**** BEGIN LOGGING AT Mon Jul 25 02:59:57 2011
Jul 25 03:50:15 <RealOpty> hey
Jul 25 03:50:18 <RealOpty> anyone around?
Jul 25 03:50:25 <RealOpty> jow_laptop, ping
Jul 25 04:00:05 <RealOpty> anyways question was... is the kernel at 3.0 now not 3.0-rc7?
Jul 25 06:28:18 <ixew> Hi, is LuCI working in openwrt-trunk? Thanks.
Jul 25 06:35:14 <CIA-89> nbd * r27764 /trunk/package/mac80211/patches/420-mac80211_ignore_invalid_ccmp_rx_pn.patch: mac80211: remove the ccmp rx pn workaround, it should no longer be necessary and it was reported to cause connection stability issues (see #9646)
Jul 25 06:36:19 <CIA-89> nbd * r27765 /branches/backfire/package/mac80211/patches/420-mac80211_ignore_invalid_ccmp_rx_pn.patch: mac80211: remove the ccmp rx pn workaround, it should no longer be necessary and it was reported to cause connection stability issues (backport of r27764, see #9646)
Jul 25 06:49:26 <philipp64|laptop> nbd: ping
Jul 25 06:50:24 <nbd> philipp64|laptop: pong
Jul 25 06:50:55 <philipp64|laptop> I've got a couple of new patches... they're all trivial... including a bump to 2.6.39.3.
Jul 25 06:51:15 <philipp64|laptop> kernel.org calls out latest/stable as 2.6.39.3.
Jul 25 06:53:17 <nbd> yes, i've seen the patches
Jul 25 06:53:18 <solca> philipp64|laptop: are you updating all targets in 2.6.39 to 2.6.39.3?
Jul 25 06:53:42 <solca> FYI ar71xx runs fine on 2.6.39.3
Jul 25 06:54:03 <nbd> his patch doesn't actually update the kernel to 2.6.39.3
Jul 25 06:54:08 <nbd> it only changes the md5sum
Jul 25 06:54:26 <solca> hmm that's not much :)
Jul 25 06:54:30 <nbd> indeed
Jul 25 06:56:43 * solca compiling ar71xx to kernel 3.0 with patches from 2.6.39
Jul 25 07:12:24 * solca is tired of refreshing/adjusting patches manually
Jul 25 07:12:41 <solca> there must be a better way...
Jul 25 07:13:02 <blogic> solca: quilt ?
Jul 25 07:13:42 <solca> blogic: can you be very kind to give me an example?
Jul 25 07:14:35 <blogic> solca: cd build_dir/linux-XYZ/linux-3.0/
Jul 25 07:14:45 <blogic> quilt push/refresh
Jul 25 07:14:47 <blogic> cd -
Jul 25 07:14:51 <blogic> make target/update
Jul 25 07:15:05 <blogic> well push -f and then fix the patches
Jul 25 07:15:21 <blogic> owrt automatically puts the patch series into quilt
Jul 25 07:17:35 <solca> blogic: great! that seems much better
Jul 25 07:17:49 <solca> will try it now...
Jul 25 07:20:47 <solca> oh I was never interested in quilt until it makes my life easier
Jul 25 07:20:49 <solca> blogic: thx!
Jul 25 07:58:46 <solca> blogic: I fixed a patch but 'quilt refresh' still gives error
Jul 25 07:58:54 <solca> applying manually works
Jul 25 08:02:01 <blogic> what error does it give
Jul 25 08:03:04 <loswillios> btw, there's a quilt guide on the wiki http://wiki.openwrt.org/doc/devel/patches?s[]=quilt#iteratively.modify.patches.without.cleaning.the.source.tree
Jul 25 08:03:51 <solca> blogic: http://pastebin.com/h81QRsES
Jul 25 08:04:21 <solca> loswillios: sure, I'm reading docs too, but practicing on the go... ;)
Jul 25 08:04:52 <loswillios> ok
Jul 25 08:06:51 <jow_laptop> solca: quilt push -f  --  that will create a .rej file, manually edit drivers/mtd/devices/m25p80.c, then refresh
Jul 25 08:08:00 <solca> jow_laptop: hmm I fixed the patch file 102-...
Jul 25 08:11:18 <solca> jow_laptop: excellent, as you said it's now refreshed
Jul 25 08:12:33 <solca> holly cow, it seems every patch is borked by tabs :(
Jul 25 08:56:15 <solca> blogic, jow_laptop, loswillios: I screwed up the patches (it contains timestamps, etc)
Jul 25 08:56:29 <solca> as I didn't put the required options in ~/.quiltrc
Jul 25 08:56:49 <solca> can I regenerate all patches again?
Jul 25 09:05:53 <jow_laptop> make target/linux/refresh
Jul 25 09:11:54 <solca> thx! now that patches apply nicely time to check why it fails to build
Jul 25 09:12:11 <loswillios> ;-)
Jul 25 09:16:02 <solca> hmm it seems ehci-ar71xx.c will need updating to 3.0 APIs :(
Jul 25 09:18:29 <KanjiMonster> solca: that's an easy fix
Jul 25 09:19:58 <solca> KanjiMonster: really? where to? :)
Jul 25 09:20:15 <nbd> solca: btw. here's another nice intro to quilt: http://www.suse.de/~agruen/quilt.pdf
Jul 25 09:21:42 <KanjiMonster> solca: see how I fixed it for bcm63xx, should give you a hint how to do it -> http://patchwork.openwrt.org/patch/1103/
Jul 25 09:22:54 <solca> nbd: thx! will read later as I'm absorbed now reading kernel headers ;)
Jul 25 09:23:08 <solca> KanjiMonster: great that's a good start, thx!
