**** BEGIN LOGGING AT Sat Sep 19 02:59:59 2009 Sep 19 03:37:32 argh Sep 19 03:37:43 udev-141 with a modified ppoll() call is still borked Sep 19 03:54:09 anyone know if tinylogin configures itself to enable shadow ? Sep 19 03:59:14 in case anyone was wondering themselves Sep 19 03:59:14 Enable using shadow passwords Sep 19 03:59:15 #define CONFIG_FEATURE_SHADOWPASSWDS Sep 19 03:59:23 from Config.h in patched tinylogin src. Sep 19 04:01:23 but from Makefile: USE_SYSTEM_SHADOW = true Sep 19 04:02:11 which means you need CONFIG_USE_BB_SHADOW=y in busybox defconfig, i think Sep 19 04:03:31 is that right? Sep 19 06:00:49 does /etc/default/volatiles only test if the file exists at boot, and create it if its not found? Sep 19 06:01:18 ie. if i make /var/log a directory instead of a symlink, and leave 00_system unchanged, will it overwrite/modify my files on boot? Sep 19 06:01:36 s/00_system/00_core Sep 19 09:06:27 Hi Sep 19 09:07:28 I'm compiling console-images. The "problem" is just that my kernel is compiles without wireless extension., so madwifi-ng fails. How do I remove madwifi from task-base ? Sep 19 09:44:08 good morning Sep 19 11:11:26 morning Sep 19 11:11:34 hmm. now shasum-native.bb is broken. Sep 19 11:14:26 hah. I had TMPDIR="tmp-foobar" Sep 19 11:14:33 without an absolute prefix. no wonder things broke Sep 19 11:23:13 heh Sep 19 11:26:35 hello pb____ Sep 19 12:07:30 pb____ the thumb/binutils thingy works Sep 19 12:07:36 uclibc + thumb = ok. Sep 19 12:12:13 czr_: excellent Sep 19 12:20:07 indeed Sep 19 12:20:17 what's the best way to stability oe for my use now? make a git branch? Sep 19 12:20:23 stabilize even Sep 19 12:22:14 btw, who's responsible for micro.conf? Sep 19 12:22:33 ah, martin lund. is he around on irc? Sep 19 12:25:36 yes, a branch would probably be a good idea. Sep 19 12:25:46 * czr_ nods Sep 19 12:25:56 next best would be to just make a conf file to nail down the PREFERRED_VERSIONs of all the packages that you're interested in Sep 19 12:26:12 how would I go about doing that? Sep 19 12:26:24 I think kergoth wrote a script for it. let me have a look. Sep 19 12:26:29 also, aren't some of the preferred versions overridden in distro.conf, sane-toolchain and such? Sep 19 12:26:36 thanks Sep 19 12:27:25 ah, try "inherit lockdown" Sep 19 12:27:38 assume I have no idea how to do that? :-) Sep 19 12:27:43 in local.conf? Sep 19 12:27:55 looks like that will generate a preferred-versions.inc in your TMPDIR containing the current versions of everything, and you can later include that to freeze everything at the same version Sep 19 12:28:00 right, add a line like: Sep 19 12:28:04 INHERIT += "lockdown" Sep 19 12:28:06 in local.conf Sep 19 12:29:24 as for micro.conf itself, yes, martin lund contributed it in the first place. he seems to have gone a bit quiet since then, though, and in practice I have mostly been maintaining it recently. Sep 19 12:29:56 ah, I'm thinking whether the PREFERRED_INSTRUCTION_SET should be set conditionally? Sep 19 12:30:08 now it's set with = to thumb. this would allow overriding it in local.conf Sep 19 12:30:45 or will this break in some intricate way somewhere down the recipes? Sep 19 12:30:52 yeah, that's not a bad idea Sep 19 12:31:20 in general I am not too keen on having lots of user-visible knobs to twiddle, but I guess there are valid reasons why one might want to build for arm. Sep 19 12:31:27 like running on a cpu with no thumb support I suppose :-} Sep 19 12:31:54 yeah. or maybe someone is bound to a vendor kernel which has no thumb enabled. and so on. Sep 19 12:32:43 btw, for the lockdown thingy, do I need to build some specific target, or any target will do? Sep 19 12:33:00 I think any target