**** BEGIN LOGGING AT Fri Dec 17 02:59:59 2010 Dec 17 03:02:45 damn it. snow storm coming in. ok, back on from home... assuming I make it. Dec 17 03:33:16 xMff: you were saying? Dec 17 04:07:17 flyn * r24632 /packages/net/krb5/Makefile: Remove non-essential software from krb5-server package to reduce size Dec 17 04:09:20 swalker * r24633 /packages/net/umurmur/ (Makefile patches/ patches/001-fix-libconfig-include.patch): [packages] umurmur: update to 0.2.5 (#8268), thanks martin, and patch the libconfig header path post r24199 Dec 17 04:25:05 swalker * r24634 /packages/utils/upx/Makefile: [packages] upx: update to 3.07 Dec 17 04:34:34 swalker * r24635 /packages/libs/libsigc++/ (Makefile patches/001-no_doc_tests_examples.patch): [packages] libsigc++: update to 2.2.8 (#8276), thanks sav, use PKG_INSTALL and refresh patches Dec 17 05:04:36 swalker * r24636 /packages/libs/xmlrpc-c/ (Makefile patches/001-config.mk.in.patch): [packages] xmlrpc-c: update to 1.24.4 (#8277), thanks sav, use PKG_INSTALL and refresh patches Dec 17 05:15:43 swalker * r24637 /packages/libs/libexif/Makefile: [packages] libexif: install the pkg-config file (#8468), thanks mike Dec 17 09:49:39 build #44 of ar71xx is complete: Failure [failed compile_11] Build details are at http://tksite.gotdns.org:8010/builders/ar71xx/builds/44 Dec 17 10:28:17 ^ [mkzynfw] *** error: file openwrt-ar71xx-generic-nbg460n_550n_550nh-jffs2-sysupgrade.bin is too big, offset = 458752, size=3538948, align = 65536 Dec 17 10:33:56 swalker * r24638 /packages/utils/usbutils/Makefile: [packages] usbutils: update to 001 Dec 17 12:49:13 build #40 of ep93xx is complete: Failure [failed compile_7] Build details are at http://tksite.gotdns.org:8010/builders/ep93xx/builds/40 Dec 17 13:20:20 build #49 of ubicom32 is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/ubicom32/builds/49 Dec 17 13:24:53 dingo * r24639 /packages/utils/triggerhappy/Makefile: [patch-team] update triggerhappy package to version 0.3.1-1 - Signed-off-by: Stefan Tomanek Dec 17 13:33:02 dingo * r24640 /packages/ (10 files in 5 dirs): [patch-team] new packages - restund, baresip, re Open-Source VoIP project - Signed-Off-By: Alfred E. Heggestad Dec 17 13:42:26 nbd: got a cheap wn2000rpt (an ar71xx with a ar8236 switch chip), so expect a patch for it "soon" (probably combined with some generic atheros switch driver rework - how about renaming it "ath8k" for total confusion? ;) Dec 17 13:46:56 sounds good, will you split out the ar8216 nastiness as well? Dec 17 13:49:21 yes Dec 17 13:49:40 ok, thanks Dec 17 13:51:34 as far as I can see, there are now three generations of switch chips in use, gen1 (ar8216, old header, old port register layout, 2k mac table, no acl), gen2 (ar8316, new header, old port register layout, 2k mac, acl) and current gen(ar8236, ar8227/8, possibly ar8327/8, new header, new port register layout, 1k mac, acl) Dec 17 13:52:07 yeah, i think atheros calls them s16, s26 and s27 Dec 17 13:52:14 split up in the same way Dec 17 13:52:36 though i'm not sure which one's s16 and which one is s26, i always get those two mixed up Dec 17 13:52:54 S26 is the ar8216 Dec 17 13:55:34 the internal one of the ar7240 is also called S26, but it seems to be something between the S26 and S16 (e.g. the mib counters are at the same place as the S16, but featurewise its the same as the S26; I'm not sure about the ath header, I don't have one) Dec 17 15:34:46 nbd * r24641 /branches/backfire/package/mac80211/patches/ (2 files): ath9k: merge fixes from r24627, r24628 Dec 17 15:40:09 Has anyone tried to compile libs/gd lately? It fails to compile due to libiconv even though the Makefile on libs/gd specifically declares '--without-iconv' in its ./configure line. Dec 17 15:40:54 Here is the message "gdkanji.c:30: error: conflicting types for 'libiconv'" Dec 17 16:07:44 works for me Dec 17 16:22:15 jow_laptop: Did you do a clean compilation? Dec 17 16:29:51 build #40 of ifxmips is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/ifxmips/builds/40 