Jul 25 09:30:33 <solca> KanjiMonster: indeed you fix the HC_LENGTH params
Jul 25 09:31:57 <solca> I know the fix but how I refresh the patch 003-ar71xx_usb_host.patch with my fix with quilt?
Jul 25 09:32:08 <solca> it seems I have to read more about quilt...
Jul 25 09:33:15 <KanjiMonster> solca: http://wiki.openwrt.org/doc/devel/patches#adding.or.editing.kernel.patches
Jul 25 09:33:46 <KanjiMonster> ah, I meant the one above
Jul 25 09:34:39 <solca> ok, reading...
Jul 25 09:35:47 <solca> ohh this is so cool!
Jul 25 09:35:51 <solca> KanjiMonster: thx!
Jul 25 09:41:20 <CIA-89> jow * r27766 /branches/backfire/target/linux/generic-2.6/patches-2.6.32/975-ssb_update.patch: [backfire] refresh 2.6.32 patches
Jul 25 09:41:59 <solca> WTF? Patch 003-ar71xx_usb_host.patch is not in series
Jul 25 09:43:10 <solca> it had to be prepended with platform/
Jul 25 09:43:48 <jow_laptop> correct
Jul 25 09:43:54 <jow_laptop> happens to me too all the time
Jul 25 09:44:08 <jow_laptop> in fact I just lost some newly added aptches due to a forgotten target/linux/update ;)
Jul 25 09:46:19 <KanjiMonster> jow_laptop: that sounds like me doing a git reset --hard without doing a git format-patch first, which I tend to do from time to time ;)
Jul 25 09:46:34 <solca> jow_laptop: hehe, hopefully you can recreate them from memory :)
Jul 25 09:46:57 <nbd> KanjiMonster: git reset is not a big deal
Jul 25 09:47:01 <nbd> there's always git reflog ;)
Jul 25 09:47:50 <KanjiMonster> oh, that's good to know
Jul 25 09:52:21 <CIA-89> jow * r27767 /branches/backfire/target/linux/ (2 files in 2 dirs):
Jul 25 09:52:21 <CIA-89> [backfire] backport upstream commit f2df1ae3fe8d44d51bd7218a9da9134b83c7531e
Jul 25 09:52:21 <CIA-89> Adds support for the Spansion S25FL016K and S25FL064K SPI flash chips.
Jul 25 09:57:42 <solca> KanjiMonster: weird 'quilt diff' shows changes to other things not appropiate to my patch that I didn't
Jul 25 09:58:21 <CIA-89> jow * r27768 /branches/backfire/target/linux/ar71xx/files/drivers/mtd/wrt160nl_part.c: [backfire] backport r27049, should fix newer WRT160NL models with Backfire
Jul 25 10:01:09 <solca> oh no, almost 4AM, 'goto bed;'
Jul 25 10:01:26 <solca> will continue later...
Jul 25 10:02:37 <CIA-89> nbd * r27769 /trunk/package/mac80211/patches/ (2 files): mac80211: rename a misnamed patch
Jul 25 10:02:52 <CIA-89> jow * r27770 /branches/packages_10.03.1/lang/php5/ (Makefile patches/950-Fix-dl-cross-compiling-issue.patch): [packages_10.03.1] backport r27763
Jul 25 10:20:14 <solca> holy s... it builds!!!
Jul 25 10:20:27 <solca> I could not make it to bed w/o checking... ;)
Jul 25 10:20:58 <solca> now the sad part is my changes end in the last patch in the whole series: 901...
Jul 25 10:38:53 <CIA-89> nbd * r27771 /trunk/ (include/kernel.mk rules.mk): only export GCC_HONOUR_COPTS for the built-in toolchain, fixes missing-fhonour-copts warnings with external toolchains
Jul 25 11:32:37 <tripolar> hi
Jul 25 11:33:04 <tripolar> when i use 3.0 the switch module seems to have a wrong symbol name - all swtich modules load except the switch_adm: Unknown symbol ssb_bcm47xx (err 0)
Jul 25 11:54:10 <tripolar> ping nbd
Jul 25 12:04:13 <blaise> shweet!
Jul 25 12:04:15 <blaise> XD
Jul 25 12:05:10 <blaise> jow_laptop: you think maybe I should try a different version of GCC ?
Jul 25 12:05:25 <jow_laptop> blaise: no, gcc is not the issue
Jul 25 12:05:35 <jow_laptop> its the gnu script crap around
Jul 25 12:05:43 <blaise> jow_laptop: what is this GNU autofail you speak of?
Jul 25 12:05:53 <blaise> lol
Jul 25 12:05:55 <jow_laptop> configure / libtool / autoconf / pkg-config
Jul 25 12:06:01 <blaise> ah.....
Jul 25 12:06:45 <blaise> now that I know where development convo belongs... I'll speak about it in here..
Jul 25 12:06:49 <tripolar> jow_laptop: i get this error switch_adm: Unknown symbol ssb_bcm47xx (err 0) when loading the swtich-adm module
Jul 25 12:06:52 <tripolar> with 3.0
Jul 25 12:06:59 <jow_laptop> tripolar: don't use 3.0 then
Jul 25 12:07:31 <jow_laptop> blaise: can you upload build_dir/target-mips_r2_uClibc-0.9.32/libotr-/libotr-3.2.0/config.log  somewhere?
