**** BEGIN LOGGING AT Thu Apr 30 02:59:57 2009 Apr 30 03:02:15 * Tartarus tries to lure mdp out Apr 30 03:02:16 ga Apr 30 06:44:00 good morning Apr 30 07:15:03 hello all Apr 30 07:15:23 I want to have avahi 0.6.21 in my root filesystem Apr 30 07:15:24 PREFERRED_VERSION_avahi = "0.6.21" Apr 30 07:15:45 this is not working how can i set avahi to this version ? Apr 30 07:47:58 03Steffen Sledz  07org.openembedded.dev * r41d2e60e3d 10openembedded.git/recipes/mii-diag/mii-diag_2.11.bb: mii-diag-2.11: GNU_HASH QA fix Apr 30 07:56:52 Guest86954: where you placed that line? Apr 30 07:57:25 mckoan: inside .bb recipe Apr 30 07:57:36 Guest86954: which .bb? Apr 30 07:57:37 now it's in .conf and it's working Apr 30 07:57:39 thx Apr 30 07:58:11 Guest86954: fine, BTW syntax is PREFERRED_VERSION_avahi ?= "0.6.21" Apr 30 07:59:09 probably = works as well :-) Apr 30 08:44:33 03Steffen Sledz  07org.openembedded.dev * r66efe0dabf 10openembedded.git/recipes/bonnie/bonnie++.inc: bonnie++: GNU_HASH QA fix Apr 30 10:03:43 i'm getting a weird error trying to "bitbake console-image" Apr 30 10:03:47 http://dpaste.com/39555/ Apr 30 10:24:21 hi all Apr 30 10:24:58 that file is not there Apr 30 10:25:44 florian: good morning Apr 30 10:36:02 hmm: http://www.nabble.com/Error-build-omap3-console-image:-control.tar.gz-not-found-td23060663.html Apr 30 10:36:32 morning all Apr 30 10:43:55 hey RP Apr 30 10:48:51 hi rp__ Apr 30 10:56:18 03Graeme Gregory  07org.openembedded.dev * r1fc726a2f3 10openembedded.git/recipes/hal/ (hal.inc hal_0.5.11.bb): Apr 30 10:56:18 hal/hal.inc : change hardcoded etc to ${sysconfdir} and remove bogus directory Apr 30 10:56:18 hal/hal_0.5.11.bb : bump PR for above change Apr 30 10:56:33 good morning Apr 30 11:03:01 it seems that a --with-zlib in EXTRA_OECONF variable in /org.openembedded.dev/recipes/python/python.inc is needed Apr 30 11:03:12 would you guys take a patch for that? Apr 30 11:04:40 kimrhh: push it to email list Apr 30 11:04:53 the python maintainer can then check it out Apr 30 11:05:44 * XorA is having fun with the glamo Xorg driver Apr 30 11:08:49 XorA: k, thz Apr 30 11:08:55 err, thx :) Apr 30 11:12:07 heh, I understood Apr 30 11:20:50 urg to early Apr 30 11:22:55 Crofton|work: its not early :-) Apr 30 11:23:10 It's always too early, just like it's always morning :) Apr 30 11:23:17 right Apr 30 11:23:31 early mean I have not finished two cups of coffee Apr 30 11:24:03 Your addiction is a medical condition and we shouldn't mock you for it. Apr 30 11:31:29 Crofton|work: then it's early for me :-) Apr 30 11:31:31 gm all Apr 30 11:32:33 broonie: are you advocating not mocking the americans? Apr 30 11:32:45 broonie: never thought I would hear such thoughts :-) Apr 30 11:33:15 We have a new leader, no more mocking Americans! Apr 30 11:33:33 well I guess he is half not american :-) Apr 30 11:33:40 so half good :-D Apr 30 11:33:42 heh Apr 30 11:51:05 XorA: -devel@ ? Apr 30 11:52:32 kimrhh: openembedded-devel@lists.openembedded.org Apr 30 11:52:59 yes, cool Apr 30 12:06:54 morning Apr 30 12:17:59 03Stanislav Brabec  07org.openembedded.dev * r5a23eb5eb0 10openembedded.git/recipes/zaurus-updater/encdec-updater.bb: encdec-updater: GNU_HASH QA fix. Apr 30 12:26:57 morning Apr 30 12:27:02 hi likewise rwhitby Apr 30 12:27:55 hi Apr 30 12:29:54 not so good morning here: attack on Queensday; http://www.geenstijl.nl/ Apr 30 12:31:56 .en version here http://news.bbc.co.uk/1/hi/world/europe/8026807.stm Apr 30 12:33:06 my interwebs are sucky today :( Apr 30 12:57:53 http://marcin.juszkiewicz.com.pl/2009/04/30/new-multi-io-card-installed/ Apr 30 12:58:50 03Michael 'Mickey' Lauer  07fso/milestone5.5 * r0234f3f02c 10openembedded.git/recipes/dbus/ (dbus.inc dbus_1.2.1.bb): dbus: bump default timeout value Apr 30 13:02:14 hrw: 6 ports? do you have a link to that product? I could only find 4-port solutions. Apr 30 13:02:26 also, a reset controller for all the boards would be nice Apr 30 13:03:03 likewise: I found it on allegro (Polish auction site) Apr 30 13:03:56 likewise: there are 2c(om)1p(arallel), 4c1p, 6c1p, 6c0p cards and then are 8com ones with hydra of cables Apr 30 13:16:41 gremlin[it]: http://www.embedded.it/?q=content/sheevaplug-mini-pluggable-computer Apr 30 13:20:07 mckoan: too bad it hasn't two RJ45 Apr 30 13:22:58 Ainulindale: indeed! I'm thinking to produce one customized ;-) Apr 30 13:23:26 mckoan: also bring out pcie, spdif, sata etc Apr 30 13:23:40 mckoan: do you do your own h/w design? Apr 30 13:23:41 sheevaplug runs the unmodified SlugOS rootfs (via nfs root) -- sweet. Apr 30 13:23:51 a perfect replacement for my Pentium III servers Apr 30 13:23:54 mwester: congrats Apr 30 13:24:13 That makes it easy to create just another image in the SlugOS distro... Apr 30 13:24:13 mckoan, goooooooooooooooooood !!! for the page and keep me in touch for the custom one ! Apr 30 13:24:45 likewise: there is (or soon be) a devboard kirkwood based for 149$ Apr 30 13:24:46 gremlin[it]: it was an idea I have no plans to produce it Apr 30 13:25:40 hrw, keep me in touch on this ...specially if it have SATA and (WiFi or VIDEO) ! Apr 30 13:25:42 hrw: from whom? Apr 30 13:26:51 mckoan, if will happen can't be so difficult to find a Chinese company to build it ;) Apr 30 13:27:16 gremlin[it]: main cost is not production, but design :-D Apr 30 13:27:25 mckoan, true ! Apr 30 13:27:39 mckoan: design is easy Apr 30 13:27:54 manufacturing cheaply is hard :) Apr 30 13:27:56 ok so let's do it Apr 30 13:28:13 hrw mentions a board is already soon there. hrw? Apr 30 13:28:15 Crofton|work: manufacturing cheaply with acceptable yield is hard :) Apr 30 13:28:28 broonie, good point :) Apr 30 13:28:30 I started a MV78100 design (Discovery Innovation), the big brother of the Sheeva Apr 30 13:29:03 6 watts instead of 2.3, two 4x PCIe, two SATA, two GbE Apr 30 13:29:11 dual-issue etc Apr 30 13:29:26 likewise: I want it ! Apr 30 13:29:44 mckoan: it will be expensive though, we add a high-end FPGA on the PCIe bus Apr 30 13:29:55 heh Apr 30 13:29:57 and videoport, hdmi, dvi, ... Apr 30 13:30:06 likewise: any link? Apr 30 13:30:07 likewise: http://www.nuhorizons.com/development/marvell/kirkwood_open_rd_base.asp Apr 30 13:30:20 but I could try to design a kirkwood design in the sideline Apr 30 13:31:16 who is pfalcon? Apr 30 13:31:45 likewise, have you seen this: http://www.flickr.com/photos/32615155@N00/3447696549/ Apr 30 13:32:01 cbrake, not sure Apr 30 13:32:08 ~seen pfalcon Apr 30 13:32:11 pfalcon was last seen on IRC in channel #handhelds, 672d 17h 33m 27s ago, saying: ';-)'. Apr 30 13:32:19 likewise: hmmmm is it sold in an usable way? Apr 30 13:34:03 Crofton|work: my memory is suggesting Paul Eggleton or Paul Sokolovsky Apr 30 13:34:23 Ainulindale: it will be a dev board, but define usable Apr 30 13:34:35 you trust your memeory more than me :) Apr 30 13:34:45 Paul S would be my guess Apr 30 13:34:49 Crofton|work: what it is? Apr 30 13:34:52 likewise: so no packaging Apr 30 13:34:52 * likewise seconds Apr 30 13:35:13 likewise, access to code for easy customization Apr 30 13:35:36 software radio Apr 30 13:35:55 the USRP daughterboards are radio front ends Apr 30 13:35:59 nice Apr 30 13:36:12 Ainulindale: you mean, no enclosure? Apr 30 13:36:20 board in between has data A/D and D/a convertors and FPGA to reduce sample rate Apr 30 13:36:56 likewise: yeah Apr 30 13:37:05 likewise: Because I was just looking for something like that Apr 30 13:37:20 i.e. two giga-rj45, sata, usb, low power consumption Apr 30 13:37:23 preferrably mini-itx Apr 30 13:37:26 for a router Apr 30 13:37:44 Didn't find anything cheap & usable properly Apr 30 13:37:50 (and noiseless) Apr 30 13:38:08 Ainulindale: use mpc8313e or kirkwood, not mv78100 Apr 30 13:38:23 likewise: "use" as in? :-) Apr 30 13:38:34 Ainulindale: buy mini-itx board :-) Apr 30 13:38:54 Any link? :-) Apr 30 13:39:00 Ainulindale: board $150 or $199, chips $20 Apr 30 13:39:14 Ainulindale: where are you located? Apr 30 13:39:15 chips? which chips? Apr 30 13:39:18 likewise: france Apr 30 13:39:49 Ainulindale: you need pcie or not? Apr 30 13:39:59 I don't care Apr 30 13:40:15 Just sata, preferrably two, and two giga ethernet Apr 30 13:41:44 * hrw tries to get ADS5121 board Apr 30 13:47:59 someone used it? Apr 30 13:52:36 hrw: MPC5121 from freescale? Apr 30 13:52:40 markos_: yes Apr 30 13:53:12 I have a problem with EXTRA_OEMAKE Apr 30 13:53:20 not sure it's a good choice nowadays, many problematic revisions and not sure if it works ok yet Apr 30 13:53:24 the .bb contains EXTRA_OEMAKE="FONTDIR=${datadir}/fonts/truetype" to pass it to make Apr 30 13:53:36 hrw: i have 2x MPC5200B here, fwiw Apr 30 13:53:43 much better in some ways Apr 30 13:53:50 and the make call is indeed "make -j 2 FONTDIR=/usr/share/fonts/truetype" Apr 30 13:53:54 but using a similar core Apr 30 13:53:56 but it gets totally ignored Apr 30 13:54:04 markos_: for 150$ ads5121 sounds good Apr 30 13:54:09 it works just fine using the same call on my ubuntu laptop Apr 30 13:54:34 markos_: I looked for cheap ppc board Apr 30 13:55:03 hrw: well, I happen to know the guy that did the opensuse port to that and Freescale sent him 5 revisions before they got one that actually worked correctly Apr 30 13:55:12 for one the 5121 is not cache coherent Apr 30 13:55:31 hrw: the efika is cheap at $99 (though no case) Apr 30 13:55:40 markos_: but who sells efika now? Apr 30 13:55:41 i have that one here Apr 30 13:55:56 hrw: if you're in the US directron.us have some in stock iirc Apr 30 13:56:25 if not, i have direct contact with Genesi and I can probably help you find one Apr 30 13:56:57 I am in europe Apr 30 13:57:45 hm, there's a guy that sells everywhere, but he's in NZ Apr 30 13:57:50 lemme check Apr 30 13:58:23 http://ausppc.blogspot.com/ that's the guy Apr 30 13:58:34 he makes custom cases for the efika he sells too :) Apr 30 13:59:33 i'll probably get 3 more efikas soon, if you still cant' find sth, let me know, i could send you one of mine Apr 30 14:00:35 i found them gathering dust somewhere and i'll get them cheap Apr 30 14:00:47 markos_: cool, I am interested Apr 30 14:01:04 ok, i'll let you know when i get them :) Apr 30 14:01:31 thx Apr 30 14:01:38 np Apr 30 14:02:40 OE runs great on it btw and right out of the box :) Apr 30 14:02:48 I know Apr 30 14:03:12 few OE devs got efikas quite early Apr 30 14:03:59 cool, that explains it Apr 30 14:04:21 markos_: efika still use OF? not u-boot? Apr 30 14:04:26 yes Apr 30 14:14:42 hrw: thanks for the openrd link (earlier)! Apr 30 14:15:40 Ainulindale: http://www.nuhorizons.com/development/marvell/kirkwood_open_rd_base.asp OR http://www.micetek.com/englishweb/product/new/PQIIproDevelopmentProducts/8313E-RDB.htm