will do Sep 19 12:33:11 03Michael 'Mickey' Lauer 07org.openembedded.dev * ra12a9f2b50 10openembedded.git/conf/checksums.ini: checksums: add libcanberra 0.17 and vala 0.7.6-fso2 Sep 19 12:33:11 03Michael 'Mickey' Lauer 07org.openembedded.dev * r07342fa6ce 10openembedded.git/recipes/pulseaudio/libcanberra_0.17.bb: libcanberra: add 0.17; misc bug fixes and includes vapi bindings now Sep 19 12:33:12 03Michael 'Mickey' Lauer 07org.openembedded.dev * rd7f889b1a9 10openembedded.git/recipes/vala/ (4 files): vala: update to 0.7.6-fso2 Sep 19 12:33:15 dos1: here you are Sep 19 12:33:17 take all three Sep 19 12:33:19 right. going to rebuild micro-base-image (have it already built anyway, so it should be quick) Sep 19 12:33:22 morning pb____ q Sep 19 12:33:25 s/q// Sep 19 12:33:30 mickeyl: thanks :) Sep 19 12:33:32 mickeyl: good afternoon q Sep 19 12:33:37 hehehe Sep 19 12:34:07 and a very pleasant afternoon it is here, too. warm and sunny, contrary to what you might expect in cambridge. Sep 19 12:34:14 off to the usa tomorrow though. Sep 19 12:34:21 where are you atm.? Sep 19 12:34:53 at home, right now. flying to massachusetts tomorrow night. Sep 19 12:35:06 ah. business or vacation? Sep 19 12:35:16 bit of both. well, business for me, vacation for my family. Sep 19 12:35:26 how long will you stay there? Sep 19 12:35:31 eight days I think Sep 19 12:35:40 sounds good... the vacation part, that is ;) Sep 19 12:35:45 pb____: not joining the rest of the world at plumbers? Sep 19 12:35:46 I have to work mon-fri next week, then we have the weekend for doing tourist things, then back home monday Sep 19 12:35:59 not too bad Sep 19 12:36:01 btw., i reckon we should take warm clothes to OEDEM, right? Sep 19 12:36:13 yeah, it will be cold, dark and wet here in november. Sep 19 12:36:17 *nod* Sep 19 12:37:20 03Phil Blundell 07org.openembedded.dev * rb50202f4be 10openembedded.git/conf/distro/micro.conf: Sep 19 12:37:20 micro: set PREFERRED_ARM_INSTRUCTION_SET weakly so local.conf can Sep 19 12:37:20 override Sep 19 12:37:42 looks like the final headcount for oedem is going to be 20-ish people, which should be fine for our office. Sep 19 12:38:33 ok, cool. that's a bit more than i had expected, for discussion we might need a moderator then i'm afraid Sep 19 12:39:17 hmm, 20 man battle royale :-) Sep 19 12:39:25 hehe Sep 19 12:39:50 yeah. of course, not all of those 20 will necessarily show up, but I would expect we will have at least 15. Sep 19 12:40:19 we should probably arrange for each session to have a moderator, yes, otherwise it will just get out of hand. Sep 19 12:40:43 hello! Anyone used the prebuilt CodeSourcery with OE? Sep 19 12:41:53 should probably get the TSC in place as an early item, then technical items can be discussed meaningfully Sep 19 12:42:53 *nod* Sep 19 12:43:48 ideally squash any eV related stuff into a morning, as that stuff can always happen via email Sep 19 12:46:05 pb____, thanks Sep 19 12:46:08 yeah, I guess it would make sense to set aside the first morning for eV Sep 19 12:46:23 * rkirti interrupts blatantly, sends in an unsolicited reminder about OE/OEDEM tshirts. *eagerly looking forward to those* ~_~ Sep 19 12:46:45 I will just leave that slot empty in the agenda and the eV board can decide what exactly they want to do with the time. Sep 19 12:47:39 am I correct to understand that most of the OE devs (and hence oedem participants) are in the uk? Sep 19 12:47:52 czr_: not at all Sep 19 12:47:52 no, not really. Sep 19 12:48:00 I think there are about four active oe people in the uk Sep 19 12:48:07 czr_: I would think .de probably has the most people per country Sep 19 12:48:12 ah, k. Sep 19 12:48:30 03Michael 'Mickey' Lauer 