Jul 25 12:07:46 <blaise> jow_laptop: absolutly.. :)
Jul 25 12:08:45 <jow_laptop> blaise: also, is there a "libgcrypt-config" command on your host?
Jul 25 12:10:05 <blaise> yes..
Jul 25 12:10:31 <tripolar> jow_laptop: id like to fix it - how can i find out why this happens?
Jul 25 12:10:55 <jow_laptop> blaise: ok then I don't need config.log
Jul 25 12:11:07 <jow_laptop> blaise: add the following to your Makefile:
Jul 25 12:11:20 <jow_laptop> blaise:  CONFIGURE_ARGS += --with-libgcrypt-prefix=$(STAGING_DIR)/usr
Jul 25 12:12:51 <jow_laptop> tripolar: the bcm47xx setup code was restructured recently, I suspect the "ssb_bcm47xx" symbol is not exported anymore. You should use bcm47xx_bus instead
Jul 25 12:13:06 <jow_laptop> tripolar: ... and treat it accordingly as either bcma or ssb bus
Jul 25 12:13:43 <jow_laptop> tripolar: depeding on the state of bcm47xx_active_bus_type which can be one of BCM47XX_BUS_TYPE_SSB, BCM47XX_BUS_TYPE_BCMA
Jul 25 12:15:07 <tripolar> thanks
Jul 25 12:15:26 <blaise> jow_laptop: oh wow, I just managed to upload it to pastebin.. http://pastebin.com/3HN0A230
Jul 25 12:15:29 <blaise> lol
Jul 25 12:15:34 <tripolar> just found the patch in generic
Jul 25 12:17:18 <nbd> tripolar: pong
Jul 25 12:18:00 <tripolar> nbd: i allready talked to jow_laptop - the switch-adm module needs a fix to use the new exported symbols for brcm47xx
Jul 25 12:20:19 <blaise> jow_laptop: after adding that configure argument, the build output looks a lot more sane now, thank you..
Jul 25 12:21:51 <blaise> never mind
Jul 25 12:22:04 <blaise> I think I added that to the Makefile wrong or something
Jul 25 12:23:37 <jow_laptop> blaise: new error?
Jul 25 12:24:02 <blaise> no, I didn't add it to the Makefile properly
Jul 25 12:24:16 <jow_laptop> ah ok
Jul 25 12:24:21 <blaise> I guess I need to create a new define.. right?
Jul 25 12:24:29 <jow_laptop> no
Jul 25 12:24:32 <blaise> define Build/Configure ?
Jul 25 12:24:45 <jow_laptop> just stick it somewhere after the include of package.mk
Jul 25 12:25:11 <jow_laptop> not indented, directly at the beginning of a line
Jul 25 12:25:16 <jow_laptop> and not within any define
Jul 25 12:25:25 <jow_laptop> like the PKG_SOURCE et.c variables
Jul 25 12:26:35 <blaise> $include $(INCLUDE_DIR)/package.mk is stated after all the PKG_*
Jul 25 12:26:55 <jow_laptop> yes
Jul 25 12:27:07 <jow_laptop> and after that  add CONFIGURE_ARGS += ...
Jul 25 12:27:12 <tripolar> jow_laptop: but how can i test in an #if - if an veriable has a value? just the same as i would in an normal if?
Jul 25 12:27:12 <blaise> you want me to put this after the include? CONFIGURE_ARGS:=--with-libgcrypt-prefix=$(STAGING_DIR)/usr
Jul 25 12:27:26 <blaise> ah, +=
Jul 25 12:27:29 <jow_laptop> not :=
Jul 25 12:27:31 <jow_laptop> but +=
Jul 25 12:27:33 <jow_laptop> exactly
Jul 25 12:27:37 <jow_laptop> overwrite vs. append
Jul 25 12:28:24 <blaise> revisiting make package/libotr/{clean,compile} V=99 now
Jul 25 12:28:28 <blaise> :)
Jul 25 12:28:34 <tripolar> is somethink like this possible?
Jul 25 12:28:34 <tripolar> #if bcm47xx_active_bus_type == BCM47XX_BUS_TYPE_SSB
Jul 25 12:28:35 <tripolar> #define ssb bcm47xx_bus.ssb
Jul 25 12:28:35 <tripolar> #endif
Jul 25 12:29:18 <jow_laptop> no
Jul 25 12:29:28 <jow_laptop> you must do it at runtime
Jul 25 12:30:10 <jow_laptop> e.g.  static inline struct ssb_device * get_ssb(void) {
Jul 25 12:30:15 <jow_laptop> #ifdef CONFIG_SSB
Jul 25 12:30:29 <jow_laptop>   if (bcm47xx_active_bus_type == BCM47XX_BUS_TYPE_SSB)
Jul 25 12:30:34 <tripolar> okay
Jul 25 12:30:39 <jow_laptop>         return bcm47xx_bus.ssb;
Jul 25 12:30:41 <jow_laptop> #endif
Jul 25 12:30:48 <jow_laptop>     return NULL;
Jul 25 12:30:49 <jow_laptop> }
Jul 25 12:31:14 <jow_laptop> or something like this
Jul 25 12:31:20 <tripolar> okay thanks
Jul 25 12:31:31 <nbd> does bcm47xx still not have gpioapi support?