Apr 30 14:16:59 markos_: let me know if you have an extra efika left, I'ld like to buy one, I blew mine during building a Wifi Car Radio from it. Apr 30 14:17:19 Ainulindale: those two platforms are a good fit to your question. Apr 30 14:18:03 likewise: ok, i'll be around anyway, when i get them i'll say it here Apr 30 14:18:43 Also, that reminds me: I have one extra SheevaPlug left, let's see who responded first on the mailing list. Apr 30 14:22:01 OK, different take. How can I pass an environment variable to "make" when inheriting autotools Apr 30 14:22:10 where do I have to export it? Apr 30 14:25:12 spaetz: openembedded calls make with the -e flag Apr 30 14:26:05 spaetz: see how the current variables (such as LDFLAGS) override the in-Makefile variables. Apr 30 14:28:07 ok, thanks Apr 30 14:28:12 will have a look Apr 30 14:29:04 likewise: no problem Apr 30 14:29:46 what is the least size i can get an image Apr 30 14:30:03 mib_l8vbfa: that rather depends on what you want to have in it Apr 30 14:30:12 0 bytes is achievable :) Apr 30 14:32:59 likewise: well I might be interested with that too Apr 30 14:33:03 s/with/in/ Apr 30 14:34:06 i dont want zero bytes Apr 30 14:34:56 i think i have udev, but for some devices nodes are not getting created in /dev/ Apr 30 14:35:09 is there a way of specify the nodes in advance Apr 30 14:36:17 http://globalscaletechnologies.com/t-openrdbdetails.aspx Apr 30 14:41:37 likewise: still got that sheeva :-) Apr 30 14:42:52 mib_l8vbfa: helloworld-image is rather small, it is statically linked hello world Apr 30 14:52:50 okay, i am using udev and what i see is the node for one of the driver does not get created Apr 30 14:53:56 which driver? Apr 30 14:56:58 morning Apr 30 14:57:19 XorA: still got? Apr 30 14:57:42 Also, that reminds me: I have one extra SheevaPlug left, let's see who responded first on the mailing list. Apr 30 14:58:54 XorA: yes, but you mean you already have a sheeva? Apr 30 14:59:01 I dont have one Apr 30 14:59:04 * likewise is a bit low on coffee today Apr 30 14:59:44 g'day kergoth Apr 30 15:01:16 * kergoth 's sheevaplug arrived yesterday, now he just needs to find the motivation to actually do something with it Apr 30 15:01:45 kergoth: it's not really hackable, all the nice i/o end in the ball grid... Apr 30 15:02:08 sata not brought out, neither is pcie, nor, spdif, neither is 2nd GbE Apr 30 15:02:13 * kergoth 's no hardware guy, happily Apr 30 15:02:20 ah :-) Apr 30 15:02:40 'embedded' counts too Apr 30 15:03:14 * mwester has a boatload of Openmoko devices for I/O hacking; he's happy with USB and SDIO for expansion on the Sheevaplug Apr 30 15:06:40 * likewise has EE background, so wants to move significant electrons around in large numbers Apr 30 15:12:18 i just wanted it so i have actual hardware to test OpenEmbedded built distros on, as well as to test it out as a box to run my torrents 24/7 so i can shut down my windows box, save power Apr 30 15:13:05 03Andrea Adami  07org.openembedded.dev * rf13ab24657 10openembedded.git/recipes/nandlogical/nandlogical-static_1.0.0.bb: Apr 30 15:13:05 nandlogical-static: actually build static... Apr 30 15:13:05 - readd -static Apr 30 15:13:05 - size is a bit big...433k Apr 30 15:13:25 guys: is there a range of MAC addresses to use for private use? such as 10.x.x.x in IP? Apr 30 15:13:50 hrw: hmm, good question, not that I know Apr 30 15:14:07 hrw: I think its any odd MAC address or something like that Apr 30 15:14:27 wikiepedia has the info Apr 30 15:14:30 hrw: http://standards.ieee.org/faqs/OUI.html Apr 30 15:14:34 but I havent looked in a while Apr 30 15:15:04 hrw: yes, there is a range for locally-assigned addresses. Apr 30 15:17:19 BUG has ethernet in base (and I have proper cable now) but no eeprom for mac Apr 30 15:17:37 and I wonder which to set (other then 001122334455 or such one) Apr 30 15:18:00 oxDEADBEAF was recognised as valid when I tried that on another board :-) Apr 30 15:18:11 and smc911 driver ignore bootloader mac ;( Apr 30 15:19:35 if (mac addr & 02:00:00:00:00:00) then is_locally_assigned_addr Apr 30 15:19:35 hrw: Three-octet values occupying the same fields as OUIs can occupy, but with the next-to-LSB of the first octet set to 1, are locally assigned and have no relationship to the IEEE-assigned values... Apr 30 15:19:54 lol, thanks mwester Apr 30 15:20:10 can i build ibgstpulse.so Apr 30 15:20:10 * mwester just went through this with the Openmoko g_ether gadget addresses. Apr 30 15:20:28 i built gstreamer-good, bad and ugly and it wont give me libgstpulse.so Apr 30 15:28:29 which rcS script gives me Angstrom logo in serial port Apr 30 15:29:57 /etc/issue? Apr 30 15:30:47 i want to insmod some modules which is the best place to insert those in /etc? Apr 30 15:33:10 mib_l8vbfa: try write names into /etc/modules but I'm unsure Apr 30 15:33:31 who calls /etc/issue Apr 30 15:33:38 i could not find any reference to that Apr 30 15:34:15 it should be called by getty afaik Apr 30 15:34:17 i want to run a script at the end of everything in /etc Apr 30 15:34:30 i dont know which is the best place to do that Apr 30 15:35:13 place script into /etc/init.d and create symlink in proper /etc/rcX.d Apr 30 15:35:22 look there for examples Apr 30 15:35:40 links are processed in alphabetical order Apr 30 