07org.openembedded.dev * rd5b4f39938 10openembedded.git/recipes/ (images/fso2-console-image.bb tasks/task-fso2-compliance.bb): fso2-console-image: new recipe for fso2-compliance Sep 19 12:48:34 well, the usa probably has the most by country. germany probably has the most as a percentage of its total population. Sep 19 12:49:03 heh time for google map of write keys :-) Sep 19 12:49:41 I think most of the oedem attendees will be from europe, but there are a couple of folks planning to travel from the usa. Sep 19 12:50:22 ah, cool. Sep 19 12:50:43 is it a core dev thing only or are oe users allowed as well? Sep 19 12:51:06 czr_: the concept of core dev died with the eV Sep 19 12:51:42 yes. electron Volt eats all. although you must mean different eV, right? Sep 19 12:52:10 czr_: not for profit organisation under .de rules Sep 19 12:52:29 ah. Sep 19 12:52:42 rkirti: are you planning to attend oedem? Sep 19 12:52:50 czr_: these are interesting times as OE is growing up Sep 19 12:53:16 03Leon Woestenberg 07org.openembedded.dev * rd9aac9ea22 10openembedded.git/ (3 files in 3 dirs): Sep 19 12:53:16 linux-kirkwood: Add kernel support for openrd-base. Sep 19 12:53:16 Signed-off-by: Leon Woestenberg Sep 19 12:53:18 XorA, indeed Sep 19 12:53:18 czr_: yes, users and other stakeholders are welcome. Sep 19 12:53:36 cool, maybe I'll have a chance to participate next year (this year came as a bit of surprise). Sep 19 12:53:36 Damn just learned about OEDEM Sep 19 12:53:40 I should read the ML more often Sep 19 12:54:31 and we really like people who bring free hardware :-) Sep 19 12:54:42 Ainulindale: there is still time to register if you want, it's not for another eight weeks. Sep 19 12:54:49 pb____: travelling from India so not possible for me :-/ (hopefully I ll make it next year) also, I have end of semester exams in November. But I had asked on the channel earlier (I think discussed with Crofton and florian), if you could bring up tshirts for OEDEM , and make them available for non-attendees Sep 19 12:54:49 Yeah I just forwarded that to my boss Sep 19 12:54:51 XorA, heh. who doesn't :-) Sep 19 12:55:10 We may come a friend and I, we're working on GTA02 stuff and other hackable devices Sep 19 12:55:25 Ainulindale: yay! Sep 19 12:55:25 Ainulindale: very good Sep 19 12:55:27 And I intend to make mickeyl pay his round of beers anyhow Sep 19 12:55:29 :) Sep 19 12:55:36 mickeyl: I could show you a nanonote :-p Sep 19 12:55:43 good Sep 19 12:56:08 (and no it's not an excuse to buy Dr. Pepper) Sep 19 12:56:38 :D Sep 19 12:56:45 rkirti: I think t-shirts may be restricted to attendees only, I'm not sure it's going to be practical to provide them for everyone. Sep 19 12:56:48 we'll see, though Sep 19 12:56:55 are you guys still in posession of a ROAD device? Sep 19 12:57:01 mickeyl: I don't think so sadly Sep 19 12:57:01 any updates wrt. global availability? Sep 19 12:57:04 khorben had one Sep 19 12:57:04 ah, bummer Sep 19 12:57:14 But I think it was only for test purposes at first Sep 19 12:57:16 apart from the palm pre, that's another shallow hope of mine Sep 19 12:57:27 mickeyl: Right now we're playing with sheevaplug and trying to make h:1 get out of its nest Sep 19 12:57:34 *nod* Sep 19 12:57:37 And I'm seriously thinking more and more about switching it from debian to OE Sep 19 12:57:41 Because debian sucks for that kind of job Sep 19 12:57:55 word! Sep 19 12:57:56 Although "debian" distro is a good thing not to fall into the "Yet another OE distro" thing Sep 19 12:58:15 I'd like to find a good middle-term solution Sep 19 12:58:26 But on the business point of view Sep 19 12:58:30 right Sep 19 12:58:36 I think a debian embedded distro as we do it is a dead end Sep 19 12:58:42 