Jul 25 12:32:46 <jow_laptop> apparently not
Jul 25 12:32:54 <nbd> oh, wait
Jul 25 12:33:03 <nbd> it does have some gpio api support
Jul 25 12:34:46 <blaise> jow_laptop: dodm
Jul 25 12:34:55 <blaise> oops,
Jul 25 12:35:12 <blaise> jow_laptop: that configure flag didn't change anything afaict
Jul 25 12:36:56 <jow_laptop> grep -r ac_cv_path_LIBGCRYPT_CONFIG build_dir/target-mips_r2_uClibc-0.9.32/libotr-/libotr-3.2.0/config.log
Jul 25 12:37:47 <blaise> ac_cv_path_LIBGCRYPT_CONFIG=/home/blaise/openwrt/trunk/staging_dir/target-mips_r2_uClibc-0.9.32/usr/bin/libgcrypt-config
Jul 25 12:38:53 <jow_laptop> ok
Jul 25 12:38:58 <jow_laptop> can you execute that path?
Jul 25 12:39:13 <blaise> yes..
Jul 25 12:39:19 <jow_laptop> what does it show?
Jul 25 12:39:35 <blaise> the same as the executable on the host system
Jul 25 12:40:05 <jow_laptop> so -I/usr/include and other stuff?
Jul 25 12:40:33 <blaise> eh?
Jul 25 12:40:41 <blaise> wait, what do you mean execute the path?
Jul 25 12:40:48 <blaise> I thought you ment the file..
Jul 25 12:41:06 <jow_laptop> you can run it as command?
Jul 25 12:41:21 <blaise> this? /home/blaise/openwrt/trunk/staging_dir/target-mips_r2_uClibc-0.9.32/usr/bin/libgcrypt-config
Jul 25 12:41:24 <jow_laptop> yes...
Jul 25 12:41:38 <blaise> which argument are you wanting me to pass to it?
Jul 25 12:42:21 <jow_laptop> --cflags
Jul 25 12:42:44 <blaise> -I/home/blaise/openwrt/trunk/staging_dir/target-mips_r2_uClibc-0.9.32/usr/include
Jul 25 12:43:07 <jow_laptop> odd
Jul 25 12:43:14 <jow_laptop> and whats the compile failure?
Jul 25 12:43:40 <blaise> otr_modify.o: In function `main':
Jul 25 12:43:40 <blaise> otr_modify.c:(.text+0xec): undefined reference to `__isoc99_sscanf'
Jul 25 12:44:08 <blaise> then ld dumps
Jul 25 12:44:14 <jow_laptop> I need the full log
Jul 25 12:44:31 <jow_laptop> the error you see is a consequence of a fault happening way earlier
Jul 25 12:47:58 <blaise> just a few molments.. piping output to file now..
Jul 25 12:54:07 <blaise> hrmm
Jul 25 12:54:47 <blaise> the log i made didn't include errors.. for some reason the errors just came on out to console and didn't go to the file
Jul 25 12:54:51 <blaise> lol
Jul 25 12:55:52 <blaise> and the build.log in logs dir isn't being created...
Jul 25 12:59:59 <blaise> jow_laptop: where can I find the build log?
Jul 25 13:00:07 <jow_laptop> make package/libotr/{clean,compile} V=99 > log.txt
Jul 25 13:00:35 <blaise> I did >> log.txt
Jul 25 13:00:49 <jow_laptop> and what does it contain?
Jul 25 13:00:53 <jow_laptop> just paste it
Jul 25 13:01:12 <blaise> it contained the make statements, but the errors didn't pipe to the file..
Jul 25 13:01:25 <jow_laptop> make package/libotr/{clean,compile} V=99 2>&1 >log.txt
Jul 25 13:02:40 <blaise> there are still some things bleeding through
Jul 25 13:02:57 <blaise> three instances of "cat: ld.so.conf.d/*.conf: No such file or directory" so far
Jul 25 13:10:58 <blaise> jow_laptop: http://pastebin.com/M84WTsiK
Jul 25 13:14:15 <jow_laptop> blaise: can you paste the generated Makefile as well please?
Jul 25 13:14:22 <jow_laptop> in the same dir as config.log
Jul 25 13:17:02 <blaise> jow_laptop: http://pastebin.com/KnkSDDFC
Jul 25 13:17:35 <^Willie^> jow_laptop: morning, when i build a openwrt image using eglibc i'm missing the links to /usr/lib mosth packages do build there libs in there except luci and the library path is then also not in the userland profile.. any idea wat i forgot ?
Jul 25 13:18:18 <^Willie^> simply link from usr/lib to /lib is not working cp the content make everything going to work as temp fix :( but i dunno if there is some menuconfig setting for this ?
Jul 25 13:21:35 <blaise> wish libotr wasn't being so troublesome...
Jul 25 13:29:27 <jow_laptop> gah libtool
Jul 25 13:29:41 <blaise> loool
Jul 25 13:29:56 <blaise> so, libtool is the culpret?
