**** BEGIN LOGGING AT Fri May 21 02:59:57 2010
May 21 05:39:06 <russiane39> hmm, why ramips target has no bootinfo.h and #define CL_SIZE                        COMMAND_LINE_SIZE ?
May 21 08:07:30 <russiane39> Unfortunately, I give up with 2.6.34 for ralink :-( https://forum.openwrt.org/viewtopic.php?id=24876
May 21 09:53:48 <CIA-38> florian * r21518 /trunk/package/iproute2/Makefile: [package] iproute2: update to v2.6.31 (#7354)
May 21 09:53:53 <CIA-38> florian * r21519 /trunk/package/iproute2/Makefile: [package] iproute2: cleanup Makefile after r21513
May 21 09:53:57 <CIA-38> florian * r21520 /trunk/package/kernel/modules/netdevices.mk: [package] add kmod-swconfig and kmod-swich-ip175c
May 21 09:54:01 <CIA-38> florian * r21521 /trunk/target/linux/rdc/ (3 files in 3 dirs): [rdc] add r6040 phylib support which allows to use the ip175c switch driver on ar525w
May 21 09:54:05 <CIA-38> florian * r21522 /trunk/package/iptables/Makefile: [package] update iptables to 1.4.8
May 21 09:54:23 <[florian]> oh crap, git svn rebase did confuse me again
May 21 10:19:59 <xl0> Can anybosy try building the pciutils package? Looks like it's triggering a toolchain problem for me.
May 21 11:09:49 <[florian]> xl0: works fine here (mips32 be)
May 21 11:24:33 <xl0> [florian]: http://dumpz.org/19595/ I'm getting this. Any ideas, what is a "hidden" symbol?
May 21 11:25:46 <[florian]> xl0: seems like you should also be linking with -lgcc_s
May 21 11:27:25 <xl0> Btw, what's the difference between the two?
May 21 11:27:46 <[florian]> between what and what?
May 21 11:28:22 <xl0> libgcc and libgcc_s.
May 21 11:29:14 <xl0> The second one seems to only come in .so, so does _s stand for shared?
May 21 11:29:16 <nbd> you should include $(LIBGCC_S) in the linker flags
May 21 11:29:36 <nbd> some platforms need to link against libgcc_s.so, some against libgcc.a
May 21 11:36:27 <xl0> That seemed to work. But why is $(LIBGCC_S) not part of the default target ld flags?
May 21 11:39:54 <xl0> And, should not gcc itself decide, whether to link to libgcc, or libgcc_s?
May 21 15:08:29 <CIA-38> nico * r21523 /trunk/package/dnsmasq/Makefile: package/dnsmasq: bump release number (missed from r21400)
May 21 15:10:34 <CIA-38> nico * r21524 /trunk/package/dropbear/Makefile: package/dropbear: bump release number (missed in r20960)
May 21 15:15:13 <rtz2> nbd: ping
May 21 15:23:56 <CIA-38> nico * r21525 /trunk/package/ebtables/Makefile: package/ebtables: update to 2.0.9-2
May 21 15:23:56 <CIA-38> nico * r21526 /trunk/package/arptables/Makefile: package/arptables: update to 0.0.3-4
May 21 15:32:31 <CIA-38> nico * r21527 /trunk/package/broadcom-diag/Makefile: package/broadcom-diag: bump release number (missed from r20737 & r20869)
May 21 15:37:02 <CIA-38> nico * r21528 /trunk/package/button-hotplug/Makefile: package/button-hotplug: bump release number (missed from r21074)
May 21 16:55:30 <builder> build #42 of uml is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/uml/builds/42
May 21 17:39:25 <countcobolt> hi qll
May 21 17:39:29 <countcobolt> *all
May 21 17:40:04 <countcobolt> I am trying to add fsvs to the package list of openwrt Kamikaze 8.9.0.2
May 21 17:40:32 <countcobolt> the build is constantly failing stating it does not find apr and apr_util
May 21 17:40:57 <countcobolt> while they are being compile before I start the compilation of fsvs
May 21 17:41:29 <rtz2> countcobolt: well, you know
May 21 17:41:40 <rtz2> countcobolt: an actual error message would kinda helpful
May 21 17:41:51 <countcobolt> okay , I was just goign there :D
May 21 17:42:01 <countcobolt> I have posted everything on the forum
May 21 17:42:08 <countcobolt> https://forum.openwrt.org/viewtopic.php?id=24870
May 21 17:43:16 <jow_laptop> countcobolt: check /home/countcobolt/8.09.2/build_dir/i386/fsvs-1.2.2/config.log
May 21 17:43:45 <countcobolt> jow_laptop, I check this and this is what I find
May 21 17:43:54 <jow_laptop> countcobolt: it must not be a missing file, could be some compile error as well due to other stuff (uses host compiler, uses unsupported flags, ...)
May 21 17:44:01 <rtz2> countcobolt: you sure you have apr?