Too much time to maintain it properly Sep 19 12:58:46 i agree Sep 19 12:58:50 Plus I'm kinda OE fluent myself Sep 19 12:58:55 Or I have been :-) Sep 19 12:59:00 we should take some time to chat about business opportunities when we meet in november Sep 19 12:59:07 If I go there :-) Sep 19 12:59:15 all these iphone jobs are boring ;) Sep 19 12:59:16 Or do you intend to come in Paris again? Sep 19 12:59:19 Well yeah Sep 19 12:59:30 if we don't meet in Cambridge, i'd come over to paris in december Sep 19 12:59:35 Well Sep 19 12:59:40 I may go to Köln in december Sep 19 12:59:48 ah, good Sep 19 12:59:50 what for? Sep 19 12:59:51 I don't know yet Sep 19 12:59:53 Well Sep 19 12:59:58 My girlfriend's sister will study there Sep 19 13:00:02 (to learn german) Sep 19 13:00:30 So picture two russian girls and a french boy, and possibly a russian-only babouchka in Köln in december :-) Sep 19 13:00:45 Got to be fun Sep 19 13:00:48 sounds.... interesting ;) Sep 19 13:00:56 at least the russions won't freeze Sep 19 13:01:03 it's gotta be cold then Sep 19 13:01:05 for us Sep 19 13:01:39 well Sep 19 13:01:45 I'd rather be cold than warm Sep 19 13:02:42 * czr_ is russian, was born on the polar ice sea shore, and feels cold all the time during winters. Sep 19 13:02:52 so. no :-). Sep 19 13:03:07 well, I'm french and born in the southern part of France Sep 19 13:03:22 And I can't bear any kind of warm weather Sep 19 13:03:26 heh Sep 19 13:03:37 I actually like warm much better, as long as it's not tropical. Sep 19 13:03:51 czr_: Meanwhile, привет :-) Sep 19 13:03:58 privet indeed :-). Sep 19 13:04:15 czr_: well anything warmer than 20°C gives me headaches Sep 19 13:04:33 maybe you should visit finland then :-) Sep 19 13:04:48 although it'd been unseasonably warm here now for last 1.5 months. Sep 19 13:04:52 well I was intending to go near Velikie Luki in december Sep 19 13:04:57 But it might be Köln instead Sep 19 13:05:03 heh. Sep 19 13:05:19 Depends on my girl :-) Sep 19 13:06:30 well, velikie luki is close enough I guess :-) Sep 19 13:36:53 mickeyl: I started updating the oedem wiki page with an initial agenda. Feel free to bash that around. Sep 19 13:39:31 pb____: url? :-) Sep 19 13:39:52 Doesn't matter, found it :-) Sep 19 14:00:36 is there a way to set the DISTRO_TYPE to debug only for one package? (dropbear, I want the passwd:less root login to work) Sep 19 14:11:45 DISTRO_TYPE_dropbear = "debug" Sep 19 14:11:55 er, not quite Sep 19 14:11:58 DISTRO_TYPE_pn-dropbear = "debug" Sep 19 14:14:03 ah. I think I tried the first one, it didn't work at least :-) Sep 19 14:14:23 what does the pn- signify? Sep 19 14:18:31 czr_: its a name space addition to show the override is only valid inside ${PN} Sep 19 14:20:16 pb____: great, thanks Sep 19 14:20:21 right. so basically by using a similar construct (FOO_pn-package-name) I could override anything? Sep 19 14:20:29 right Sep 19 14:20:49 there's no magic to the pn- bit, see the definition of OVERRIDES in bitbake.conf Sep 19 14:20:50 cool. are these things documented somewhere? Sep 19 14:20:59 czr_: yes, Ive used that to great effect to make DISTRO hacks to packages without editing the .dbb Sep 19 14:21:32 I think OVERRIDES is documented in the fine manual Sep 19 14:21:46 and it's almost certainly in the poky manual if it isn't in the bitbake one Sep 19 14:22:15 right. one last question. I've rebuilt dropbear now, how do I rebuild a micro-base-image that will contain the new dropbear? I tried -c clean micro-base-image and then rebuilding it, but it's not part of it. Sep 19 14:23:03 s/not part of it/not replaced/ Sep 19 14:24:30 oh well, maybe I'll just rebuild the whole thing :-). np Sep 19 14:24:41 03Michael 