Jul 25 13:30:00 <jow_laptop> please try adding this next to the other PKG vars:
Jul 25 13:30:04 <jow_laptop> PKG_FIXUP:=autoreconf
Jul 25 13:30:11 <blaise> XD
Jul 25 13:30:39 <jow_laptop> yes most likely
Jul 25 13:31:09 <jow_laptop> its authors spend years of work and hundrets of thousand lines of code to make it *not* work in every possible situation
Jul 25 13:31:23 <blaise> that's awesome.. D:
Jul 25 13:31:34 <jow_laptop> thats free software... fuck yeah
Jul 25 13:32:30 <jow_laptop> anyway, try it, if it works you should see "OpenWrt-libtool" in the log
Jul 25 13:32:34 <blaise> I'm currently operating with libtool-2.2.10
Jul 25 13:32:46 <jow_laptop> doesn't matter, all versions are broken
Jul 25 13:32:51 <jow_laptop> just differently
Jul 25 13:32:58 <blaise> shizzle!
Jul 25 13:33:25 <blaise> I could jump up to 2.4-r1, maybe it's not broken as bad?
Jul 25 13:33:43 <jow_laptop> ^Willie^: I think you have to add ldconfig and/or populate ld.conf.so or so
Jul 25 13:33:58 <jow_laptop> blaise: doesn't matter, you do not want the host libtool at all
Jul 25 13:34:21 <blaise> ah..
Jul 25 13:34:29 <jow_laptop> in part because cross compiled packages must be independant from the host system
Jul 25 13:34:40 <jow_laptop> you can't force users to upgrade libtool just because libotr does not build
Jul 25 13:34:46 <blaise> so the crossdev libtool is a different version entirely...
Jul 25 13:34:53 <^Willie^> jow_laptop: hmm i try to answer a question to someone else right now but am i forced to boot a wrt image before i can cat the content of ld.so.conf :S
Jul 25 13:34:54 <jow_laptop> yes
Jul 25 13:35:13 <^Willie^> nice
Jul 25 13:35:16 <blaise> I see.., cuz libotr builds just fine on the host.. :D
Jul 25 13:37:37 <nbd> blaise: have you tracked down yet where the -I/usr/include comes from?
Jul 25 13:37:44 <jow_laptop> nbd: libtool
Jul 25 13:37:48 <nbd> ah
Jul 25 13:37:52 <nbd> that stupid thing
Jul 25 13:38:30 <jow_laptop> libtool... I'm in your toolchain mangling ur cflags
Jul 25 13:38:48 <^Willie^> jow_laptop: where do i have to add ldconfig ?
Jul 25 13:39:04 <blaise> lol
Jul 25 13:39:08 <jow_laptop> ^Willie^: dunno, should be buried in the glibc manual somewhere
Jul 25 13:39:15 <^Willie^> jow_laptop: since the ld export to /usr/lib is gone when use anything diffrent then uclibc in default mode.
Jul 25 13:39:20 <jow_laptop> ^Willie^: you can also check your host system
Jul 25 13:39:37 <jow_laptop> ^Willie^: well uclibc handles that copletely different
Jul 25 13:39:40 <blaise> new libotr build log...... http://pastebin.com/M84WTsiK
Jul 25 13:40:06 <^Willie^> it is fixed in uclibc builds and happend when use glibc or eglibc thats why i ask you people hald luci libs must be copyed to /lib etc etc ..
Jul 25 13:40:13 <blaise> how do I find out what version of libtool is being used in the crossdev toolchain?
Jul 25 13:40:41 <nbd> look at package/libtool/Makefile
Jul 25 13:41:12 <blaise> libtool-2.4-1
Jul 25 13:41:17 * blaise facepalms
Jul 25 13:41:54 <blaise> I wonder if maybe I should build libtool-2.2.10
Jul 25 13:42:01 <blaise> since it worked fine on the host system....
Jul 25 13:42:36 <blaise> or, hell.. maybe even fall back to libtool-1.3.5
Jul 25 13:44:02 <loswillios> no, you don't want that
Jul 25 13:44:29 <blaise> nbd: I can totally imagine jow_laptop going ballistic on an inatamate object right now..
Jul 25 13:45:02 <jow_laptop> blaise: please paste the generated .../toolkit/Makefile as well
Jul 25 13:45:14 <CIA-89> nbd * r27772 /trunk/toolchain/uClibc/ (9 files in 3 dirs): uClibc: update to latest 0.9.32.1 git
Jul 25 13:45:15 <jow_laptop> one directory deeper than the other one
Jul 25 13:45:31 <blaise> uno momento
Jul 25 13:46:16 <loswillios> thanks nbd
Jul 25 13:46:37 <blaise> uh,
Jul 25 13:47:59 <blaise> that's a really big Makefile
Jul 25 13:48:09 <jow_laptop> I only need the first hundret lines or so
Jul 25 13:50:16 <blaise> jow_laptop: http://pastebin.com/JkgwdRZY
Jul 25 13:51:26 <^Willie^> jow_laptop: anyway did you look at the ancient rrdtool on x86 ?
Jul 25 13:51:43 <^Willie^> jow_laptop: the gcc 4.5 and rrdtool issue resulting in not updating rrd's ?
Jul 25 13:52:03 <jow_laptop> ^Willie^: sorry, no not yet - I played with the rt-n16
Jul 25 13:52:09 <^Willie^> hmm
Jul 25 13:52:24 <^Willie^> i do not feel happy about the idea using a second machine for rrdtool :S
Jul 25 13:52:42 <^Willie^> 2 setups one with graph stuff and one as data collector
Jul 25 13:53:25 <^Willie^> will try if native building is going to work if the build essentials even install .. else i dunno
Jul 25 13:53:54 <^Willie^> last try uclibc .. intl/iconv stuff is more linkly glibc targets.
