**** BEGIN LOGGING AT Wed Dec 29 02:59:57 2010 Dec 29 04:05:43 03Mike Westerhof  07org.openembedded.dev * r0799fe9dd1 10openembedded.git/ (4 files in 4 dirs): Dec 29 04:05:43 SlugOS: fix module loading problem and image size problems: Dec 29 04:05:43 - add explicit dependency so that cross depmod will at image creation Dec 29 04:05:43 - switch to busybox 1.18.0 and add defconfig for SlugOS Dec 29 04:05:43 - remove cpio from image (experiment with busybox version) Dec 29 04:05:44 - remove sysvinit-pidof and dependencies from image Dec 29 04:05:45 - clean up syntax and add qemuarm for compatible machines Dec 29 04:12:20 03Mike Westerhof  07org.openembedded.dev * r2279283caf 10openembedded.git/recipes/sysvinit/sysvinit_2.86.bb: Dec 29 04:12:20 SlugOS: sysvinit_2.86.bb - override dependency that pulls in sysvinit-utils Dec 29 04:12:20 (SlugOS, like micro, uses busybox to provide the sysvinit-utils functionality) Dec 29 06:05:33 03Martin Jansa  07master * r45316c9713 10openembedded.git/conf/distro/include/preferred-shr-versions.inc: Dec 29 06:05:34 shr: prefer newer perl 5.10.1 Dec 29 06:05:34 Signed-off-by: Martin Jansa Dec 29 06:33:11 Does Bitbake support more languages? Let's say I wanna include software written with D, Vala or golang? Dec 29 06:34:15 The easy way would be precompile the binaries and use package just to place them in the right directories. Dec 29 06:45:56 davidlt: you can add the compiler into the toolchain Dec 29 07:45:25 gm Dec 29 07:52:08 terrible nick Dec 29 07:52:15 morning eFfeM_work Dec 29 07:52:32 hi ka6sox Dec 29 07:52:36 my nick ? Dec 29 07:52:44 no, Unhelpful Dec 29 07:52:46 :( Dec 29 07:52:53 ah ok Dec 29 07:53:08 my wife sometimes says that to me (when I make fun of something usually) Dec 29 07:53:16 :-) Dec 29 09:09:03 good morning Dec 29 09:09:53 morning all ! Dec 29 11:10:08 03Koen Kooi  07org.openembedded.dev * rce37bc90d4 10openembedded.git/recipes/linux/ (17 files in 6 dirs): Dec 29 11:10:08 linux-omap 2.6.37rc: update SRCREV and add more machines Dec 29 11:10:08 Signed-off-by: Koen Kooi Dec 29 11:44:57 How to turn kernel messages about I/O on angstrom? Dec 29 11:45:13 echo 1 > /proc/sys/vm/block_dump, doesn't to it. Dec 29 11:45:33 dmesg doesn't give any WRITE/READ messages. Dec 29 13:51:21 hi, I've a strange compilation problem: Dec 29 13:51:22 http://pastebin.com/32yE5wji Dec 29 13:51:31 does gcc really work? Dec 29 13:51:36 is it a race? Dec 29 13:53:29 crtbeginS.o should be what is included in any compiled program Dec 29 13:53:40 for the startup of the program Dec 29 13:53:41 no? Dec 29 13:56:59 hmmm Dec 29 13:57:02 /home/gnutoo/embedded/oe/oetmps/htcdream/sysroots/x86_64-linux/usr/armv6-novfp/lib/gcc/arm-oe-linux-gnueabi/4.5.3/crtbegin.o is present Dec 29 13:57:10 strange that it picks the old gcc Dec 29 13:59:58 GNUtoo|laptop, hi Dec 29 14:00:07 pespin|away, I've the same error than you Dec 29 14:00:13 so far I've found that: Dec 29 14:00:26 oh then it's not my disk's fault ;) Dec 29 14:00:36 arm-oe-linux-gnueabi-g++: /home/gnutoo/embedded/oe/oetmps/htcdream/sysroots/x86_64-linux/usr/armv6-novfp/lib/gcc/arm-oe-linux-gnueabi/4.5.2/crtbeginS.o: No such file or directory Dec 29 14:00:38 but.... Dec 29 14:00:43 building libdb Dec 29 14:00:45 /home/gnutoo/embedded/oe/oetmps/htcdream/sysroots/x86_64-linux/usr/armv6-novfp/lib/gcc/arm-oe-linux-gnueabi/4.5.3/crtbegin.o is present Dec 29 14:01:04 and I crtbegin* is ultra important btw Dec 29 14:01:14 it's the startup of any program Dec 29 14:01:15 I think Dec 29 14:01:18 I've to verify tough Dec 29 14:01:30 basically it get inserted at the beginning of each program Dec 29 14:02:28 GNUtoo|laptop, what about linking the 4.5.2 .o file to 4.5.3? :P Dec 29 14:02:46 hmmm Dec 29 14:02:52 note the S too Dec 29 14:02:59 ah Dec 29 14:03:12 btw, why is it building with 4.5.2 if 4.5.3 is available? -.