May 21 17:44:06 <countcobolt> yes
May 21 17:44:16 <countcobolt> I have it on the host system
May 21 17:44:17 <rtz2> countcobolt: well, check the log then
May 21 17:44:21 <countcobolt> and in buildroot
May 21 17:44:24 <jow_laptop> host system is wrong
May 21 17:44:27 <countcobolt> I can even find them
May 21 17:44:27 <jow_laptop> ok
May 21 17:44:31 <countcobolt> in the staging dir
May 21 17:44:39 <jow_laptop> put config.log into pastebin
May 21 17:44:44 <countcobolt> okay
May 21 17:47:14 <countcobolt> http://pastebin.org/262167
May 21 17:47:16 <countcobolt> here you go
May 21 17:49:07 <countcobolt> even the libsqlite is available in the build_dir /usr/lib/
May 21 17:50:46 <countcobolt> jow_laptop,  this http://pastebin.org/262171 is an ls of my staging_dir
May 21 17:52:33 <countcobolt> @rtz I needed apr to have subversion
May 21 18:20:05 <jow_laptop> countcobolt: see line 194
May 21 18:20:13 <jow_laptop> countcobolt: it says exactly what to do
May 21 18:21:17 <countcobolt> jow_laptop, check the other pastebin
May 21 18:21:25 <countcobolt> libexpat is available on staging_dir
May 21 18:21:28 <jow_laptop> countcobolt: add -Wl,-rpath-link=$(STAGING_DIR)/usr/lib to TARGET_CFLAGS and/or -rpath-link=$(STAGING_DIR)/usr/lib to TARGET_LDFLAGS
May 21 18:21:41 <jow_laptop> the linker does not look there by default
May 21 18:21:55 <jow_laptop> it does not matter whether the libs are there or not as long as the linker does not use this dir
May 21 18:30:51 <countcobolt> I will give this a try asap, unexpected visitors here
May 21 19:29:03 <rtz2> jow_laptop: you still here?
May 21 20:35:42 <CIA-38> jow * r21529 /trunk/package/uci/Makefile: [package] uci: bump pkg release after r21489-r21494 (#7312)
May 21 20:36:11 <xMff> rtz2: now yes
May 21 20:38:14 <rtz2> xMff: do you remeber the problem with mmaping mtd devices?
May 21 20:38:25 <xMff> vaguely
May 21 20:38:45 <rtz2> xMff: do you remeber which mmap flag fixed it?
May 21 20:39:25 <xMff> trac does... https://dev.openwrt.org/changeset/19486/trunk/package/nvram
May 21 20:42:01 <rtz2> xMff: thanks
May 21 20:42:09 <rtz2> xMff: mtd seems to have a similiar problem
May 21 20:42:11 <rtz2> somehow
May 21 20:42:13 <rtz2> sometimes
May 21 20:42:55 <xMff> ah
May 21 20:43:16 <xMff> that could explain the sysupgrade failures on ar71xx
May 21 20:43:30 <xMff> if I recall correctly the nvram issue was also only seen on ar71xx
May 21 20:45:13 <rtz2> hmmmm
May 21 20:45:26 <rtz2> let me take a look at the flash driver
May 21 21:02:01 <rtz2> ok, MAP_LOCKED and mtd is not a good combination
May 21 21:25:40 <CIA-38> florian * r21530 /trunk/package/kernel/modules/netdevices.mk: [package] fix kmod-swconfig module loading typo
May 21 21:25:48 <CIA-38> florian * r21531 /trunk/target/linux/rdc/base-files/lib/preinit/05_set_ether_mac_rdc: [rdc] load all r6040 modules dependencies in preinit hook
May 21 21:25:53 <CIA-38> florian * r21532 /trunk/target/linux/rdc/profiles/ar525w.mk: [rdc] include swconfig in ar525w profile
May 21 21:29:18 <[florian]> rtz2: did you have some time to test r6040?
May 21 21:41:46 <rtz2> [florian]: only a few short tests, but it seemes to work
May 21 21:41:54 <rtz2> [florian]: I didn't have any problems
May 21 21:44:46 <rtz2> [florian]: the phys are there (duplicated)
May 21 21:50:43 <[florian]> rtz2: yeah, I do not really know to prevent that however
May 21 21:51:00 <[florian]> rtz2: it does not seem to be an issue anyway, since it's same the switch in my case for instance
May 21 21:53:04 <rtz2> [florian]: you have to create only one mdio_bus for both eth devices, but it's probably not worth the effort
May 21 21:53:49 <[florian]> rtz2: yes that's right
May 22 02:01:22 <CIA-38> jow * r21533 /trunk/package/firewall/files/lib/core_rule.sh: [package] firewall: count rules per chain and family, fix wrong order of ip6tables rules when ipv4 only or dual family rules are defined
**** ENDING LOGGING AT Sat May 22 02:59:57 2010