Jul 25 13:54:09 <blaise> I use rrdtool-1.3.8, havn't tried the newer versions though
Jul 25 13:54:11 <^Willie^> and not even shure if the stubs will do.
Jul 25 13:54:18 <^Willie^> blaise: with cairo or without ?
Jul 25 13:54:29 <blaise> with..
Jul 25 13:54:33 <^Willie^> blaise: as long i build without graph creation(cairo) it works
Jul 25 13:54:35 <^Willie^> wtf how ?
Jul 25 13:54:46 <^Willie^> cairopango fails on iconv in glib2
Jul 25 13:54:59 <blaise> I dunno, it just compiled..
Jul 25 13:55:06 <^Willie^> do you have a patch or working makefile posible that can point my at my issue.
Jul 25 13:55:24 <^Willie^> compiled on the target machine like i'm going todo now ? (installing build essentials)
Jul 25 13:56:17 <^Willie^> blaise: compiled but how.. cross compiled using the feeds or on the wrt target setup it self.
Jul 25 13:56:32 <blaise> unrelated...
Jul 25 13:56:38 <^Willie^> no that is related
Jul 25 13:56:41 <blaise> it's on the host system..
Jul 25 13:57:08 <^Willie^> by modify simply the md5 and version and thats it in the makefile in feeds/packages/utils/rrdtool ?
Jul 25 13:57:11 <blaise> not built from crossdev toolchain
Jul 25 13:57:12 <^Willie^> i do not belive that
Jul 25 13:57:22 <^Willie^> hosts building system..
Jul 25 13:57:28 <^Willie^> target machine runs wrt.
Jul 25 13:57:29 <^Willie^> for the infpo
Jul 25 13:58:17 <blaise> yeah, and the machine which I am developing on is a gentoo system..
Jul 25 13:58:46 <^Willie^> blaise: i use debian right now but on gentoo no diff in the results.
Jul 25 13:59:14 <^Willie^> one point i don't get in your statement.. "build on the host" like ssh to the wrt setup and there building it right ?
Jul 25 13:59:56 <blaise> no, when I say host, I mean where I'm typing from.. the box which hosts me..
Jul 25 13:59:56 <^Willie^> blaise: else pm ?
Jul 25 14:00:00 <blaise> sure//
Jul 25 14:00:08 <^Willie^> since i'm curious..
Jul 25 14:08:37 <jow_laptop> blaise: I'll try to build it myself now, can't figure it out just by looking at it
Jul 25 14:09:02 <jow_laptop> blaise: the CFLAGS crap is puked out by automake, but didn't find exactly where
Jul 25 14:09:04 <blaise> jow_laptop: okay, thank you for all your patience and help..
Jul 25 14:10:11 <jow_laptop> building libgcrypt takes a few minutes, I'll get back to you
Jul 25 14:11:38 <blaise> no problem
Jul 25 14:12:06 <blaise> kinda wish google wave hadn't gone out of style, it would make this process a lot easyer
Jul 25 14:12:09 <blaise> lol
Jul 25 14:39:56 <^Willie^> hmm i have seen this before..
Jul 25 14:40:03 <^Willie^> Download failed. --2011-07-25 16:32:34--  http://mirrors.rcn.net/pub/sourceware/gcc/releases/gcc-linaro/gcc-linaro.tar.bz2
Jul 25 14:40:21 <^Willie^> enable build essentials and it is missing the version number there.
Jul 25 14:40:30 <^Willie^> this is not the toolchain gcc
Jul 25 14:44:54 <EqUaTe> blaise: google wave was basically a forum with email-style access.
Jul 25 14:45:06 <^Willie^> hmm wats wrong with this -> PKG_VERSION:=$(strip $(subst ",, $(CONFIG_GCC_VERSION)))#"))
Jul 25 14:45:13 <^Willie^> outcome is nothing
Jul 25 14:46:00 <blaise> EqUaTe: a bunch of people actually wrote syntax highlighting so a group of people could work on a peice of code or document simultaniously
Jul 25 14:46:10 <EqUaTe> heh
Jul 25 14:46:11 <blaise> all that's gone now though :(
Jul 25 14:46:56 <blaise> we had C, ASM, perl, python, and several other plugin's..
Jul 25 14:47:28 <EqUaTe> cool, but possibly overkill
Jul 25 14:48:11 <blaise> I'm kinda wondering maybe if g+ can be used the same way..
Jul 25 14:48:40 <blaise> I better get some sleep..
Jul 25 14:49:53 <blaise> jow_laptop: I'll be available some time after 1200GMT
Jul 25 14:54:03 <^Willie^> jow_laptop: the md5sum for gcc-linaro 4.5 is the md5 from make-3.81.tar.bz2
Jul 25 14:54:11 <^Willie^> that is not a typo i think
Jul 25 14:54:45 <^Willie^> when trying to install build essentials in the openwrt target image.
Jul 25 14:58:12 <^Willie^> works for now by using cp dl/gcc-linaro-4.5-2011.06-0.tar.bz2 dl/gcc-linaro.tar.bz2
Jul 25 14:58:16 <^Willie^> and then patches fail.