^ Dec 29 14:03:20 crtbeginS.o is present in 4.5.3 dir too Dec 29 14:03:31 hmmm Dec 29 14:03:37 are you sure it's building 4.5.2 Dec 29 14:03:38 ? Dec 29 14:03:42 GNUtoo|laptop, I have both crtbegin.o crtbeginS.o present in 4.5.3 Dec 29 14:03:43 and not rather *using* Dec 29 14:03:47 ok me too Dec 29 14:04:19 | make: *** [libdb-5.0.la] Error 1 Dec 29 14:04:21 that's not gcc Dec 29 14:04:24 that's libdb Dec 29 14:04:29 so it's using 4.5.2 stuff Dec 29 14:04:43 let me look something Dec 29 14:04:54 ok Dec 29 14:05:14 there are only directories in /home/gnutoo/embedded/oe/oetmps/htcdream/sysroots/x86_64-linux/usr/armv6-novfp/lib/gcc/arm-oe-linux-gnueabi/4.5.2/ Dec 29 14:05:24 which I think is caused by some uninstall issues Dec 29 14:05:28 like wrong rm command Dec 29 14:05:33 or something like that Dec 29 14:06:42 maybe move that dir, clean the libdb and restart compiling Dec 29 14:07:01 to see if it uses 4.5.3 then? Dec 29 14:07:12 no to workarround the issue Dec 29 14:08:02 if you have time you could signal it in the openembedded mailing list Dec 29 14:08:17 GNUtoo|laptop, I'm not in the ml Dec 29 14:08:23 ok Dec 29 14:08:34 then signal it to JaMa|Ski Dec 29 14:09:01 ok I'll try to remember Dec 29 14:32:25 03Koen Kooi  07org.openembedded.dev * rdbd3c51c7b 10openembedded.git/recipes/linux/ (31 files in 3 dirs): Dec 29 14:32:25 linux-omap 2.6.37rc: add more sakoman patches and fix COMPATIBLE_MACHINE Dec 29 14:32:25 Signed-off-by: Koen Kooi Dec 29 14:37:34 hmmm doesn't seem to fix it Dec 29 14:37:49 i'm using DEPENDS := "${@oe_filter_out('^(libice|libsm|virtual/libx11)$', '${DEPENDS}', d)}" in recipes/dbus/amend.inc, to filter out dependencies on x11 Dec 29 14:38:07 how can i filter out dependencies, that are generated by BBCLASSEXTEND = "native"? Dec 29 14:38:14 e.g. libsm-native Dec 29 14:38:56 dbus-native still depends on libsm-native Dec 29 14:39:10 adding libsm-native to the string above doesn't help Dec 29 14:39:25 and DEPENDS_virtclass-native := "${@oe_filter_out('^(libice-native|libsm-native|virtual/libx11-native)$', '${DEPENDS_virtclass-native}', d)}" doesn't work either Dec 29 14:40:37 obi hm yiu have to wait for kergoth Dec 29 14:45:45 GNUtoo|laptop, I'm triing linking 4.5.3 to 4.5.2 Dec 29 14:45:55 ok Dec 29 14:47:18 GNUtoo|laptop, it worked :) Dec 29 14:47:28 ok I'll do that Dec 29 15:02:14 hi mickeyl Dec 29 15:18:48 sigh Dec 29 15:18:49 | checking for XML::Parser... configure: error: XML::Parser perl module is required for intltool Dec 29 15:18:52 depending on Dec 29 15:19:09 libxml-parser-perl-native doesn't fix the issue Dec 29 15:19:36 Ug indeed Dec 29 15:20:10 We so need to whack perl-native to be used just for building target perl and configure cpan/etc stuff to just install into tmpdir and use host perl Dec 29 15:20:22 Not that it's directly related to your problem atm, that just reminded me :) Dec 29 15:20:36 * kergoth dreams of recipe-specific sysroots Dec 29 15:20:48 hi kergoth Dec 29 15:20:52 hey Dec 29 15:23:30 kergoth, that'd be an also nice thing, yes :) Dec 29 15:29:46 so what should I do for my problem? Dec 29 15:29:54 look into it as usual? Dec 29 15:29:59 or is there an easier fix? Dec 29 15:36:57 hey, i am trying to build an image again, and i keep