**** BEGIN LOGGING AT Mon Jun 08 02:59:57 2009
Jun 08 03:22:54 <cshore> Anyone if the kamikaze revision number is a makefile variable?
Jun 08 03:23:06 <cshore> or svn commit?
Jun 08 03:24:04 <rwhitby> generated from svn commit rev
Jun 08 03:24:27 <cshore> So $(shell svn commit rev) should get me it as a string in the makefile?
Jun 08 03:24:45 <rwhitby> dunno exactly how it's done, sorry.
Jun 08 04:54:35 <russell_> yeah, weird.  something not triggering new images
Jun 08 04:56:39 <russell_> even after forcing the rebuild of the kernel
Jun 08 05:03:41 <russell_> (also, the thing it did build with make target/install failed to boot)
Jun 08 07:32:22 <russell_> even after a make dirclean, bin contains no images
Jun 08 08:24:18 <danage> salut [florian], j'ai le routeur chez moi et reste a ta disposition
Jun 08 08:29:00 <[florian]> danage: hey danage
Jun 08 08:29:20 <[florian]> danage: ok, I am at work now, but later when I get back home we can try something
Jun 08 08:29:27 <menardorama> hi, is somebody could give me a hand in setting up a rt2500 usb wireless card ?
Jun 08 08:29:34 <menardorama> I've rebuilt from trunk installed the required driver but i don't know what to put in /etc/config/wireless
Jun 08 08:34:48 <danage> [florian]: yes, no hurries, just wanted to let you know. i'm at work until ~5pm too. i have the router here but home is better for me as well
Jun 08 08:35:20 <[florian]> danage: cool
Jun 08 08:38:45 <russell_> anybody build trunk recently?  /me getting weird non-image-rebuilding behavior.
Jun 08 08:39:03 <[florian]> russell_: will do, hold on
Jun 08 08:39:28 <russell_> i build with -j17, but retrying now without paralleling
Jun 08 08:39:54 <[florian]> trunk should not be broken with parallel builds, packages/ can be however
Jun 08 08:40:03 <russell_> yeah, they seem okay
Jun 08 08:40:38 <russell_> nope, not even serial, bin just has packages/... no images
Jun 08 08:41:13 <[florian]> strange
Jun 08 08:41:38 <russell_> i'm at r16375.
Jun 08 08:42:20 <russell_> nbd pointed me at make target/install (which i am retrying now) and that seemed to help, but he said that should run with a vanilla make
Jun 08 08:43:36 <russell_> "make BUILD_LOG=1 IGNORE_ERRORS=m V=99 -j17 target/install" created images
Jun 08 08:44:45 <russell_> (building for wgt634u)
Jun 08 09:06:35 <russell_> hmm.  the image i did get (with target/install) fails to boot, stalling at
Jun 08 09:06:35 <russell_>   Please be patient, while OpenWrt loads ...
Jun 08 09:06:35 <russell_>   Algorithmics/MIPS FPU Emulator v1.5
Jun 08 09:12:26 <CIA-29> marek * r16376 /packages/net/ (7 files in 5 dirs): [batman] removing batman-adv userspace as it is not supported anymore (see: http://www.open-mesh.net/changeset/1283)
Jun 08 11:31:19 <CIA-29> florian * r16377 /trunk/target/linux/generic-2.6/patches-2.6.28/ (208-rtl8110sb_fix.patch 980-vm_exports.patch): [kernel] refresh 2.6.28 patches
Jun 08 11:59:42 <CIA-29> jow * r16378 /trunk/package/nvram/ (Makefile src/nvram.c): [package] nvram: refuse nvram operations on the WGT634U
Jun 08 12:01:00 <CIA-29> jow * r16379 /branches/8.09/package/nvram-brcm47xx/ (. Makefile src/nvram.c): Merge r16378 to 8.09
Jun 08 13:42:13 <CIA-29> florian * r16380 /trunk/target/linux/brcm63xx/files/drivers/pcmcia/bcm63xx_pcmcia.c: [brcm63xx] remove this line which got accidentaly succesfully applied, but should not have (#5285)
Jun 08 14:15:27 <CIA-29> florian * r16381 /trunk/target/linux/brcm47xx/ (3 files in 2 dirs): [brcm47xx] add watchdog driver (#5238)
Jun 08 15:50:55 <moonflux> Kaloz: ping
Jun 08 15:51:45 <moonflux> for the broken redirection for the wiki, something like this http://pastebin.ca/1451970 should work
Jun 08 18:49:34 <Bartman007> w/in 33
Jun 08 18:51:38 <[florian]> you won
Jun 08 18:52:11 <Bartman007> not sure what prize 33 is though.