15:37:41 leon@witty:~/sandbox/sheevaplug/openembedded/tmp/rootfs$ ../cross/armv5te/bin/arm-angstrom-linux-gnueabi-objdump -x sbin/sysctl.procps | grep NEEDED Apr 30 15:37:43 NEEDED libproc-3.2.7.so Apr 30 15:37:51 but libproc was not installed in the rootfs... Apr 30 15:46:51 how can i disable the cursor in my lcd, i dont want angstrom and login prompt to be disabled in my lcd Apr 30 15:49:26 Package: procps Depends: update-rc.d, libc6 (>= 2.6.1), ncurses (>= 5.4), libgcc1 (>= 4.2.4) Apr 30 15:51:40 where are the (library) dependencies resolved? Apr 30 15:52:04 mib_l8vbfa: you want to disable the cursor, but keep the login prompt? you'd have to patch the startup scripts to send the appropriate cursor-off sequence to the console. Apr 30 15:56:06 Hi guys, I'm having problems building ti-codec-engine. Is there anybody that can and wants to help? Apr 30 16:00:56 pb_: i dont want to display anything on the lcd including the angstrom logo, i get them on my serial port Apr 30 16:01:14 if i dump something on lcd, i get the cursor blinking Apr 30 16:02:46 Any git people around? Apr 30 16:03:09 Is there any way to amend a commit a few commits back, that you haven't pushed, without having to do all the ones that followed? Apr 30 16:03:45 the ones that follow have to change too, because the older commit changes, and all commits contain references to the parent commit. but making that change is transparent, and you dont have to worry about it yourself Apr 30 16:03:53 just do git rebase -i HEAD~5 (for 5 commits back) Apr 30 16:04:00 mark the one you want to change as edit in the text editor Apr 30 16:04:02 rather than pick Apr 30 16:04:12 itll drop you into a shell where youc an do a ci --amend Apr 30 16:04:19 thanks Apr 30 16:04:22 then git rebase --continue and itll re-apply the other commtis on top of it Apr 30 16:04:24 np Apr 30 16:05:02 mib_l8vbfa: disable getty on tty0-x Apr 30 16:05:07 via /etc/inittab Apr 30 16:06:32 jay7: i dont see that in my inittab Apr 30 16:06:36 i use console-image Apr 30 16:12:40 Jay7: my inittab, please let me know how to disable it.http://pastebin.ca/1408572 Apr 30 16:13:23 mib_l8vbfa: comment out last line with tty1 Apr 30 16:17:54 03Tom Rini  07org.openembedded.dev * r1f7055b6bb 10openembedded.git/recipes/gcc/ (6 files in 2 dirs): gcc 4.0.0: Fix fortran building, disable libmudflap/ssp, add gcc-posix-open-fix.patch, add missing cross-initial/cross-intermediate files, really honor LDFLAGS, bump PR Apr 30 16:17:54 03Tom Rini  07org.openembedded.dev * r73a5d9a06f 10openembedded.git/recipes/gcc/ (41 files in 10 dirs): gcc (various): Drop incomplete LDFLAGS patch and really honor them, bump PR Apr 30 16:17:55 03Tom Rini  07org.openembedded.dev * racfd59a939 10openembedded.git/recipes/gcc/ (gcc-4.1.0.inc gcc-cross_4.1.0.bb gcc_4.1.0.bb): gcc-4.1.0: Use patches from 4.1.1, really honor LDFLAGS, bump PR Apr 30 16:17:56 03Tom Rini  07org.openembedded.dev * r3628be8901 10openembedded.git/recipes/gcc/ (4 files in 3 dirs): gcc-3.4.x: Rename gcc-posix.patch to gcc-posix-open-fix.patch, put in files/ Apr 30 16:17:59 03Tom Rini  07org.openembedded.dev * r8592e94690 10openembedded.git/recipes/gcc/ (6 files in 2 dirs): gcc 4.0.2: Disable libmudflap/ssp, add gcc-posix-open-fix.patch, really honor LDFLAGS, bump PR Apr 30 16:18:02 03Tom Rini  07org.openembedded.dev * ra087626bd7 10openembedded.git/recipes/gcc/ (18 files in 4 dirs): Apr 30 16:18:05 gcc-4.3.x: Really honor LDFLAGS, bump PR. Apr 30 16:18:07 Rework SYSROOT_CFLAGS_FOR_TARGET.patch each time, and rename to -4.3.2/4.3.3. Apr 30 16:24:48 florian: ping Apr 30 16:26:41 likewise: pong Apr 30 16:40:29 florian: congr's with new hosting :) Apr 30 16:41:16 Jay7: thanks... we still need to get it integrated :) Apr 30 16:41:31 hehe.. Apr 30 16:41:43 Jay7: but after moving the databases and some minor changes the main server works much better Apr 30 16:42:20 yeah, it's visible :) Apr 30 16:46:46 florian: we already are hosted elsewhere?? Apr 30 16:48:36 like|run: no... ltg has a vserver Apr 30 16:48:39 bbl Apr 30 16:58:19 for some reason I can't a variety if internet addresses from home Apr 30 16:58:25 including my mail server Apr 30 16:58:34 but I can ping from other locations, weird Apr 30 16:58:47 Hmmm Apr 30 16:59:39 okay what can cause: failed sanitiy test (workdir) in path [...]staging/i686-linux/usr/lib Apr 30 16:59:40 ? Apr 30 16:59:47 * kergoth 's never seen this particular qa failure Apr 30 17:00:48 pkg-config? Apr 30 17:01:02 its the .la files failing the sanity test Apr 30 17:01:10 not sure why workdir would end up there Apr 30 17:01:11 OK, here's a hard question Apr 30 17:01:44 Why do we want a compiled for host libiberty.a in the target staging dir? Apr 30 17:01:48 https://www.leopardboard.org/ Apr 30 17:02:20 Crofton|work: impressive video Apr 30 17:08:02 hmmm Apr 30 17:15:41 Crofton|work: nslookup to see if it's a DNS problem Apr 30 17:16:01 done all that :) Apr 30 17:16:21 Crofton|work: change isp Apr 30 17:16:29 heh Apr 30 17:16:31 I wish Apr 30 17:31:03 hi, is there already a oe build for sheevaplug? Apr 30 17:31:25 and if so, is it possible to build for two targets using the same tree (of course with different tmp dirs Apr 30 17:31:44 eFfeM, yes Apr 30 17:32:08 I'm not sure if there is a build, but multimachine takes care of different machine in same tmp dir Apr 30 17:32:49 how do i do that ? is there a apge on that? Apr 30 17:32:52 page Apr 30 17:34:47 eFfeM: MACHINE=mach1 bitbake something;MACHINE=mach2 bitbake something Apr 30 17:34:52 thats whole magic Apr 30 17:35:19 or even: "for mach in mach1 mach2;do MACHINE=$mach bitbake something;done" which I usually use Apr 30 17:35:31 and sheevaplug is supported Apr 30 17:35:42 hrw, ok thanks, so no change needed in local.conf or so Apr 30 17:35:43 thx to chouimat and koen Apr 30 17:35:58 eFfeM: unless you have MACHINE="machine1" in conf instead of ?