**** BEGIN LOGGING AT Tue Jun 21 02:59:56 2011 Jun 21 04:04:38 swalker: ping Jun 21 10:07:53 florian * r27246 /trunk/target/linux/brcm63xx/ (patches-2.6.39/240-spi.patch patches-3.0/240-spi.patch): [brcm63xx] use macros to generate SPI registers table and switches Jun 21 11:49:38 <^Willie^> wtf? ubicom.com is not listed in google's dns anymore :S Jun 21 11:50:18 <^Willie^> hmm dig command results in finding the records normal ping requests fail to lookup there hostname .. weird Jun 21 13:06:55 philipp64|laptop: pingf Jun 21 13:25:15 nbd: did you find out sth. regarding the conditional-dependency-issue? Jun 21 13:28:11 mirko: not yet Jun 21 16:18:40 mirko: apparently miniperl needs -lbsd, but I can't figure out how to get it passed in. Jun 21 16:25:40 build #55 of s3c24xx is complete: Failure [failed compile_2] Build details are at http://buildbot.openwrt.org:8010/builders/s3c24xx/builds/55 Jun 21 16:43:04 nbd mirko I did a clean build 2 days ago trunk with eglibc for wzr300. Hotplug had still one build problem because libbsd that it required for eglibc/glibc builds did not get build. Following helped me: http://pastie.org/2102210 Jun 21 16:46:58 lamikr: did you look at r27209 ? Jun 21 16:48:45 philipp64|laptop from where? Jun 21 16:49:23 oh, hang on.... you're saying 27209 was the damage... btw, you can write: ifneq ($(CONFIG_USE_EGLIBC)$(CONFIG_USE_GLIBC),) as well. Jun 21 16:49:35 "svn diff -c27209" Jun 21 16:51:15 * lamikr wondering can I find svn rev numbers somehow when using git with openwrt... Jun 21 16:51:45 you're on your own. Jun 21 16:52:15 nbd: I've got 6 patches pending... can I please get some commits? Jun 21 16:56:45 lamikr: in the log Jun 21 16:56:50 in the commit messages Jun 21 16:57:12 philipp64|laptop: maybe you should use git-svn and push your tree somewhere Jun 21 16:59:04 nbd Ok, thanks found now. Checked and the syntax "DEPENDS:=+USE_EGLIBC:libbsd +USE_GLIBC:libbsd" from 27209 did not work for me at least in this hotplug2 Makefile. Jun 21 17:00:24 yeah, there's still a bug there Jun 21 17:01:20 nbd I assume you fix it faster than I wrote send a patch... Jun 21 17:01:39 haven't tracked down the exact source of the bug yet Jun 21 17:12:17 nbd: the problem is that that wouldn't help me... I'm working with other people who need to see my changes in trunk when they pull. Jun 21 17:13:49 plus it saves mirko and me the effort of both of us simultaneously working to fix the same bugs... Jun 21 17:20:46 build #55 of at91 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/55 Jun 21 17:43:52 build #53 of ubicom32 is complete: Failure [failed compile_3] Build details are at http://buildbot.openwrt.org:8010/builders/ubicom32/builds/53 Jun 21 17:45:20 what is the nominal setup for openwrt + https + certificate? Jun 21 18:41:52 hauke * r27247 /trunk/target/linux/generic/ (14 files in 2 dirs): kernel: refresh patches for kernel 3.0-rc4 Jun 21 19:22:42 florian * r27248 /packages/utils/pcmciautils/Makefile: [package] pcmciautils: depends on libsysfs, not sysfsutils Jun 21 19:22:49 florian * r27249 /packages/utils/pcmciautils/Makefile: [package] pcmciautils: bump to 018 Jun 21 19:51:40 OMG THE BUILDSLAVES ARE BACK UP Jun 21 19:51:40 \o/ Jun 21 19:51:54 who do I send the beer to? Jun 21 20:45:55 hauke * r27250 /trunk/package/mac80211/Makefile: mac80211: remove rests of ar9170 driver Jun 21 20:46:55 hauke * r27251 /trunk/package/mac80211/Makefile: Jun 21 20:46:55 mac80211: fix debugfs for ath9k_htc Jun 21 20:46:55 if debugfs was enabled ath9k_htc did not load. Jun 21 20:52:16 hauke * r27252 /trunk/package/mac80211/Makefile: (log message trimmed) Jun 21 20:52:16 mac80211: use firmware from linux-firmware git Jun 21 20:52:16 Use the firmware from linux-firmware git where possible. Jun 21 20:52:16 Update some firmware files in that process and check for missing firmware files for some modules. Jun 21 20:52:16 * update Ralink firmwares Jun 21 20:52:16 * update ath9k_htc firmware Jun 21 20:52:17 * add firmware for mwl8k Jun 21 20:54:58 florian * r27253 /trunk/ (2 files in 2 dirs): Jun 21 20:54:58 [kernel] fix and make usable kmod-pcmcia-rsrc Jun 21 20:54:58 This module is also needed for e.g: kmod-bcm63xx-pcmcia Jun 21 20:55:03 florian * r27254 /trunk/target/linux/brcm63xx/ (2 files in 2 dirs): [brcm63xx] fix tab damage in 141-led_count.patch Jun 21 21:04:47 nbd: ordering MIPS firmware for bdi3000 today. Jun 21 21:07:01 sbak, how much $$ ? Jun 21 21:10:44 obinou: 1000$ Jun 21 21:11:05 i have a bdi3000 with arm9 firmware and PPC firmware, no MIPS though :( Jun 21 21:11:59 ouch Jun 21 21:12:07 $1000 juste for the piece of code Jun 21 21:12:31 (I suppose you already paid taht much for ARM & PPC, not couting the hardware...) Jun 21 21:12:56 it is a racket! but i guess they are running a business. Jun 21 21:13:00 A clear show-stopper for a small company, let alone an individual Jun 21 21:13:50 (In fact I use sub-$100 JTAG probes. Less expensive, but slower & more complicated to work with ) Jun 21 21:14:19 obinou: what do you use? i'd be interested for personal use. Jun 21 21:15:55 does it have an ICE? Jun 21 21:16:04 hauke * r27255 /trunk/tools/firmware-utils/src/trx2edips.c: Jun 21 21:16:04 firmware-utils: fix build on big endian systems Jun 21 21:16:04 This closes #9446 Jun 21 21:16:46 I use Swiss-based Amontec probes : http://www.amontec.com/jtagkey2.shtml (Avoid the tiny version , which had signal-stability problems). Jun 21 21:18:36 obinou: do you use openOCD with that? does it allow you to step through kernel code? Jun 21 21:18:37 praise the almighty bdi3000 ;) Jun 21 21:19:26 sbak, yes, and yes Jun 21 21:19:51 it's more complicated than with the bdi3000 though Jun 21 21:20:14 you have to load the .elf of the kernel into gdb Jun 21 21:20:49 then load the kernel on the plateform, use jtag to set a breakpoint, and here you go Jun 21 21:21:17 but you often need to modify openocd to suit your needs, depending on the plateform Jun 21 21:21:55 well for the price thats a deal. I got work to get the bdi3000, but couldn't afford 2400$ for a bdi3000 for personal use. Jun 21 21:22:08 thanks for the link Jun 21 21:22:43 here is the equivalent, from a canadian company : http://www.xverve.com/ Jun 21 21:22:55 (saves the shipping & taxes if you are in the us) Jun 21 21:23:16 just, be extra-carefull of the I/O voltage Jun 21 21:23:25 thank you! Jun 21 21:23:35 i fried some chips with that Jun 21 21:23:53 don't expect magic, thought: The speed is nothing like the bdi Jun 21 21:24:30 (in genera, at 6MHZ it take a good 20Mn to flash a kernel on a NAND Jun 21 21:24:42 yea, we had a bdi2000 and bdi3000 just lying on a shelf! Jun 21 21:24:48 lucky me Jun 21 21:25:00 however, for debugging it's ok - the speed is not an issue Jun 21 21:25:54 sbak, it says 'arm-jtag' because openocd is mainly arm-focused Jun 21 21:26:16 but I used it for MIPS and ST-Micro and atmel as well Jun 21 21:26:21 i was going to ask. so openOCD needs MIPS tweaking as you stated, correct? Jun 21 21:27:13 isn't there some JTAG plugin for eclipse that i can use the BDI3000 with? Jun 21 21:28:20 ah, LinuxScope JTD Jun 21 21:28:26 Yes, openocd is not as great for mips as it is for jtag. There are some patch lying around , throught. A bit time-consuming , less than $1000 in my opinion though Jun 21 21:29:10 Concerning eclipse , i don't know - I rerely use it (not enought ram on my laptop, and the screen is too small) Jun 21 21:29:31 I just use gdb and a bunch of xterm :-) Jun 21 21:31:33 figures... ive done all my embedded devel with codewarrior, rowley cross studio, and code composer studio. having an IDE to look at the call stack and registeres is all i know. Jun 21 21:31:52 is it easy to acomplish this with gdb? Jun 21 21:32:16 well its probably as cryptic as changing from eclipse to emacs Jun 21 21:32:23 lol Jun 21 21:32:51 yes you can do that, but you need to use learn some commands Jun 21 21:32:56 or sue a graphical frontend for gdb Jun 21 21:33:00 *use Jun 21 21:34:27 i've used DDD - i'd rather just learn gdb. I suppose the time it takes will pay for iteslf Jun 21 21:34:42 yes Jun 21 21:36:26 sbak, maybe with DDD it's possible Jun 21 21:36:49 I remember that there is Insight that also has a gdb front-end Jun 21 21:37:04 http://beej.us/guide/bggdb/ Jun 21 21:37:10 its a good howto imho Jun 21 21:37:10 and , of course, eclipse. Now, will il work with jtag.... I don't know. Jun 21 21:37:18 very basic but gives a good overview Jun 21 21:38:47 sbak : However, I feel it's *clear* than using such USB-Jtag probes + gdb + a gui (eclispe) is clearly more difficult to setup than the IDE you mentionned Jun 21 21:39:23 it will require more time .... the price to pay to have cheap JTAG probes :-( Jun 21 21:46:47 blogic: thanks for the link. im gonna get my hands dirty and read up on gdb tonight. I should have the mips firmware tomorrow provided they allow me to download. hopefully the abatron reseller doesn't mail it!!! Jun 21 21:47:21 thanks for the help, dinner time! Jun 21 22:04:49 good night Jun 22 00:51:35 <^Willie^> hmm must be luck.. write a makefile for perl-json and it works out well in just one shot :) (must be luck) Jun 22 02:10:20 KanjiMonster,acoul:Are you around? Jun 22 02:11:13 ehci_ar71xx compile errors when building 3.0 Jun 22 02:11:14 http://pastebin.com/i7j3ENQz **** ENDING LOGGING AT Wed Jun 22 02:59:57 2011