Jun 08 18:56:06 <danage> what did you win?
Jun 08 18:57:53 <maligor> z/out?
Jun 08 19:14:43 <danage> [florian]: should i hook up the router now?
Jun 08 19:15:12 <CIA-29> florian * r16382 /packages/net/hiawatha/Makefile: [package] update hiawatha to 6.14.1 (#5307)
Jun 08 19:16:41 <[florian]> danage: not yet
Jun 08 19:16:47 <[florian]> danage: reviewing the system code atm
Jun 08 19:17:19 <russell_> hmm.  my non-image-building behaviour might be due to image size?
Jun 08 19:17:58 <danage> [florian]: cool, i have some other stuff to do as well, so that's good
Jun 08 19:18:16 <xMff> russell_: quite possible
Jun 08 19:18:39 <xMff> if nothing else fails...
Jun 08 19:18:56 <russell_> just did a much more vanilla-ish .config and it built
Jun 08 19:19:20 <russell_> only difference is "m" vs "y" on some packages
Jun 08 19:19:44 <xMff> then you hit a size constraint
Jun 08 19:20:37 * russell_ wonders if it's something fundamental, or just an arbitrary limitation.  wgt has 8 meg of flash.
Jun 08 19:21:55 <Bartman007> russell_: CFE on some devices can't handle images over ~5.8MB
Jun 08 19:22:33 <russell_> i know about the CFE flashing limitation, i was using mtd in this case
Jun 08 19:27:53 <[florian]> danage: could you try the following patch: http://openwrt.pastebin.com/m74f91704
Jun 08 19:29:31 <[florian]> danage: what happens when you load an openwrt kernel with your 6345 device? nothing shows up right?
Jun 08 19:35:09 <Bartman007> russell_: the CFE flashing limitation is somewhere around 3MB IIRC
Jun 08 19:35:30 <Bartman007> the limit is currently set at 0x5A0000
Jun 08 19:36:25 <Bartman007> when I tried moving it roughly a year ago, 0x700000 sent my wgt into a reset loops, and 0x600000 caused CFE to hang
Jun 08 19:38:17 <danage> [florian]: correct
Jun 08 19:38:47 <russell_> hmm.  what i noticed was that despite adding packages, i had to force image creation with target/install, the trx image size didn't change, and the boot hung.
Jun 08 19:39:10 <danage> let me set up my testing environment, give me a couple of minutes please
Jun 08 19:39:35 <[florian]> danage: ok, so this might be the cause of the UART not showing anything
Jun 08 19:39:44 <[florian]> danage: instead of enabling the uart clock, we were disabling it
Jun 08 19:39:50 <danage> d'oh
Jun 08 19:41:45 <CIA-29> florian * r16383 /packages/libs/libtorrent/Makefile: [package] update libtorrent and cleanup makefile (#5305)
Jun 08 19:42:19 <CIA-29> florian * r16384 /packages/net/rtorrent/Makefile: [package] update rtorrent and cleanup makefile (#5305)
Jun 08 19:42:58 <russell_> the trx image size was 4198400 bytes
Jun 08 19:43:47 <CIA-29> florian * r16385 /packages/libs/xmlrpc-c/Makefile: [package] update libxmlrpc-c to 1.18.5 (#5304)
Jun 08 19:44:15 <russell_> that's 0x401000
Jun 08 19:46:44 <russell_> the wgt CFE flashing limitation is 0x3C0000 (that big works, anyway)
Jun 08 19:52:31 <danage> [florian]: http://openwrt.pastebin.com/m52062dcb and don't laugh please in case i did something stupid
Jun 08 19:56:19 <[florian]> danage: which revision are you at ?