= Apr 30 17:36:02 yeah i knew koen had one and was working on it Apr 30 17:36:24 hrw, i probably have MACHINE=beagleboard Apr 30 17:36:30 thanks for the warning Apr 30 17:36:50 eFfeM: most of my OE multibuilds do not have MACHINE set in config at all Apr 30 17:36:59 i'll just pull MACHINE from the env Apr 30 17:37:23 ah.. one thing Apr 30 17:37:24 hrw, I did whatever was in the elinux beagleboard page on oe as I was a git n00b Apr 30 17:37:35 export BB_ENV_EXTRAWHITE="MACHINE" is needed Apr 30 17:37:45 ah ok, thanks Apr 30 17:37:49 eFfeM: but noe OE n00b iirc Apr 30 17:37:57 fscking e17 Apr 30 17:37:58 03Stanislav Brabec  07xora/angstrom-srcpv * r5a23eb5eb0 10openembedded.git/recipes/zaurus-updater/encdec-updater.bb: encdec-updater: GNU_HASH QA fix. Apr 30 17:38:00 03Andrea Adami  07xora/angstrom-srcpv * rf13ab24657 10openembedded.git/recipes/nandlogical/nandlogical-static_1.0.0.bb: Apr 30 17:38:00 nandlogical-static: actually build static... Apr 30 17:38:00 - readd -static Apr 30 17:38:00 - size is a bit big...433k Apr 30 17:38:03 03Tom Rini  07xora/angstrom-srcpv * r3628be8901 10openembedded.git/recipes/gcc/ (4 files in 3 dirs): gcc-3.4.x: Rename gcc-posix.patch to gcc-posix-open-fix.patch, put in files/ Apr 30 17:38:06 03Tom Rini  07xora/angstrom-srcpv * r73a5d9a06f 10openembedded.git/recipes/gcc/ (41 files in 10 dirs): gcc (various): Drop incomplete LDFLAGS patch and really honor them, bump PR Apr 30 17:38:09 03Tom Rini  07xora/angstrom-srcpv * racfd59a939 10openembedded.git/recipes/gcc/ (gcc-4.1.0.inc gcc-cross_4.1.0.bb gcc_4.1.0.bb): gcc-4.1.0: Use patches from 4.1.1, really honor LDFLAGS, bump PR Apr 30 17:38:14 03Tom Rini  07xora/angstrom-srcpv * r8592e94690 10openembedded.git/recipes/gcc/ (6 files in 2 dirs): gcc 4.0.2: Disable libmudflap/ssp, add gcc-posix-open-fix.patch, really honor LDFLAGS, bump PR Apr 30 17:38:17 03Tom Rini  07xora/angstrom-srcpv * ra087626bd7 10openembedded.git/recipes/gcc/ (18 files in 4 dirs): Apr 30 17:38:20 gcc-4.3.x: Really honor LDFLAGS, bump PR. Apr 30 17:38:22 hrw, indeed no OE n00b but was away for a while, so my knowledge was still iwth mtn, not git Apr 30 17:38:22 Rework SYSROOT_CFLAGS_FOR_TARGET.patch each time, and rename to -4.3.2/4.3.3. Apr 30 17:38:26 03Graeme Gregory  07xora/angstrom-srcpv * r786c6fb3c1 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git@git.openembedded.net/openembedded into xora/angstrom-srcpv Apr 30 17:38:29 03Tom Rini  07xora/angstrom-srcpv * r1f7055b6bb 10openembedded.git/recipes/gcc/ (6 files in 2 dirs): gcc 4.0.0: Fix fortran building, disable libmudflap/ssp, add gcc-posix-open-fix.patch, add missing cross-initial/cross-intermediate files, really honor LDFLAGS, bump PR Apr 30 17:38:43 libecore-fb - 2:0.9.9.050+svnr40125-r2.1 - Ecore is the Enlightenment application framework library Apr 30 17:38:46 libecore-fb-ver-pre-svn-00-0 - 2:0.9.9.050+svnr40247-r4.1 - Ecore is the Enlightenment application framework library Apr 30 17:38:54 * Crofton|work senses he should go rebuild the world ..... Apr 30 17:40:02 Crofton|work: have fun :-) that'll take a while Apr 30 17:40:17 no.. Apr 30 17:40:25 just few minutes will tak Apr 30 17:40:26 e Apr 30 17:40:38 OE has too many missing deps here and there so world fails Apr 30 17:40:53 -k world is your friend :) Apr 30 17:41:08 kergoth: last time -k was not helpful Apr 30 17:41:10 Crofton|work: i was half way through a world build a few weeks ago. it took 48 hours to get that far Apr 30 17:41:14 hrw: works fine for me Apr 30 17:41:25 bitbake spent most of time in resolving deps here Apr 30 17:41:25 it takes a LONG time to get past all the dep failures though when it calcs the runqueue Apr 30 17:41:36 does work though Apr 30 17:43:31 well, ok Apr 30 17:43:36 I meant console-image Apr 30 17:43:41 hehe Apr 30 18:12:59 have a nice weekend Apr 30 18:14:14 * Tartarus begs for more review and acks for his LIBC introduction Apr 30 18:27:39 humm i have appended a distro-specific configfile to alsa-lib and it does install properly when doing ipkg install but when i build an image, the configfile is not being copied into the /etc directory of the rootfs created :/ Apr 30 18:30:55 what could i be missing? Apr 30 18:31:26 http://pastebin.com/d356dfd54 it's about the asound.conf Apr 30 18:32:06 oh wait a minute... could it be that it has to go into do_stage_append_opendreambox instead of do_install_... ? Apr 30 18:32:47 nope Apr 30 18:32:52 Are you sure the right one is being packaged? Apr 30 18:32:57 do_rootfs just installs the package Apr 30 18:33:01 not entirely Apr 30 18:33:09 heh Apr 30 18:33:11 well i cclean'ed the package Apr 30 18:33:17 You can do -c clean then -c unpack Apr 30 18:33:24 Then look in the work dir and see if you have the right file in there Apr 30 18:33:35 If not, try bitbake -e