'Mickey' Lauer 07org.openembedded.dev * r86d4dd34f8 10openembedded.git/recipes/vala/vala_0.7.6-fso2.bb: vala 0.7.6: make it build Sep 19 14:25:22 hm, dunno. it should pick it up automatically, no need even to clean. Sep 19 14:25:57 hmm. I'll check it later, need to watch a movie with gf now. thanks again for all the help Sep 19 14:30:05 hi Sep 19 14:32:48 there are sombody? Sep 19 14:49:16 03Stanislav Brabec 07org.openembedded.dev * rbdd37476c0 10openembedded.git/recipes/audiofile/ (5 files in 2 dirs): audiofile: Security fix CVE-2008-5824, coding style fixes from openSUSE, recipe simplified. Sep 19 15:22:33 good morning florian Sep 19 15:24:07 hi all Sep 19 15:25:59 ~curse DHL Sep 19 15:26:00 May the fleas of a thousand camels infest your most sensitive regions, DHL ! Sep 19 15:37:33 hello all Sep 19 15:37:35 hallo eFfeM Sep 19 15:39:57 morning kergoth, likewise Sep 19 15:41:37 hi Sep 19 15:45:10 gm all Sep 19 15:49:50 hi likewise Sep 19 16:04:30 hoi likewise Sep 19 16:04:36 everything ok ? Sep 19 16:05:05 sure, playing with the openrd-base today Sep 19 16:05:17 cool Sep 19 16:05:22 going well? Sep 19 16:05:45 eFfeM: yup, pci express and fpga looking good Sep 19 16:06:01 you've added an fpga? or is it already on the board? Sep 19 16:07:36 eFfeM: altera development board Sep 19 16:07:45 eFfeM: not on-board Sep 19 16:08:03 likewise: this connects through pci express? Sep 19 16:09:29 hm, client says: (WW) OS did not count PCI devices, guessing wildly Sep 19 16:14:04 03Leon Woestenberg 07org.openembedded.dev * re5fc5e963f 10openembedded.git/recipes/linux/linux-kirkwood_2.6.31.bb: Sep 19 16:14:04 linux-kirkwood-2.6.31: Fetch newer mach-types for MACHINE openrd-base only. Sep 19 16:14:04 Signed-off-by: Leon Woestenberg Sep 19 16:14:07 hmm. does anyone know a reason why strace is still at 4.5.15? Sep 19 16:14:12 (there's a 4.5.18 out) Sep 19 16:14:18 probably noone bothered to upgrade. Sep 19 16:14:25 too many recipes for so few folks Sep 19 16:14:28 patches appreciated :) Sep 19 16:14:41 np :-). will do. Sep 19 16:14:47 cool Sep 19 16:14:59 I have a recipe for 4.5.18 already Sep 19 16:15:08 ah, very good Sep 19 16:15:12 but I haven't tested against dev yet. it was against angstrom-2007-smt. Sep 19 16:15:18 | {standard input}:4346: Error: selected processor does not support `mla r3,r1,r6,r3' Sep 19 16:15:19 still needed one patch though. Sep 19 16:15:20 hmm Sep 19 16:15:29 openocd_svn is stuck at 570 something in sane-rev, we have 2725-something out. Who may decide to move the version number? (May I?) Sep 19 16:15:48 likewise: iirc i added it, feel free to bump Sep 19 16:16:01 Who decides about what's "sane" in .dev? Sep 19 16:16:02 at least with openocd you know where you stand. pick any rev, none will work :-) Sep 19 16:16:07 * czr_ hides & runs Sep 19 16:16:11 likewise: the "maintainer" to its best knowledge Sep 19 16:16:22 I ended up doing a binary search with openocd:s revs to find out a version that worked. Sep 19 16:16:37 czr_: yeah, openocd is VERY machine and version dependent Sep 19 16:16:44 yeah. don't I know it :-) Sep 19 16:16:53 plus, I never got it to work with pxa270 anyway. Sep 19 16:18:13 btw, when adding new patches, how do I tell where I found the patch at? Sep 19 16:18:15 or do I need to? Sep 19 16:18:26 (this would be for the 4.5.18 strace) Sep 19 16:19:58 well, ideally you encode this knowledge as comment in the file Sep 19 16:20:03 (the .diff) Sep 19 16:20:20 ah yes. I guess I could just add anything just before the first hunk Sep 19 16:20:29 or grab the patch from the original location in the first place Sep 19 16:20:39 I found it on a mailinglist. Sep 19 16:20:48 hmm, ok, then as you said Sep 19 16:22:01 I'll need some help from