Dec 17 16:41:13 Hi to all. I have a problem related to USB block devices: Dec 17 16:43:12 When owrt starts recognize usb device (a pen with multiple partitions) but doesn't mount them. The only way I found is to set (testint) "/usr/bin/env -i ACTION=add DEVPATH=/dev/sda1 /sbin/hotplug-call block" in boot script but I think I need something a loop to check for all the partitions on the (unique) usb device. Dec 17 17:11:17 juhosg * r24642 /trunk/target/linux/ar71xx/base-files/etc/uci-defaults/wzr-hp-g300nh: Dec 17 17:11:17 ar71xx: populate LED configuration for WZR-HP-G300NH Dec 17 17:11:17 Signed-off-by: Dave Lichterman Dec 17 17:11:17 [nand-disk trigger has been removed - juhosg] Dec 17 17:11:21 juhosg * r24643 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/mach-ubnt.c: (log message trimmed) Dec 17 17:11:21 ar71xx: UBNT Secondary MAC address duplicate fix Dec 17 17:11:21 There is Secondary MAC address duplicate problem with some UBNT Dec 17 17:11:21 RouterStation and RouterStation Pro HW batches as Primary MAC addresses Dec 17 17:11:22 are not increased by 2 per device in board data. Fix is to use 'Locally Dec 17 17:11:22 Administrated bit' for Secondary MAC address instead of increasing Dec 17 17:11:23 Primary MAC addresses by 1 which could overlap with other device Dec 17 17:11:23 juhosg * r24644 /trunk/ (11 files in 11 dirs): generic: nuke 2.6.33 specific stuff, is not used by any platform Dec 17 17:11:27 juhosg * r24645 /trunk/target/linux/generic/ (5 files in 5 dirs): generic: add usb_find_device_by_name helper Dec 17 17:11:30 juhosg * r24646 /trunk/ (10 files in 8 dirs): generic: add LED trigger for USB device presence/activity Dec 17 17:11:32 juhosg * r24647 /trunk/package/base-files/ (Makefile files/etc/init.d/led): base-files: add support for the usbdev LED trigger Dec 17 17:11:35 juhosg * r24648 /trunk/target/linux/ar71xx/ (7 files in 2 dirs): ar71xx: use usbdev trigger for USB LEDs by default Dec 17 17:11:38 juhosg * r24649 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/ (dev-ap94-pci.c dev-ap94-pci.h mach-wndr3700.c): ar71xx: add helper function to setup ath9k LED/GPIO on the AP94 based boards Dec 17 17:11:40 juhosg * r24650 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/mach-dir-825-b1.c: ar71xx: fix WLAN LEDs on the DIR-825 board Dec 17 17:54:34 then after boot, I can't add or remove any unit. ... Some ideas? Consider that I installed correctly the drivers, ntfs-3g helper and busybox corresponding helper but the mechanism won't work ;( Dec 17 17:58:43 juhosg * r24651 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/dev-ap94-pci.h: ar71xx: remove stray semicolons Dec 17 18:16:23 jow_laptop: I deleted/retrieved a current libs/gd/Makefile and did a recompilation with the same error messages. Here is the http://pastebin.com/UFrjjTh7 that shows the error in compiling libs/gd on my OpenSuSE v11.3. Dec 17 18:18:17 lol: "Decreasing size of jffs overlay" Dec 17 18:21:25 someone did not understood the fs overlay Dec 17 18:23:39 apparently Dec 17 18:23:58 I think there's a new entry on the wiki which explains things Dec 17 18:36:15 jow_laptop: Can you confirm libiconv is not selected and built on your OpenWRT system? Dec 17 18:37:30 jow_laptop: On my OpenWRT, libiconv is selected by other package and building lib/gd show a conflict with libiconv as shown in the http://pastebin.com/UFrjjTh7 link. Dec 17 18:39:15 loswillios: googling for an existing explanation, i found (first hit) a circa 2005 explanation from mbm on a sticky forum page with some obsolete information about symlinks (pre-minifo, apparently) Dec 17 18:40:56 that might be it Dec 17 18:42:29 https://forum.openwrt.org/viewtopic.php?id=3474 Dec 17 20:22:18 <[mbm]> hmm yeah, seem to recall that a long time ago there actually was an overlay based on symlinks Dec 17 20:26:15 cshore: ping? Dec 17 20:35:48 juhosg * r24652 /trunk/target/linux/generic/patches-2.6.32/009-jffs-fix-min-max-confusion.patch: generic: fix min/max confusion in jffs2_sum_init on 2.6.32 Dec 17 20:35:52 juhosg * r24653 /trunk/tools/firmware-utils/src/mkzynfw.c: firmware-utils/mkzynfw: fix available flash space calculation Dec 17 20:37:06 is there a security person here? Dec 17 20:37:16 as in an openwrt security person? Dec 17 20:42:55 dingo * r24654 /packages/ (libs/re/Makefile net/baresip/Makefile net/restund/Makefile): [patch-team] reconvert spaces to tabs - Signed-off-by: Alfred E. Heggestad Dec 17 20:44:08 ioerror: what do you mean by security person? Dec 17 20:44:43 security@ generally Dec 17 20:45:06 Or someone who builds the binary packages - so when there is a bug in a piece of software, openwrt people can build new packages with a security patch or otherwise Dec 17 20:46:09 ioerror: we do not have a special handling for security problems yet Dec 17 20:47:59 ah, ok Dec 17 20:48:05 well, how can i help make that happen? Dec 17 20:48:25 Or better - how do i ensure that a specific package gets upgraded ASAP with new binary builds released? Dec 17 20:51:56 build #55 of brcm63xx is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/brcm63xx/builds/55 Dec 17 20:53:02 ioerror: about what package are your talking in this case? Dec 17 20:55:08 I'd rather not say here at the moment Dec 17 21:12:34 xMff: Remember that a couple of days ago, you helped me to fix the compilation of libs/libshout due to libvorbisidec? Now, I did a 'make dirclean' and a 'make' to find out the compilation of libs/libshout gives out error messages as shown on http://pastebin.com/kHBHFMNY . Dec 17 21:24:55 ioerror: then send a mail to openwrt-hackers on lists.openwrt.org Dec 17 21:26:48 xMff: is there a pgp key or is it more open? Dec 17 21:27:10 qjust send it there Dec 17 21:38:23 ok Dec 17 21:38:49 xMff: so in the future, i may have security sensitive things (bugs, patches, etc) and my goal is not to disclose them until they are fixed in OpenWRT Dec 17 23:08:50 pong ioerror Dec 17 23:29:24 hi cshore - i have another tor patch coming soon Dec 17 23:33:00 ok, cool Dec 17 23:37:45 i'm trying to create a tor-alpha package as well Dec 18 00:08:18 any asm gurus out there? Dec 18 00:10:04 I'm looking at http://pastebin.com/ssDFXRbL Dec 18 00:11:37 and I'm thinking, what's the difference between ""jr\t%0"::"r"(0xbfc00000) and "jr\t%0": :"r"(0xbfc00000) ? Dec 18 00:38:06 build #41 of adm5120 is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/adm5120/builds/41 Dec 18 01:08:05 xMff: R U there? Dec 18 01:17:39 xMff: Here is the error messages in libs/libshout compilation: http://pastebin.com/zg5sQC6y Dec 18 02:01:33 nbd * r24655 /trunk/package/mac80211/ (29 files in 2 dirs): mac80211: update to wireless-testing 2010-12-16 Dec 18 02:03:20 dingo * r24656 /packages/net/tor-alpha/ (6 files in 3 dirs): [patch-team] - tor-alpha new package - Signed-off-by: Jacob Appelbaum Dec 18 02:04:35 build #50 of orion is complete: Failure [failed compile_7] Build details are at http://tksite.gotdns.org:8010/builders/orion/builds/50 Dec 18 02:12:31 more trac spam, #8475, along with #8464 & #8387 Dec 18 02:14:35 ping acoul Dec 18 02:15:14 grrrr, what time is it in .gr? Dec 18 02:16:09 defect #8211 apparently also breaks multiwan now Dec 18 02:16:13 nbd * r24657 /trunk/package/mac80211/patches/531-ath9k_legacy_chainmask_fix.patch: ath9k: do not force the chainmask to 1 for legacy mode - it makes no sense and it reduces txpower and unnecessarily disables MRC Dec 18 02:17:00 having a less-than-functional iptables binary on a router is just silly Dec 18 02:22:43 nbd * r24658 /branches/backfire/package/mac80211/patches/ (3 files): mac80211: merge pending regulatory fixes Dec 18 02:22:45 nbd * r24659 /branches/backfire/package/mac80211/patches/ (2 files): ath9k: merge chainmask fix from r24657 Dec 18 02:24:08 xMff/jow_laptop: you worked on libtool too, so maybe you can also take a look Dec 18 02:24:25 _Ernesto_: you reported this, right? **** ENDING LOGGING AT Sat Dec 18 03:00:01 2010