**** BEGIN LOGGING AT Wed May 22 02:59:58 2013 May 22 07:52:37 good morning May 22 08:11:20 hi all May 22 09:26:35 otavio: ping May 22 11:45:55 hi, how can I add kernel-dev to the sdk.sh installer generated using populate-sdk ? May 22 12:59:25 hi there. i am trying to find relevant documentation on SDK generation, but i don't think i found what i needed. basically i am making my own distro, and i need to 'generate a SDK' which I can distribute to whoever wants to build a lib, an application to run on distro/device. that seems to be the job of ADT, but i don't quite get it. May 22 13:00:23 i think i need a cross compiler + custom rootFS which has all my -dev packages pre-installed. how do i make that? May 22 13:00:58 you probably want "bitbake your-image -c populate_sdk" May 22 13:01:02 which will make an ADT for that image May 22 13:01:30 that being an installable cross compiler and development files for all packages installed in the image May 22 13:59:05 It seems bitbake broke GIT fetcher in master May 22 13:59:10 panda84kde: pong May 22 14:06:10 otavio: what's broken? May 22 14:55:00 bluelightning: May 22 14:55:02 The stack trace of python calls that resulted in this exception/failure was: May 22 14:55:03 File: 'write_srcrev', lineno: 43, function: May 22 14:55:03 0039: if os.path.exists(srcrevfile): May 22 14:55:03 0040: os.remove(srcrevfile) May 22 14:55:04 0041: May 22 14:55:07 0042: May 22 14:55:10 *** 0043:write_srcrev(d) May 22 14:55:25 Exception: AttributeError: 'tuple' object has no attribute 'startswith' May 22 14:55:41 File: 'buildhistory.bbclass', lineno: 20, function: _get_srcrev_values May 22 14:56:00 this is related to last bb changes May 22 17:11:30 Any knows a resolution to this issue: http://bit.ly/12N1IG8? May 22 17:12:32 I am trying to build the angstrom with the latest graphics SDK and I am running into this. May 22 17:13:36 SlashV: you might have more luck in the dedicated angstrom channels May 22 17:13:44 Apparently the drm sources were removed from the SDK, and it is now taking the source from the kernel May 22 17:14:22 rburton: this issue is in meta-ti and posted on the yocto list. May 22 17:16:08 rburton: please look at the issue before telling me to go somewhere else just because you see the word "angstrom" ty May 22 17:17:20 I am *interested* in the resolution within poky, so I can apply it with angstrom May 22 17:18:48 SlashV: its a failure in the beagle BSP, so i recommend asking the beagle people May 22 17:49:04 otavio: did you resolve that somehow? I see the same after bitbake upgrade May 22 17:49:13 otavio: it's one of last 37 commits May 22 17:50:53 even HEAD~15 still works May 22 17:54:07 rev = ud.method.sortable_revision(scm, ud, d, name) May 22 17:54:07 if rev.startswith(autoinc_templ): May 22 17:54:22 and it returns tuple now autoinc, rev = ud.method.sortable_revision(scm, ud, d, name) May 22 17:54:33 so buildhistory.bbclass needs to be updated May 22 18:27:00 JaMa: I reverted back May 22 18:39:53 * mranostay pokes in May 22 21:53:23 So I'm using 'bitbake -c populate_sdk ' to build an SDK. A nativesdk package included by drops some files in /sbin, but the environment-setup-... script doesn't include sbin in the PATH. I found the 'toolchain_create_sdk_env_script' function in meta/classes/toolchain-scripts.bbclass where the environment-setup script is getting created, and I don't see any hooks for including other variables. Is there something I'm May 22 21:53:58 I guess I could do a toolchain_create_sdk_env_script_append() and add it May 22 21:54:31 yeah, I see that done in at least one place May 22 21:54:38 (sorry for the noise) May 22 21:57:02 Yeah, we've done something similar before - https://github.com/MentorEmbedded/meta-sourcery/blob/master/classes/toolchain-scripts-external.bbclass. would be nice to have explicit hooks, though May 22 21:58:27 cool — thanks May 22 23:50:42 sgw_: see my reply regarding consolidated pull. also, I can't see all the changes mentioned in the cgit browser. May 22 23:51:31 Net147: are you Jonathan? yes, I already updates my branch May 23 00:02:38 sgw_: yes I am May 23 00:12:09 Net147: I was not able to apply your updated systemd patch May 23 00:31:54 sgw_: I will look into it. what's the time over there? May 23 00:38:28 Net147: it's 1738 here, where are you located? May 23 00:39:21 sgw_: Sydney, Australia. GMT+10 May 23 00:40:10 sgw_: it looks like the systemd patch still depends on the FHS patch May 23 00:40:33 sgw_: FHS patch modifies systemd_199.bb, systemd patch renames it May 23 00:41:54 sgw_: would it be more convenient to reorder them? May 23 00:42:09 yes, that would help. May 23 00:42:25 sgw_: ok, so FHS patch would depend on systemd update May 23 00:46:42 sgw_: okay, systemd update V4 sent May 23 00:49:37 Net147: thank you, and I apperciate the effort you are putting into this, sorry for requesting multiple changes. May 23 00:52:45 sgw_: you're welcome. I appreciate the feedback. May 23 00:57:45 sgw_: I also have an enhancement that uses MBR disk signature (for kernel 3.8 and later) instead of hardcoded root=/dev/sda2 for boot-directdisk class. would it be useful to submit a patch for it? May 23 01:01:37 Net147: it would be interesting to see, that's for sure, the SYSLINUX_ROOT is override-able, but there may be other places that need to fixed to make it work correctly. May 23 01:03:21 sgw_: http://pastebin.com/t6n7hBpD May 23 01:05:50 sgw_: I have been using it for a week or so and it works. I can boot in both QEMU and on real hardware without changing root= in syslinux.cfg (QEMU uses /dev/hda2 instead of /dev/sda2) May 23 01:12:29 Net147: are you trying to fake out a random uuid? If so, why not use uuid to generate the signature? May 23 01:13:49 sgw_: uuidgen? May 23 01:15:00 yes May 23 01:16:43 sgw_: I have actually submitted patch to parted upstream for that - http://lists.alioth.debian.org/pipermail/parted-devel/2013-May/004374.html May 23 01:17:24 sgw_: but yes I could use the random parts of uuidgen's output (rather than the parts based on MAC address, etc.) May 23 01:18:35 That's probably the better approach then have parted do the correct thing! May 23 01:18:52 Time for dinner here, I might be back on later tonight. May 23 01:19:01 sgw_: so prefer patching upstream parted? May 23 01:56:26 sgw_: okay, I will adapt it to use uuidgen **** ENDING LOGGING AT Thu May 23 02:59:59 2013