people testing it on non-arms as well. Sep 19 16:22:16 but, maybe I'll just submit the patch and see what will happen to it Sep 19 16:23:54 czr_: We also have DEFAULT_PREFERENCE = "-1" if you want your version not to be the preferred version (due to lack of testing). People can still hand-pick that version for testing. Sep 19 16:24:07 excellent. Sep 19 16:26:16 hmm. actually, I didn't find it on a mailinglist (I found a reference to it on ML). Anyone know what would be the canonnical source for this? : http://bugs.gentoo.org/attachment.cgi?id=181917&action=view Sep 19 16:26:58 (that's needed for strace 4.5.17 and 4.5.18) Sep 19 16:29:23 such SRC_URIs are always tough for our fetcher Sep 19 16:29:27 just grab it into oe and add a note Sep 19 16:29:33 * czr_ nods Sep 19 16:31:22 hmm. whats the easy way to install anything from ipk on a micro-base-image? (it doesn't have ipkg by default) Sep 19 16:31:57 nm though, I'll just copy the binary Sep 19 16:34:09 hmm. interesting. the recipe that worked for angstrom-2007.12 doesn't work against current oe :-). any idea where to look for? : http://pastebin.com/m2ab18313 Sep 19 16:34:28 -> afk for food Sep 19 16:46:19 mickeyl: that is a bit odd. what cpu are you building for? Sep 19 16:46:51 armv4t Sep 19 16:47:08 i guess the offending package is using some eleet optimizations without CPU ARCH guides Sep 19 16:47:12 guards, even Sep 19 16:47:17 oh, are you building for thumb? Sep 19 16:47:28 hmm, om-gta02 actually Sep 19 16:47:31 mla is a perfectly good v4t instruction in arm mode, but it won't work in thumb mode Sep 19 16:47:40 ah Sep 19 16:47:52 perhaps it's just that Sep 19 16:47:56 so, might be eleet asm optimisations without checking for thumb mode Sep 19 16:48:06 where can we switch on/off thumb mode these days? Sep 19 16:48:14 * mickeyl lost in all the switches Sep 19 16:48:26 do you want to turn it off globally, or just for this one package? Sep 19 16:48:37 for this one package first Sep 19 16:48:42 well, I guess it doesn't matter actually, the answer is the same :-} Sep 19 16:48:45 hehe Sep 19 16:48:49 PREFERRED_ARM_INSTRUCTION_SET = "thumb" Sep 19 16:48:52 er Sep 19 16:48:52 ah, right Sep 19 16:48:54 PREFERRED_ARM_INSTRUCTION_SET = "arm" Sep 19 16:48:57 thanks Sep 19 16:49:14 or, in a bb file, you can do it without the preferred bit: just ARM_INSTRUCTION_SET = "arm" Sep 19 16:49:42 which package is it? I'll have a quick look and see if there is any easy way to make that code compatible with thumb Sep 19 16:50:39 it was bailing out in pulseaudio when compiling libbluetooth_sbc Sep 19 16:50:45 thanks! Sep 19 16:51:28 oh, sbc, I think I saw that before Sep 19 16:52:14 see bluez/bluez4/sbc-thumb.patch, I wonder if the pulseaudio package just needs the same treatment Sep 19 17:02:46 angstrom/uclibc building for people currently? Sep 19 17:02:57 I got: ERROR: log data follows (/home/trini/work/droid/droid-oe/tmp/work/db1200-angstro Sep 19 17:02:57 m-linux-uclibc/uclibc-initial-0.9.30-r32.0/temp/log.do_stage.28709) Sep 19 17:02:58 | make: *** No rule to make target `install-headers'. Stop. Sep 19 17:05:00 hmm, looks like that should be install_headers Sep 19 17:05:04 i'll push if so Sep 19 17:05:53 moved past at least :) Sep 19 17:05:55 hi is anybody here using OE under arch linux? Sep 19 17:06:40 hehe. I found solution to the cachectl thingy. khem actually already patches it (uclibc export issue). Sep 19 17:07:59 03Tom Rini 07org.openembedded.dev * r122d4d8bd6 10openembedded.git/recipes/uclibc/uclibc-initial_0.9.30.bb: uclibc-initial-0.9.30: Fix a typo, install_headers not install-headers Sep 19 17:49:35 crumbs, building pulseaudio on top of micro-base-image requires running 1130 tasks Sep 19 17:49:45 