sequence. so much for having the patches apply properly in any order. Nov 12 00:48:14 [g2]: thanks for testing. (Trust but Verify!!) Nov 12 00:48:58 <[g2]> GPSFan I did test by manually applying the patch Nov 12 00:49:33 <[g2]> then when I clobbered and built the patch failed Nov 12 00:50:12 <[g2]> well patching is order dependent Nov 12 00:51:16 that raises the question about adding additional patches in the future. should they go at the end, or is it important that they go in the middle (what ever that means) especially patches that modify the same file. such as 8,40,42,97 all modify avila-setup.c Nov 12 00:52:42 <[g2]> GPSFan there's a source dependency Nov 12 00:52:48 s/42/43 Nov 12 00:53:18 <[g2]> that patches are only dependant based on the code Nov 12 00:53:23 <[g2]> it's a static thing Nov 12 00:54:00 indeed, one must start from a clean source and apply in the "correct" order. Nov 12 00:54:21 <[g2]> GPSFan there are only some interdependencies Nov 12 00:56:25 automated tools like quilt should help. this is the first time I've used patch sets . I usually do just one big patch that covers the whole kernel. I know, of course, that this is no longer the "proper" way, but hey, I aint no programmer... Nov 12 01:03:29 03g2-tbillman * r569 10kernel/trunk/patches/2.6.19/ (3 files): GPSFan's patch needs to be last for now. Nov 12 02:33:49 * rwhitby is back Nov 12 02:37:22 [g2]: which patch are the nslu2 guys not keen on? Nov 12 02:38:25 <[g2]> rwhitby I was talking about the 33- improve mac patch, but it looks like it's in the latest series Nov 12 02:38:46 yeah, dwery has a plan to bypass l-a-k and go through netdev :-) Nov 12 02:39:04 <[g2]> there's an update to that patch that handles the condition the Christian commented on Nov 12 02:39:18 <[g2]> specifically it certianly won't fly on netdev Nov 12 02:39:23 do we have news from Christian? Nov 12 02:39:55 <[g2]> dwery modifiying the mac on if open is a no-go Nov 12 02:40:19 anyone got any other ideas on how to achieve it? Nov 12 02:40:47 <[g2]> there's also a bug in the current patch what if you try to change the hw addr when the intf is down it won't work. Nov 12 02:41:04 no ideas right now.. it's quite late here Nov 12 02:41:13 <[g2]> the patch I've got fixes that as it simply check for a zero mac ether addr Nov 12 02:41:43 [g2]: so you've got a patch that works and you think will be accepted by net-dev ? Nov 12 02:42:05 <[g2]> rwhitby I've got patch that works, but won't be accepted by netdev Nov 12 02:42:23 does anyone have anything better at the moment? Nov 12 02:42:24 I think we can handle that when the other bits are upstream.... Nov 12 02:42:33 <[g2]> the eeprom notifer isn't in the repo either, I'm about to fix that Nov 12 02:43:03 <[g2]> make that the "loft/avila platform specific eeprom notifier" Nov 12 02:43:27 <[g2]> the notifier is there, it's just not called from the Loft/Avila platform Nov 12 02:43:33 cool. I'll even be able to start testing again this week. Nov 12 02:44:07 <[g2]> rwhitby I think the big issue right now is what's up with ka6sox ? Nov 12 02:45:02 <[g2]> the whole GPL driver upstream is a big bundle and I'm sure lennert's going to have a bunch of patches / suggestions Nov 12 02:45:17 yep Nov 12 02:46:05 ka6sox is a bit busy at the moment changing his business stuff over to the new server Nov 12 02:46:26 good night Nov 12 02:46:38 but I'm pretty sure he's planning to get the Loft board up and running this week Nov 12 02:46:42 night dwery Nov 12 02:46:43 <[g2]> Ok cool Nov 12 02:47:01 <[g2]> dwery sweet dreams.... eeprom notifier compiles, I'm rebooting now Nov 12 02:47:30 like all of us, ka6sox needs to put his business first Nov 12 02:47:33 <[g2]> ok that works again Nov 12 02:47:51 <[g2]> rwhitby big nod from me. Nov 12 02:48:07 [g2]: is the common Makefile working for you now? Nov 12 02:48:30 (i.e. after I merged Makefile-BE back in)? Nov 12 02:49:06 <[g2]> rwhitby I haven't tried BE, but I'm running the Loft my just changing the defconfig Nov 12 02:49:26 ok. you asked about the apex build before? Nov 12 02:50:27 anyone know how to reset an ixp420 via JTAG with just TMS TCK and TDI ? Nov 12 02:52:48 <[g2]> rwhitby nod apex is building Nov 12 02:52:58 <[g2]> I'm guessing you added that Nov 12 02:53:08 yeah, but the defconfigs need work Nov 12 02:54:48 [g2]: should the default endiannes of our svn kernel repo makefile be LE or BE ? Nov 12 02:55:55 I figure since the Debian folk will be testing LE, then perhaps our kernel repo should be BE by default. Nov 12 02:57:03 <[g2]> rwhitby depends more on the other platforms Nov 12 02:58:04 <[g2]> I'm building both now I am playing with a very tiny BE system < 1MB Nov 12 02:58:28 I also have to rename the debianslug target to something else. Nov 12 02:58:36 I reckon leslug **** ENDING LOGGING AT Sun Nov 12 02:59:57 2006