getting an error about not being able to stat an lsb_release file Dec 29 15:37:14 this just started happening int he past couple of days Dec 29 15:46:10 perl: symbol lookup error: /home/gnutoo/embedded/oe/oetmps/htcdream/sysroots/x86_64-linux/usr/lib/perl/5.8.8/auto/XML/Parser/Expat/Expat.so: undefined symbol: Perl_Tstack_sp_ptr Dec 29 15:46:15 here's the real error Dec 29 15:46:18 obtained with: Dec 29 15:46:24 perl -e "require XML::Parser" Dec 29 15:46:26 in devshell Dec 29 15:47:19 I'll rebuild from scratch Dec 29 15:49:54 Oh good, so it is semi related to my whish list Dec 29 15:50:14 gandhijee: git pull and try again? Dec 29 15:57:08 03Koen Kooi  07org.openembedded.dev * re6ac906835 10openembedded.git/recipes/linux/linux-omap-2.6.37rc/beagleboard/defconfig: Dec 29 15:57:08 linux-omap 2.6.37rc: enable USRP in main config Dec 29 15:57:08 * this will only get built for machines that have the usrp patch set, but this allows to use use a single config for all machines Dec 29 15:57:08 Signed-off-by: Koen Kooi Dec 29 16:27:30 http://pastebin.ca/2033363 Dec 29 16:27:39 how can i see what version i have checked out? Dec 29 16:30:24 git log -n 1 Dec 29 16:30:28 Is that a recent date? Dec 29 16:52:24 kergoth: do you know a solution to what i described above? Dec 29 16:59:23 obi, done any oe_filter_out on DEPENDS_virtclass-native ? Dec 29 16:59:53 gandhijee, I already saw that, some user reported that already on irc(here) Dec 29 16:59:58 but it was some time ago Dec 29 17:02:48 Tartarus: yes. as i wrote, DEPENDS_virtclass-native := "${@oe_filter_out('^(libice-native|libsm-native|virtual/libx11-native)$', '${DEPENDS_virtclass-native}', d)}" doesn't work either Dec 29 17:07:54 hello, basic question about local.conf: (located in ~/overo-oe/build/conf/) DISTRO = "angstrom-2008.1", but in the console it is displayed: DISTRO_VERSION = "2010.7-test-20101229", is this normal? Who can help? Dec 29 17:15:13 GNUtoo|laptop: oh sorry, is there a fix for it? Dec 29 17:18:26 would be thankful if anyone could help out: basic question about local.conf: (located in ~/overo-oe/build/conf/) there is a line DISTRO = "angstrom-2008.1", but in the console it is displayed: DISTRO_VERSION = "2010.7-test-20101229", how can this be explained? Dec 29 17:21:07 what do you mean? Dec 29 17:21:14 local.conf is far from the only place variables are defined. Dec 29 17:21:20 DISTRO and DISTRO_VERSION are two entirely different variables Dec 29 17:23:40 gm Dec 29 17:23:58 gandhijee, I could look at irc logs Dec 29 17:27:45 DISTRO and DISTRO_VERSION are two entirely different variables >> this helps me. Dec 29 17:31:12 gandhijee, get a more recent bitbake or comment the line Dec 29 17:31:24 there were 2 other users with that error Dec 29 17:31:27 one comented the line Dec 29 17:31:37 another was told to upgrade bitbake version Dec 29 17:36:57 o Dec 29 17:38:17 hi, does anyone manage to build glib-2.0_2.26.1 ? Here (oe master + angstrom 2010.x + beagleboard target), it fails with dbus/dbus.h: No such file or directory Dec 29 17:39:07 hi ericben Dec 29 17:39:13 hi GNUtoo|laptop Dec 29 17:39:27 I think mickeyl fighted with it too Dec 29 17:40:46 ericben, did you sent me some mails ? (like for the internship) Dec 29 17:40:51 or for the mx31? Dec 29 17:40:57 because I didn't receive them if so Dec 29 17:41:03 GNUtoo|laptop: not yet, end of year was terrible Dec 29 17:41:06 ok Dec 29 17:41:07 np at all Dec 29 17:41:11 I just wanted to know Dec 29 17:41:37 btw I'm building from scratch, let's see if I've the dbus.h issue