alsa-whatever | less Apr 30 18:33:41 See if OVERRIDES has your distro in it Apr 30 18:34:18 it must have it since the ipk built does install the conf file when i directly install it on the box using ipkg install -force_reinstall Apr 30 18:34:36 rebaking right now... Apr 30 18:36:27 So, OK. From a 'normal' build what are the things in 'out' we care about, explicitly? Apr 30 18:36:28 ga Apr 30 18:36:46 pardon? Apr 30 18:38:47 wrong channel Apr 30 18:39:05 okay well i have the asound.conf in /dream/dm7025/build/tmp/work/alsa-lib-1.0.11-r1/./image/etc/asound.conf as expected Apr 30 18:39:12 OK Apr 30 18:39:34 Then the wrong ipk is being used in do_rootfs Apr 30 18:39:45 find tmp/deploy/ipk -name alsa-whatever* | xargs rm Apr 30 18:39:45 i didn't unpack anything though, that's just what came out of the bake process that i just did Apr 30 18:39:49 then build it again Apr 30 18:39:53 (it==package) Apr 30 18:48:19 I propose we rewrite OE in verilog! Apr 30 18:48:29 rofl ^^ Apr 30 18:48:46 vhdl forever Apr 30 18:48:52 rofl Apr 30 18:50:27 i still wonder how i manager to get an a in that class at uni, which doesn't mean that you must ever consult me though ^^ i have no idea about it Apr 30 18:50:55 Tartarus: still no asound.conf -_- Apr 30 18:51:03 mickeyl: good evening Apr 30 18:51:15 good evening pb__ Apr 30 18:53:52 http://pastebin.com/d56a971dc Apr 30 18:54:04 how to see exactly which patch commands are being run in which dir? Apr 30 18:55:25 03Stanislav Brabec  07org.openembedded.dev * rca74043bac 10openembedded.git/recipes/mtkbabel/mtkbabel_0.8.bb: mtkbabel: Allow to set faster logging. Apr 30 18:59:42 i'll also delete those couple other ipks created and retry Apr 30 19:01:48 Crofton|work: one advantage is we can parallellize the build :-) Apr 30 19:02:00 Crofton|work: in *extreme* ways Apr 30 19:02:04 :) Apr 30 19:02:14 and go seriously insane ... Apr 30 19:02:38 I have been doing some GPU programming lately: *one* thread per pixel on 1920x1080 pixels Apr 30 19:02:57 o.ó Apr 30 19:03:12 I didn't believe it until I learned how GPU's work (or at least, learned a bit) Apr 30 19:03:56 NOTE: base_set_filespath usage is deprecated, initscripts-slugos-1.0 should be fixed Apr 30 19:04:04 Tartarus: didn't help either! bothersome! Apr 30 19:04:14 Anyone care to offer a pointer on how/what I should replace the base_set_filespath with? Apr 30 19:04:15 I know guys doing signal processing on gou's Apr 30 19:04:31 mwester, I think kergoth may know Apr 30 19:04:46 I think he has been working in that area Apr 30 19:05:42 Perhaps he will pop in and comment.... a pointer to a recipe that has been "corrected" would be helpful, and I'll correct the one it's currently whining about. Apr 30 19:06:01 yeah Apr 30 19:06:20 we have some pretty good patterns for the GNU Hash stuff now Apr 30 19:08:06 mwester: FILESPATH is now constructed from FILESPATHBASE, FILESPATHPKG, and OVERRIDES Apr 30 19:08:20 for each path in base, + for each path in pkg + overrides = each entry in the filespath Apr 30 19:08:46 if you look at the default FILESPATHPKG< thats what contains your ${BPN}, ${PN}, ${PN}-${PV} Apr 30 19:08:57 so its likely all you need to do is set/modify FILESPATHPKG from the recipe Apr 30 19:09:22 I'm open to a better way of doing it, if you can think of one, but this seemed a lot cleaner than having to call base_set_filespath, or hardcode a single location with FILESDIR Apr 30 19:10:01 Ok. I have a package named silly.bb -- it includes realpackage.bb, and wishes to ensure that it can use the files from files/ as well as silly/ .... Apr 30 19:10:01 FILESPATHPKG =. "mypackage" in mypackage-foo.bb would ensure it picks up mypackage's files Apr 30 19:10:06 err, =. "mypackage:" that is Apr 30 19:10:08 colon separated Apr 30 19:10:16 Ok, I'll give it a shot. Apr 30 19:10:17 FILESPATHPKG =. "realpackage:" Apr 30 19:10:39 then itd check realpackage *before* silly. keep that in mind. prepend itll check it first, append itll check it after all the others, including "files" Apr 30 19:10:43 hello Apr 30 19:10:54 so if you want it between files and silly, you'd have to just override FILEPATHPKG directly rather than appending/prepending Apr 30 19:10:55 In this case, *before* is actually correct. Apr 30 19:11:00 * kergoth nods Apr 30 19:11:07 just something to keep in mind Apr 30 19:11:13 it uses the first one it finds when searching FILESPATH Apr 30 19:11:23 Actaully, I should subsume silly.bb into the real one, but that might not be so easy just now... Apr 30 19:11:38 thanks Apr 30 19:11:54 can you think of a better way of dealing with the filespath stuff? like i said, open to alternatives Apr 30 19:12:10 grr Apr 30 19:18:46 03ghost  07org.openembedded.dreambox * r8dd4e5d440 10openembedded.git/packages/dreambox/dreambox-dvb-modules.bb: (log message trimmed) Apr 30 19:18:46 dreambox-dvb-modules.bb: update dm8000 and dm800 driver modules Apr 30 19:18:46 changes in both drivers: Apr 30 19:18:46 - working on dvb-t tuner initialisation problems Apr 30 19:18:48 - demux "feed_cb but filter not GO" bug fixed (i.e. this fixes frequently crashes on cyfrowy polsat channels) Apr 30 19:18:50 additional changes in dm800 drivers: Apr 30 19:18:52 - allow playback of radio services with invalid pcr data or non pcr data in pcr pid Apr 30 19:19:06 03Roman I Khimov  07org.openembedded.dev * ra02e11ade1 