Jul 25 15:04:53 <KanjiMonster> ^Willie^: the problem is that CONFIG_GCC_VERSION is "linaro" for current trunk, so it obviously doesn't include any version number
Jul 25 15:05:30 <^Willie^> fixed that, bumped into next it is applying the toolchain patches to the target gcc :(
Jul 25 15:07:32 <^Willie^> hmm fixing the ld patch in menuconfig somehow looks like easyer then this :(
Jul 25 15:07:39 <^Willie^> match=path
Jul 25 15:08:03 <^Willie^> gcc is not maintained for a long time and no gcc 4.4 :(
Jul 25 15:09:43 <^Willie^> KanjiMonster: any idea how to get working build essentials in the target image ?
Jul 25 15:10:09 * ^Willie^ keeps working around problems this way.
Jul 25 15:10:48 <KanjiMonster> no
Jul 25 15:11:09 <KanjiMonster> never needed it
Jul 25 15:11:45 <^Willie^> when i switch from uclibc to glibc or eglibc there is no ldpath defined in /etc/profile and must copy all content from /usr/src to make luci work.
Jul 25 15:11:55 <^Willie^> iirc some switch is missing there.
Jul 25 15:12:31 <^Willie^> i have no clue wich option out of many has been disabled.
Jul 25 15:15:43 <^Willie^> or locate/gettext support enabled in uclibc is also fine.
Jul 25 15:15:48 <^Willie^> locale*
Jul 25 15:15:54 <^Willie^> ok bbl i can't even type right now
Jul 25 15:16:07 <^Willie^> sucks when all features that are needed are disabled.
Jul 25 16:02:07 <builder_> build #48 of ixp4xx is complete: Success [build successful]  Build details are at http://buildbot.openwrt.org:8010/builders/ixp4xx/builds/48
Jul 25 16:18:20 <philipp64|laptop> KanjiMonster: if I'm using GPIO line 3 in a driver file, does that have a corresponding mask of 0x08 or 0x04 on the command line? in other words, is the GPIO numbering 1-based or 0-based in the drivers?
Jul 25 16:24:33 <philipp64|laptop> solca: isn't that usually done by the platform owners?
Jul 25 16:38:53 <builder_> build #46 of kirkwood is complete: Success [build successful]  Build details are at http://buildbot.openwrt.org:8010/builders/kirkwood/builds/46
Jul 25 16:59:51 <jow_laptop> blaise: found it, was a badly written automake tempalte
Jul 25 17:00:31 <^Willie^> fixed the rrdtool-1.4.5 configure errors here. bbl diner.
Jul 25 17:00:49 <^Willie^> +perl +python +cairo +libxml2
Jul 25 17:00:50 <jow_laptop> blaise: rm -rf package/libotr/; wget -qO- http://luci.subsignal.org/~jow/libotr.patch | patch -p0
Jul 25 17:11:59 <swalker> blaise: unneeded & misspelled libgpg-error depends in the Makefile
Jul 25 17:13:09 <jow_laptop> I updated the patch
Jul 25 17:13:23 <swalker> nbd: http://git.uclibc.org/uClibc/patch/libc/sysdeps/linux/common/epoll.c?id=8245f3b4638fdff2011c2657af1bb211def704bc is needed on x86 after the uClibc change
Jul 25 17:44:02 <philipp64|laptop> anyone seen cshore lately? I've not seen him in over a week...
Jul 25 17:46:59 <_trine> has any platform moved over the kernel 3 yet
Jul 25 17:51:09 <mazilo> The STAGING_DIR_HOST variable points to "staging_dir/host" directory, Can anyon here please tell me what variable points to "staging_dir/target-<platform>/host" directory?
Jul 25 18:01:23 <builder_> build #42 of ep93xx is complete: Failure [failed compile_4]  Build details are at http://buildbot.openwrt.org:8010/builders/ep93xx/builds/42
Jul 25 18:21:57 <jow_laptop> mazilo: STAGING_DIR
Jul 25 18:22:24 <mazilo> jow_laptop: Thanks.
Jul 25 18:28:42 <builder_> build #40 of rdc is complete: Failure [failed compile_3]  Build details are at http://buildbot.openwrt.org:8010/builders/rdc/builds/40
Jul 25 18:31:51 <RealOpty> _trine, brcm47xx
Jul 25 18:32:20 <CIA-89> nbd * r27773 /trunk/toolchain/uClibc/patches-0.9.32/150-x86_epoll_fix.patch: uClibc: backport an x86 compile fix from mainline (thx, swalker)
Jul 25 18:32:34 <RealOpty> nbd, thanks for that patch, just in time
Jul 25 18:54:11 <_trine> RealOpty, ah thanks
Jul 25 18:54:45 <RealOpty> _trine, x86 can do 3.0 if u set it to that kernel also
Jul 25 18:54:46 <_trine> I'm just trying a compile on ar71xx using 3.0
Jul 25 19:44:40 <RealOpty> the following config symbols are missing from the kernel  config 3.0 for x86, possibly others. -----> DEBUG_PERF_USE_VMALLOC, ACPI_CUSTOM_METHOD, ACERHDF, HP_ACCEL, XO15_EBOOK, DEBUG_NX_TEST, CPA_DEBUG, DEBUG_STRICT_USER_COPY_CHECKS,
Jul 25 19:47:01 <RealOpty> Kaloz, maybe you can add these to config-default on x86 target, as this patch was incomplete https://dev.openwrt.org/changeset/27762/trunk
Jul 25 19:47:01 <Hauke> RealOpty: could you make a patch and place the symbols into the x86 specific config and the general config
Jul 25 19:47:49 <RealOpty> Hauke, sure np
Jul 25 19:59:05 <CIA-89> jogo * r27774 /trunk/target/linux/brcm63xx/ (patches-2.6.39/240-spi.patch patches-3.0/240-spi.patch):
Jul 25 19:59:06 <CIA-89> bcm63xx: Fix SPI commands and register offsets for BCM6348
Jul 25 19:59:06 <CIA-89> In 240-spi.patch, spi registers for bcm6348 were all messed up. This patch fixes that.