Dec 29 17:43:16 that's strange because dbus is not listed as a dependency here : http://library.gnome.org/devel/glib/2.26/glib-building.html Dec 29 17:44:01 ericben its the internal glib implementation of dbus Dec 29 17:44:09 which comes with 2.26 Dec 29 17:44:13 ask mickeyl Dec 29 17:45:05 woglinde: ok will do if I don't find a solution, in the makefile there are CFLAGS & LIBS flags for dbus-1.0 Dec 29 17:47:15 yes, he needs it for fso Dec 29 17:52:46 interesting, this .h is only needed for one test application in glib-2.0 ! Dec 29 17:53:01 so a quick fix is not to compile this test app Dec 29 17:53:44 hehe Dec 29 17:54:10 maybe they forget to remove it when migrating to their internal dbus implementation Dec 29 17:54:19 hihi Dec 29 17:54:45 and as dbus need glib ... we can't have dbus when compiling glib Dec 29 17:55:40 yes Dec 29 17:55:48 thats what mickeyl hit too Dec 29 17:55:54 now preparing a patch Dec 29 17:56:35 and maybe I'll manage to run something compiled here on my beagleboard XM ;) Dec 29 17:57:41 till later Dec 29 17:58:19 :) I would have expected beagleboard + angstrom 2010 to work out of the box ... Dec 29 18:00:28 interesting configure believes dbus is present which explains why this test is enabled Dec 29 18:01:24 ericben, stuff is always breaking ..... Dec 29 18:01:58 ericben: http://comments.gmane.org/gmane.comp.handhelds.openembedded/40368 Dec 29 18:02:39 JaMa: I think it's not worth to add dbus as a dependency for glib as it's only to build a test application Dec 29 18:04:03 now removing dbus-dev from my host to see if configure finds the host's dbus ... Dec 29 18:05:07 ok Dec 29 18:11:40 it seems that pkgconfig finds the dbus-1.pc which is in the i686-linux sysroot (because of dbus-native which was build before) Dec 29 18:12:35 hi all Dec 29 18:18:26 hi morphis Dec 29 18:18:33 I've the same problem than you Dec 29 18:18:37 we created a machine Dec 29 18:18:42 palmpre in your case Dec 29 18:18:42 is it normal to have PKG_CONFIG_PATH set to STAGING_LIBDIR/pkgconfig *and* STAGING_LIBDIR_NATIVE/pkgconfig ? Dec 29 18:18:45 htcdream in mine Dec 29 18:19:00 and similar machine appeared that have little differences Dec 29 18:19:13 palm-pre plus and palm-pre 2 in your case Dec 29 18:19:16 htcmagic in mine Dec 29 18:19:18 so.... Dec 29 18:19:22 there is machine class Dec 29 18:19:37 ericben: iirc, yes, because there are cases where a recipe asks pkg-config for something it wants to *run* to build, as well as things it wants to link against Dec 29 18:19:44 could be remembering incorrectly, though Dec 29 18:19:53 is there a solution that would prevent re-making all overrides Dec 29 18:20:14 hmmm Dec 29 18:20:22 because most of it would be the same Dec 29 18:20:27 like configs Dec 29 18:20:31 like xorg.conf Dec 29 18:20:42 only kernel is changing in the case of the palm-pre Dec 29 18:20:44 hmmm Dec 29 18:20:50 kergoth: ok, in the present case this breaks cross compilation as dbus-native was built and thus dbus-1.pc is present in i686-linux sysroot so pkgconfig finds it when trying to configure glib-2.0 for arm ... Dec 29 18:21:00 morphis, do you have a lot of overriden configs? Dec 29 18:22:01 for palm-pre I bet that could be done Dec 29 18:22:09 put palmpre as machine class Dec 29 18:22:17 so it keeps everything Dec 29 18:22:28 and override the last bits that are machine specific? Dec 29 18:23:19 morphis_, back? Dec 29 18:23:35 GNUtoo|laptop: jepp Dec 29 18:23:39 ericben: either we need to ensure it finds the target version before the build version, if that's