10openembedded.git/ (7 files in 3 dirs): Apr 30 19:19:06 coreutils: add version 7.2 Apr 30 19:19:07 Acked-by: Otavio Salvador Apr 30 19:19:09 03Roman I Khimov  07org.openembedded.dev * r80671af8df 10openembedded.git/recipes/coreutils/ (4 files): Apr 30 19:19:12 coreutils-native: remove autotools and quilt dependencies Apr 30 19:19:14 Allows to build coreutils-native at very early stage in build process. Apr 30 19:19:18 Acked-by: Otavio Salvador Apr 30 19:19:20 03Roman I Khimov  07org.openembedded.dev * rd5631dc054 10openembedded.git/ (12 files in 10 dirs): Apr 30 19:19:21 hi, does somone knows how to fix GNU-hash on packages built with bjam? or should I look into it? Apr 30 19:19:23 remove install-native package Apr 30 19:19:25 Sane 'install' program is provided by coreutils-native now. Apr 30 19:19:27 Acked-by: Otavio Salvador Apr 30 19:19:29 03Roman I Khimov  07org.openembedded.dev * rdeb50c8b70 10openembedded.git/classes/base.bbclass: Apr 30 19:19:29 (I've tried a trick but it didn't work) Apr 30 19:19:34 coreutils-native: make mandatory dep for all packages Apr 30 19:19:36 Gives us guaranteed sane GNU coreutils on any host. Should greatly ease Apr 30 19:19:40 support for various non-GNU host systems and allow to safely remove some other Apr 30 19:19:44 native packages like 'install-native'. Apr 30 19:19:50 Acked-by: Otavio Salvador Apr 30 19:20:26 03Stanislav Brabec  07org.openembedded.dev * r1503c8c955 10openembedded.git/recipes/mtkbabel/ (files/fast-logging.patch mtkbabel_0.8.bb): mtkbabel: Allow to set faster logging. Apr 30 19:20:31 03Stanislav Brabec  07org.openembedded.dev * rf6e145bbad 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 30 19:20:32 03Stanislav Brabec  07org.openembedded.dev * r260e1ccc21 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 30 19:21:19 03Roman I Khimov  07xora/angstrom-srcpv * ra02e11ade1 10openembedded.git/ (7 files in 3 dirs): Apr 30 19:21:24 coreutils: add version 7.2 Apr 30 19:21:26 Acked-by: Otavio Salvador Apr 30 19:21:28 03Roman I Khimov  07xora/angstrom-srcpv * rdeb50c8b70 10openembedded.git/classes/base.bbclass: Apr 30 19:21:29 Bah damnit Apr 30 19:21:31 coreutils-native: make mandatory dep for all packages Apr 30 19:21:33 Gives us guaranteed sane GNU coreutils on any host. Should greatly ease Apr 30 19:21:35 support for various non-GNU host systems and allow to safely remove some other Apr 30 19:21:35 Anyone seen this before on x8664: Apr 30 19:21:37 native packages like 'install-native'. Apr 30 19:21:39 Acked-by: Otavio Salvador Apr 30 19:21:41 03Roman I Khimov  07xora/angstrom-srcpv * r80671af8df 10openembedded.git/recipes/coreutils/ (4 files): Apr 30 19:21:44 coreutils-native: remove autotools and quilt dependencies Apr 30 19:21:48 Allows to build coreutils-native at very early stage in build process. Apr 30 19:21:50 Acked-by: Otavio Salvador Apr 30 19:21:54 03Roman I Khimov  07xora/angstrom-srcpv * rd5631dc054 10openembedded.git/ (12 files in 10 dirs): Apr 30 19:21:57 remove install-native package Apr 30 19:22:01 Crofton|work, hi do you know bjam well? Apr 30 19:22:02 Sane 'install' program is provided by coreutils-native now. Apr 30 19:22:03 NOTE: generating locale ve_ZA (UTF-8) Apr 30 19:22:03 mmap: No such device or address Apr 30 19:22:03 NOTE: Task failed: localedef returned an error (command was PATH="/ Apr 30 19:22:05 ... Apr 30 19:22:06 Acked-by: Otavio Salvador Apr 30 19:22:10 03Stanislav Brabec  07xora/angstrom-srcpv * rca74043bac 10openembedded.git/recipes/mtkbabel/mtkbabel_0.8.bb: mtkbabel: Allow to set faster logging. Apr 30 19:22:17 03Stanislav Brabec  07xora/angstrom-srcpv * r260e1ccc21 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 30 19:22:23 03Stanislav Brabec  07xora/angstrom-srcpv * r1503c8c955 10openembedded.git/recipes/mtkbabel/ (files/fast-logging.patch mtkbabel_0.8.bb): mtkbabel: Allow to set faster logging. Apr 30 19:22:30 03Graeme Gregory  07xora/angstrom-srcpv * r27fb2a3330 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git@git.openembedded.net/openembedded into xora/angstrom-srcpv Apr 30 19:22:31 qemu-arm blowing up Apr 30 19:22:37 03Stanislav Brabec  07xora/angstrom-srcpv * rf6e145bbad 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 30 19:22:52 * XorA does wish that not every branch was reported Apr 30 19:23:18 lol Apr 30 19:23:20 Gnutoo, all I know is I hate boost for using bjam Apr 30 19:23:28 yes... Apr 30 19:23:46 Crofton|work, did you build boost recently or fixed its GNU-hash? Apr 30 19:23:52 Trying to build omap3-sgx-modules; a patch is not getting applied correctly, but do_patch doesn't complain. If I extract source and apply the patch by hand, the source line looks ok. How can I get more info about how the patch is being applied? (bitbake -v didn't help) Apr 30 19:24:40 no Apr 30 19:24:43 ok Apr 30 19:24:51 Tartarus: meep? Apr 30 19:24:52 so I'll have to look into it Apr 30 19:24:55 thanks Apr 30 19:25:41 fraxinas, no idea, sorry Apr 30 19:26:01 what else could i try? Apr 30 19:27:14 my /dream/dm7025/build/tmp/deploy/ipk/alsa-lib_1.0.11-r1_mipsel.ipk definitely contains the /etc/asound.conf file inside the data.tar.gz Apr 30 19:27:29 And that's the only alsa-lib ipk? Apr 30 19:27:36 Check the log for do_rootfs and see if thats the one it grabs Apr 30 19:27:56 aye Apr 30 19:28:30 And you're sure you're trying