it's at 493 now so I guess I have a way to go Sep 19 17:51:52 * Tartarus hates when he forgets to bump threads and -j up, bah Sep 19 17:57:47 heh, I don't think my laptop will cope with massive -j4 settings. Sep 19 17:57:52 should get myself a better computer I suppose Sep 19 17:57:56 heh Sep 19 17:58:02 probably just slows down :-) Sep 19 17:58:17 yeah, only 2GB of memory on this machine, and evolution eats most of it Sep 19 17:58:27 * czr_ nods Sep 19 17:58:34 I think it'll just start swapping with more than a couple of threads Sep 19 17:58:44 how many cores do you have? Sep 19 17:59:02 2, I think Sep 19 17:59:25 yeah, probably no sense using more than 3 then Sep 19 17:59:31 unless you're on a very slow link or smt Sep 19 17:59:34 yeah, seems to be a core 2 duo Sep 19 17:59:40 have to afk now, have fun :-) Sep 19 17:59:44 I dunno Sep 19 17:59:58 I think cores=x2 for threads, and then coresx1.5 or so for jobs Sep 19 18:00:18 Er, no, 1/1.5 Sep 19 18:00:25 * Tartarus does 4/6 on his 4 core boxes Sep 19 18:18:31 yeah, I routinely run -j2 plus 2 threads on the dual-cores at the office and that seems to be fine. Sep 19 18:19:09 but, on this little thinkpad, there is a noticeable amount of disk thrashing going on even with just a single bitbake running in the background. I don't think 4 threads would work well. Sep 19 18:19:15 heh Sep 19 18:19:31 pb___, any chance you're building uclibc/mipsel and if so, running it? Sep 19 18:19:42 uclibc yes, but arm not mips Sep 19 18:19:46 bah Sep 19 18:19:50 there's some mips problem Sep 19 18:19:53 I do most of my testing with qemuarm at the moment Sep 19 18:19:59 shared stuff seg faults Sep 19 18:20:07 oh, still? i thought that was fixed with that always_inline patch that otavio landed yesterday Sep 19 18:20:15 (seen it on real hw, but confirming builds atm in qemu) Sep 19 18:20:25 hm Sep 19 18:20:29 is there a qemu target for mips? I can take a look at it if so, but I don't have any mips hardware on hand. Sep 19 18:20:34 yeah, malta Sep 19 18:20:52 i'm building for just db1200 and passing to malta, that works fine enough :) Sep 19 18:21:01 but lemme double check that that patch is in Sep 19 18:21:31 ok Sep 19 18:21:37 mips ld.so definitely was busted until about yesterday Sep 19 18:21:39 Well, hmm Sep 19 18:21:49 That patch isn't being used for 0.9.30 which is what angstrom is on Sep 19 18:21:58 ah, that might be a problem Sep 19 18:22:16 poking it now Sep 19 18:22:40 applies fine, heh Sep 19 18:22:45 might be worth upgrading to 0.9.30.1, it seems to be working ok for us arm folk :-) Sep 19 18:23:03 yeah, I don't think that code has changed in a while. the patch should apply to any recent-ish version Sep 19 18:23:13 I think it's landed in upstream uclibc now so it should be in the next release Sep 19 18:23:55 angstrom is 0.9.30 for all Sep 19 18:25:11 03Thomas Zimmermann 07shr/import * rad62752f4e 10openembedded.git/conf/distro/include/sane-srcrevs.inc: Sep 19 18:25:11 epdfview: set SRCREV in sane-srcrevs.inc too Sep 19 18:25:11 Signed-off-by: Klaus Kurzmann Sep 19 18:25:14 03Thomas Zimmermann 07shr/import * ra7b6f5686e 10openembedded.git/ (3 files in 2 dirs): Sep 19 18:25:14 Pisi, om-neon: prepare for OE.dev merge Sep 19 18:25:16 Signed-off-by: Klaus Kurzmann Sep 19 18:25:18 03Thomas Zimmermann 07shr/import * r978948636b 10openembedded.git/ (4 files in 3 dirs): Sep 19 18:25:21 epdfview: prepare for OE.dev merge Sep 19 18:25:23 Signed-off-by: Klaus Kurzmann Sep 19 18:25:27 03Thomas Zimmermann 07shr/import * r78b04f36ad 10openembedded.git/ (conf/checksums.ini recipes/python/python-notify_0.1.1.bb): Sep 19 18:25:30 Python: Add python-notify recipe (version 0.1.1) Sep 19 18:25:32 Signed-off-by: Klaus Kurzmann