the problem, or its a feature being enabled based on whether something is available, which we should never let builds do, make it explicit instead Dec 29 18:23:46 afaict, anyway Dec 29 18:24:12 ah I remember that oe couldn't be touched while compiling Dec 29 18:24:18 it could cause some issues Dec 29 18:24:21 * kergoth shrugs Dec 29 18:24:43 GNUtoo|laptop: I don't do something like that ... gnome just not reacts anymore Dec 29 18:24:48 (theses christmaths fest are tyring in the physical sense....so I'm tired) Dec 29 18:25:11 morphis, ??? Dec 29 18:25:13 gnome issue? Dec 29 18:25:16 dmesg? Dec 29 18:25:30 maybe OOM Dec 29 18:25:56 I had that issues when I didn't have enough swap Dec 29 18:27:05 nithing Dec 29 18:27:10 s/nithing/nothing/ Dec 29 18:28:03 nothing about MCE either? Dec 29 18:28:09 MCE? Dec 29 18:28:09 (machine check exception) Dec 29 18:28:16 no Dec 29 18:28:18 basically overheating of CPU Dec 29 18:28:36 no, as I can use the desktop but can't change control focus Dec 29 18:28:38 joke: use xmonad Dec 29 18:28:44 I am then stuck to the last control I used Dec 29 18:28:45 :) Dec 29 18:28:47 or awesome Dec 29 18:28:50 yes Dec 29 18:28:54 maybe I should do that ... Dec 29 18:29:19 you could also wipe some settings Dec 29 18:29:22 remove 3d Dec 29 18:29:29 (the effects) Dec 29 18:30:31 kergoth: you are right. I'll add a configure_append to fix config.h Dec 29 18:31:00 GNUtoo|laptop: jepp, will do that when I have some more time Dec 29 18:31:02 ericben: which of the cases was it? conditionally enabling a feature, or finding native before target? Dec 29 18:31:51 kergoth: none ! target dbus is not yet compiled because it depends on glib-2.0 which configure detects dbus-1.pc of the native dbus Dec 29 18:32:30 so you likely need to force it to not enable dbus despite its detection then? Dec 29 18:32:35 JaMa, hi, GNUtoo|laptop and me had problems when building libdb (db package). db is looking for some gcc files from 5.4.2 (which is weird because 5.4.3 is available). The 5.4.2 dir only have the subdirs that are too in 5.4.3, but not the files which db is willing to find there Dec 29 18:32:53 kergoth: unfortunatly there is no flag to disable it so I'll do this with sed Dec 29 18:33:08 removing the 5.4.2 dir and linking that to 5.4.3 worked for us as a workaround Dec 29 18:33:17 ericben: or patch configure.ac to add a flag ;) Dec 29 18:33:36 pespin, I'm rebuilding from scratch Dec 29 18:34:01 GNUtoo|laptop, ah ok :) Dec 29 18:36:12 kergoth: ping Dec 29 18:36:22 obi: pong Dec 29 18:37:47 kergoth: did you read my question above, regarding the modification of dependencies of recipes which use BBCLASSEXTEND = "native"? Dec 29 18:38:24 4 hours ago Dec 29 18:38:55 i'm not sure what the best solution to that is -- its one of those cases where the order that things happen in is critical Dec 29 18:39:48 would it make a difference if i used bbappend instead of amend.inc? Dec 29 18:40:09 possibly, due to the implementation Dec 29 18:40:18 bbappend is pulled in prior to overrides/anonymous python execution Dec 29 18:40:32 whereas amend.inc happens *during* anonymous python execution, becaues that's how its implemented Dec 29 18:40:41 and of course, native.bbclass does its depends mangling in anonymous python Dec 29 18:40:48 so if it runs before/after the amend.inc one... Dec 29 18:40:51 you can imagine the potential issues Dec 29 18:41:10 * kergoth hasn't had enough caffeine yet today though, not 100% Dec 29 18:41:20 ok Dec 29 18:41:23 anon python should die Dec 29 18:41:50 kergoth: what if we cut on target DEPENDencies Dec 29 18:42:11 will then BBCLASSEXTEND ignore appending -native Dec 29 18:42:15 to the one we punt out Dec 29 18:42:24 ideally we'd definitely cut down on anonymous python, and what we *do* need to do that way should be moved to event handlers based on the RecipeParsed event, to make it completely clear when they're processed Dec 29 18:42:36 hmm, yeah, should Dec 29 18:42:58 yeah Dec 29 18:53:21 03Martin Jansa  07master * r52d7143606 10openembedded.git/recipes/linux/ (linux-kexecboot_git.bb linux_git.bb): Dec 29 18:53:21 linux(-kexecboot): bump SRCREV for rc8 Dec 29 18:53:21 Signed-off-by: Martin Jansa Dec 29 19:27:10 should i be able to build the same stuff as i did with 1.8 as with 1.10? Dec 29 19:31:10 Yes Dec 29 19:31:30 1.10 contains various improvements Dec 29 19:31:49 and bitbake master (as in, getting it from its git tree) is faster still, especially with the initial parse Dec 29 19:35:49 hmmm i busted up something then Dec 29 19:55:51 hi rp Dec 29 20:16:34 re Dec 29 20:17:50 re florian Dec 29 22:21:58 obi Dec 29 23:34:00 03Richard Purdie  07master * r683894e3a9 10bitbake.git/lib/bb/fetch/__init__.py: Dec 29 23:34:01 fetch: Ensure SRCREV is still set correctly if the pn-X override has been expanded Dec 29 23:34:01 Signed-off-by: Richard Purdie Dec 29 23:34:11 03Chris Larson  07master * rd68a4e7ccf 10bitbake.git/lib/bb/__init__.py: Dec 29 23:34:12 bb.debug: handle caller not passing a debug level Dec 29 23:34:12 This is a compatibility measure, as previous versions did not error out when Dec 29 23:34:12 this occurred, and will most likely be dropped in the next bitbake minor Dec 29 23:34:12 version in favor of a fatal error. Dec 29 23:34:12 Signed-off-by: Chris Larson Dec 29 23:34:13 03Chris Larson  07master * r8c0afc13ad 10bitbake.git/lib/bb/fetch/git.py: Dec 29 23:34:13 fetch: resurrect branch inclusion in git revision key Dec 29 23:34:14 This is similar to Richard Purdie's commit, but falls back to checking the Dec 29 23:34:14 persist_data database for the old key if the new does not exist, so as to Dec 29 23:34:15 retain compatibility for at least one bitbake minor version. Dec 29 23:34:15 Signed-off-by: Chris Larson Dec 29 23:34:16 03Richard Purdie  07master * r2b9cd8c4d8 10bitbake.git/lib/bb/fetch/__init__.py: Dec 29 23:34:16 fetch: Fix named SRCREVS variable name ordering Dec 29 23:34:17 Retains backward compatibility. Dec 29 23:34:17 Signed-off-by: Richard Purdie Dec 29 23:39:11 bah, missed a spot, fixed latest_revision but not sortable_revision Dec 29 23:39:12 * kergoth mutters Dec 29 23:39:15 * kergoth fixes Dec 29 23:42:55 I thought only I muttered to myself Dec 29 23:49:37 03Chris Larson  07master * r85e5a357ab 10bitbake.git/lib/bb/fetch/git.py: Dec 29 23:49:37 fetch: fix git rev key compat for sortable Dec 29 23:49:37 Signed-off-by: Chris Larson Dec 29 23:57:43 okay, any volunteers to test this bitbake branch? Dec 30 00:54:59 * kergoth ponders Dec 30 00:55:13 * kergoth just had an idea for a different take on the environment filtering / reexporting / etc issue Dec 30 00:56:16 03Chris Larson  07master * r4d9e3f12a2 10bitbake.git/lib/bb/utils.py: Dec 30 00:56:17 utils: kill unnecessary duplicated chmod Dec 30 00:56:17 Signed-off-by: Chris Larson Dec 30 01:43:04 * kergoth grumbles Dec 30 01:44:16 03Chris Larson  07master * rd88dfcc749 10bitbake.git/lib/bb/fetch/git.py: Dec 30 01:44:16 fetch: handle no sortable_buildindex for git Dec 30 01:44:16 Signed-off-by: Chris Larson **** ENDING LOGGING AT Thu Dec 30 02:59:58 2010