Jun 08 19:56:38 <[florian]> danage: I think the patch got corrupted while downloading from pastebin.com, can you apply it by hand?
Jun 08 19:56:56 <danage> Checked out revision 16385.
Jun 08 19:57:01 <danage> yeah
Jun 08 19:57:14 <danage> i used the "download" button on pastebin, i'll do it manually
Jun 08 19:57:22 <[florian]> yes, that's a one liner ;)
Jun 08 19:59:18 <[florian]> brb
Jun 08 20:03:41 <russell_> i think maybe IGNORE_ERRORS=m is hiding what's failing for me, trying without
Jun 08 20:07:01 <danage> what command makes me a clean kernel? make clean enough?
Jun 08 20:07:15 <[florian]> danage: make target/linux/clean
Jun 08 20:13:57 <[florian]> danage: does it build?
Jun 08 20:16:09 <CIA-29> florian * r16386 /packages/ipv6/aiccu/files/aiccu.init: [package] fix aiccu heartbeat option parsing, thanks to Malte S. Stretz
Jun 08 20:17:06 <danage> [florian]: yes. do you know the serial bps rating that is used commonly?
Jun 08 20:17:11 <danage> i.e. most likelyx?
Jun 08 20:19:05 <[florian]> danage: 115200 8n1 on all brcm63xx I have seen so far
Jun 08 20:21:54 <danage> argh why am i seeing garbage then
Jun 08 20:26:51 <danage> ok finally. sorry sorry. next time i'll write it down
Jun 08 20:27:28 <danage> it's stilly compiling anyways
Jun 08 20:33:33 <[florian]> danage: no problems
Jun 08 20:35:06 <CIA-29> florian * r16387 /trunk/docs/ (build.tex network-scripts.tex network.tex): [docs] comment and fix some parts of the documentation, thanks to Harsh Trivedi
Jun 08 20:37:09 <russell_> hmm.  no more clue than before.  the last bit in the make output is a series of these:
Jun 08 20:38:11 <russell_> Configuring kmod-ppp...Done.
Jun 08 20:38:11 <russell_> IPKG_TMP=/aux/src/openwrt/tmp/ipkg IPKG_INSTROOT=/aux/src/openwrt/build_dir/target-mipsel_uClibc-0.9.30.1/root-brcm47xx IPKG_CONF_DIR=/aux/src/openwrt/staging_dir/target-mipsel_uClibc-0.9.30.1/etc IPKG_OFFLINE_ROOT=/aux/src/openwrt/build_dir/target-mipsel_uClibc-0.9.30.1/root-brcm47xx /aux/src/openwrt/scripts/ipkg -force-defaults -force-depends install /aux/src/openwrt/bin/packages/target-mipsel_uClibc-0.9.30.1/kmod-pppoe_2.6.28.10-brcm47xx
Jun 08 20:38:15 <russell_> -1_mipsel.ipk
Jun 08 20:39:40 <russell_> oh.  wait.  then I get a "Configuring kmod-pppoe..." and then it exits silently
Jun 08 20:40:26 <[florian]> russell_: building on debian squeeze?
Jun 08 20:40:35 <russell_> debian unstable
Jun 08 20:40:39 <russell_> whatever that is
Jun 08 20:40:47 <russell_> (/me hates debian code names)
Jun 08 20:41:55 <[florian]> that is debian squeeze
Jun 08 20:42:26 <russell_> one modification from vanilla .config i routinely do is to <m> ppp support, but /me realizes now that there are related kmod-ppp things still built in.  trying again without.
Jun 08 20:42:56 <xMff> russell_: you'll need a patch
Jun 08 20:43:07 <russell_> which patch?
Jun 08 20:43:09 <xMff> russell_: http://forum.openwrt.org/viewtopic.php?pid=87427#p87427
Jun 08 20:43:15 * russell_ looks
Jun 08 20:45:23 <russell_> hmm.  interesting
Jun 08 20:46:10 <[florian]> danage: any updates ?