Jul 25 19:59:06 <CIA-89> It also fixes some spi commands for all bcm63xx.
Jul 25 19:59:06 <CIA-89> Signed-off-by: Anthony Blakemore <stokie-ant@raverbaby.co.uk>
Jul 25 20:10:29 <RealOpty> Hauke, can i bump x86 up to 3.0 while im at it?
Jul 25 20:14:20 <RealOpty> Hauke, here is the patch https://dev.openwrt.org/attachment/ticket/9719/add_3.0_config_symbols.patch
Jul 25 20:16:53 <RealOpty> Hauke, apply this patch too https://dev.openwrt.org/attachment/ticket/9718/add_virtualbox_profile.2.patch :)
Jul 25 21:27:20 <CIA-89> hauke * r27775 /trunk/target/linux/ (generic/config-3.0 x86/config-default):
Jul 25 21:27:20 <CIA-89> x86: add missing config symbols
Jul 25 21:27:20 <CIA-89> This adds some config symbols for kernel 3.0.
Jul 25 21:27:20 <CIA-89> Thank you realopty for the patch.
Jul 25 21:27:20 <CIA-89> This is from #9719.
Jul 25 21:30:10 <Hauke> RealOpty: the virtual box ticket is owned by [florian]
Jul 25 21:31:02 <Hauke> RealOpty: I do not want to switch x86 to linux 3.0 now as default as it is not that stable
Jul 25 21:36:45 <RealOpty> but i thought backfire was for stability and trunk was for testing?
Jul 25 21:37:54 <jow_laptop> yeah but testing does not mean sabotage :P
Jul 25 21:38:15 <jow_laptop> you can still flip the number in target/linux/x86/Makefile
Jul 25 21:38:29 <jow_laptop> patches etc. are there regardless of whats the default kernel atm
Jul 25 21:38:53 <RealOpty> yeah
Jul 25 21:39:57 <RealOpty> jow_laptop, how about copying the x86 config-default to config-3.0 :)
Jul 25 22:15:20 <tripolar> hello - i get this  switch_adm: Unknown symbol ssb_bcm47xx (err 0)
Jul 25 22:15:31 <tripolar> i created a patch - but i dont know if its correct ...
Jul 25 22:15:36 <tripolar> http://nopaste.dk/p3645
Jul 25 22:15:43 <tripolar> can someone tell me if it looks okay?
Jul 25 22:17:26 <tripolar> ping nbd
Jul 25 22:26:13 <CIA-89> jow * r27776 /trunk/package/base-files/ (Makefile files/lib/upgrade/common.sh): [package] base-files: use "reboot -f" after sysupgrade to prevent executing deleted init scripts
Jul 25 22:27:01 <CIA-89> jow * r27777 /branches/backfire/package/base-files/ (Makefile files/lib/upgrade/common.sh): [backfire] backport r27776
Jul 25 22:51:50 <builder_> build #44 of iop32x is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/44
Jul 25 22:51:50 <builder_> build #43 of etrax is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/etrax/builds/43
Jul 25 22:51:52 <builder_> build #40 of gemini is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/gemini/builds/40
Jul 25 22:51:54 <builder_> build #40 of adm5120 is complete: Exception [exception interrupted]  Build details are at http://buildbot.openwrt.org:8010/builders/adm5120/builds/40
Jul 25 23:20:13 <builder_> build #60 of ubicom32 is complete: Failure [failed compile_3]  Build details are at http://buildbot.openwrt.org:8010/builders/ubicom32/builds/60
Jul 25 23:30:24 <ixew> Hi, I am trying to build trunk with broadcom-wl (since b43 doesn't work well in brcm47xx and I can't use a 2.4 kernel). However, compiling package/broadcom-wl fails because of undefined BCMA_HOSTTYPE_SOC and more... help please.
Jul 25 23:35:35 <jow_laptop> it currently only works with kernel 3.0
Jul 25 23:36:11 <jow_laptop> alternatively revert r27751
Jul 25 23:36:30 <jow_laptop> and r27752
Jul 25 23:37:47 <ixew> kernel 3.0 is not available for brcm47xx yet, right?
Jul 25 23:39:28 <jow_laptop> it is
Jul 25 23:39:57 <jow_laptop> just change LINUX_VERSION in target/linux/brcm47xx/Makefile to 3.0
Jul 25 23:40:16 <ixew> thanks, will try
Jul 25 23:41:14 <ixew> I guess it should be a menuconfig option
Jul 25 23:42:55 <ixew> it is now: LINUX_VERSION:=2.6.39.2 and should be 2.6.39.3 ?
Jul 25 23:45:29 <ixew> didn't know 3.0 is out :-)
Jul 25 23:59:03 <_trine> linux 3 does not build on AR71xx
Jul 26 00:02:16 <builder_> build #62 of at91 is complete: Failure [failed compile_4]  Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/62
Jul 26 01:57:16 <philipp64|laptop> jow_laptop: ping
**** ENDING LOGGING AT Tue Jul 26 02:59:56 2011