Jun 08 20:46:25 <danage> still compiling
Jun 08 20:58:52 <danage> [florian]: flashing...
Jun 08 21:01:20 <danage> [florian]: Starting program at 0x80010000, nothing more. how can i verify the patch got applied correctly (the ~) and i got it put into the kernel?
Jun 08 21:02:07 <[florian]> danage: you can verify that the sources got patched in build_dir/linux-brcm63xx/linux-2.6.27.24/arch/mips/bcm63xx/prom.c
Jun 08 21:02:38 <[florian]> danage: I may have a slighlty different approach
Jun 08 21:03:38 <CIA-29> florian * r16388 /packages/net/transmission/Makefile: [package] update to transmission 1.71 (#5292)
Jun 08 21:06:06 <[florian]> danage: let me check something
Jun 08 21:06:57 <danage> [florian]: this for verification purposes: http://openwrt.pastebin.com/m4853114d
Jun 08 21:08:47 <[florian]> danage: can you try that one: http://openwrt.pastebin.com/m5e7fe32d ?
Jun 08 21:09:40 <[florian]> danage: basically, what I think is happening, is either the CPU is not correctly detected, or the UART is disabled too quickly
Jun 08 21:11:45 <[florian]> danage: I will use a jtag probe tomorrow I think
Jun 08 21:14:58 <CIA-29> florian * r16389 /packages/admin/muninlite/ (3 files in 2 dirs): [package] update muninlite to 1.0.2, fix xined dependency (#5291)
Jun 08 21:18:01 <CIA-29> lars * r16390 /trunk/scripts/ipkg:
Jun 08 21:18:01 <CIA-29> Workaround a bug(?) in tar in debian squeeze, which causes the build process to
Jun 08 21:18:01 <CIA-29> fail without any reason.
Jun 08 21:18:22 <danage> [florian]: ok this may sound silly, but i dont seem to manage to apply any of these patches using the patch command
Jun 08 21:18:28 <danage> how do you usually do it?
Jun 08 21:19:18 <[florian]> danage: wget -O bcm63xx.patch "http://openwrt.pastebin.com/pastebin.php?dl=m5e7fe32d"
Jun 08 21:19:24 <[florian]> patch -p0 --dry-run < bcm63xx.patch
Jun 08 21:19:29 <[florian]> to make sure that it applies cleanly
Jun 08 21:19:35 <[florian]> if it does, patch -p0 < bcm63xx.patch
Jun 08 21:20:29 <danage> see, that's what i did, same crud about ending in the middle of the line
Jun 08 21:20:45 <[florian]> humm, I had the same problem indeed
Jun 08 21:21:37 <danage> i had used wget initially too, thinking i was all smart
Jun 08 21:24:58 <CIA-29> florian * r16391 /packages/admin/muninlite/patches/002-hostname.patch: [package] add missing muninlite patch after 1.0.2 update
Jun 08 21:29:12 <danage> [florian]: should i patch it in manually?
Jun 08 21:29:29 <[florian]> danage: if you can, please do
Jun 08 21:29:41 * danage can
Jun 08 21:29:44 <danage> :)
Jun 08 21:29:44 <[florian]> danage: otherwise, I will just make the board back online tonight
Jun 08 21:29:50 <[florian]> danage: I will anyway ;)
Jun 08 21:32:15 <danage> make target/linux/clean ?
Jun 08 21:35:42 <russell_> [florian], xMff: patch seems to have fixed me. thx!
Jun 08 21:36:02 <xMff> russell_: np, it's in svn now
Jun 08 21:50:57 <danage> [florian]: flashing...
Jun 08 21:52:22 <danage> [florian]: http://openwrt.pastebin.com/m12eca5e8
Jun 09 01:35:25 <Weedy> so uhh
Jun 09 01:35:41 <Weedy> anyone here work on ubiquity hardware?
**** ENDING LOGGING AT Tue Jun 09 02:59:57 2009