your new rootfs image and not an old one? Apr 30 19:28:39 If the ipk contains what you want and the log says the right one is installed Apr 30 19:28:45 it's time to look for brown paper bag errors Apr 30 19:29:38 Gnutoo, sorry about that Apr 30 19:29:42 I wish I could be more help Apr 30 19:29:44 well! it only says stuff about extracting the control.tar.gz from that Apr 30 19:29:46 np Apr 30 19:30:04 * mwester loves it when he solves a bug by _removing_ significant portions of recipes. Apr 30 19:30:05 * Tartarus forces his buildbot to not do binary locales anymore Apr 30 19:30:39 Crofton|work, there is someone from shr that made a boost-38 recipe...could I modify it and could we include it(after review of course)? Apr 30 19:30:59 i'll actually cclean and redo the image again real quick Apr 30 19:31:14 remind me who shr is? Apr 30 19:31:21 but, it sounds like an idea Apr 30 19:32:21 SHR is a distro for the openmoko Apr 30 19:32:27 ah Apr 30 19:32:42 we have some shr-dev that have commit access in oe Apr 30 19:32:49 maybe aniudale and quickdev Apr 30 19:34:04 03Angus Ainslie  07fso/milestone5.5 * rc3b88a2970 10openembedded.git/recipes/webkit/webkit-gtk_svn.bb: webkit-gtk: change libcurl dependency to libsoup Apr 30 19:34:14 03Angus Ainslie  07fso/milestone5.5 * r0bde794d81 10openembedded.git/: Merge branch 'fso/milestone5.5' of git@git.openembedded.net:openembedded into fso/milestone5.5 Apr 30 19:35:02 ha!!! funny as hell xD Apr 30 19:35:37 the thing is Tartarus.... alsa-lib is not being installed at all! only alsa-conf and alsa-lib is just a build dep for that Apr 30 19:35:59 then i'll move my config file over Apr 30 19:38:44 03Angus Ainslie  07fso/milestone5.5 * r2374709cf5 10openembedded.git/recipes/images/fso-image.bb: fso-image : renamed to fso-zhone-image to match dev Apr 30 19:46:30 03Stanislav Brabec  07org.openembedded.dev * rc7ca3bf3ef 10openembedded.git/recipes/opkg/ (opkg-nogpg-nocurl_svn.bb opkg-nogpg_svn.bb opkg_svn.bb): opkg_svn.bb: Explicitly package datadir/opkg/intercept (not happens automatically for opkg-nogpg). Apr 30 19:49:26 Tartarus: regarding your mmap: I have "vm.mmap_min_addr = 0" in my /etc/sysctl.conf, maybe it's related Apr 30 19:50:01 Tartarus: It should be 0 for OE, I *think* because of locate generation using qemu. Apr 30 19:53:07 yaaay Tartarus: fixed Apr 30 19:53:11 OE won't start w/o that Apr 30 19:53:14 thx for the clues Apr 30 19:53:28 but that key is gone in 2.6.29, but i also think i've seen this on .28 where it's still around Apr 30 19:53:32 fraxinas, np :) Apr 30 19:53:57 likewise, but it's also a big CPU sink for something I don't need Apr 30 19:54:19 I'm (a) the SDK guy and (b) just doing build tests for this stuff w/ buildbot Apr 30 19:57:06 hi, is there a way to specify oe to include even -doc -dev ... packages for a recipe in the image? Apr 30 19:57:31 by default it includes only the "main" package Apr 30 19:57:59 Well Apr 30 19:58:08 You could make a task for the doc or dev stuff you want Apr 30 19:58:13 and add that in Apr 30 19:58:36 I'd recommend task rather than directly to make sure debian.bbclass rename stuff is handled right by ipkg Apr 30 19:59:09 claudiu_ita, or, look at the hooks that poky has for such things and add them to mainline OE Apr 30 19:59:27 if I add for example myhelloworld-doc in IMAGE_INSTALL it' ok? Apr 30 19:59:49 claudiu_ita, for the most part, yes. Apr 30 19:59:58 But for example, if your distro uses debian.bblcass Apr 30 20:00:21 And you want 'libfoo-dev' from 'foo.bb' you're best off making a task that wants foo-dev Apr 30 20:00:31 And installing your task in IMAGE_INSTALL Apr 30 20:02:25 I want to include dhcp in my image.. i've put in IMAGE_INSTALL "dhcp" but do_rootfs fails because it can't find the package "dhcp", infact I have not "dhcp" package between .ipk-s, I have dhcp-client_version.ipk, dhcp-server_version.ipk ... Apr 30 20:02:44 claudiu_ita, that's a different problem :) Apr 30 20:02:51 if in IMAGE_INSTALL i put "dhcp-client" it works? Apr 30 20:02:52 You want dhcp-client or dhcp-server, or both Apr 30 20:02:54 not 'dhcp' Apr 30 20:03:00 dhcp is just the recipe name Apr 30 20:03:00 both Apr 30 20:03:01 door Apr 30 20:03:25 how ca I tell him to put dhcp-client in the image? Apr 30 20:03:45 (and avoid do_rootfs error) Apr 30 20:05:27 claudiu_ita, change it to 'dhcp-client' and it should work Apr 30 20:05:47 i try right now..thanks Apr 30 20:07:56 so I just have to replace "dhcp" with "dhcp-client" in IMAGE_INSTALL ? Apr 30 20:08:21 yeah Apr 30 20:08:32 thank you Apr 30 20:10:19 ***found source of omap3-sgx-modules problem: do_compile was using sed to modify the file in question... so the patch worked, but then the file got trashed. Ug. Now fixed Apr 30 20:16:23 worked, thank you a lot Tartarus Apr 30 20:19:39 * XorA is always constantly surpised about what bad C code old gcc compilers accepted Apr 30 20:31:49 imho ~50% of all software will break when compiling with -Werror :) Apr 30 20:32:11 Jay7: sadly Apr 30 20:34:18 Jay7: could be worse, when I used to do WinNT work, the header files were full of errors, so your code could never be clean Apr 30 20:44:55 XorA: on positive its much better now Apr 30 20:45:16 XorA: infact GCC has gotten smarter than humans Apr 30 20:45:17 now Apr 30 20:45:19 khem: getting better :-) Apr 30 20:45:54 uclibc has changed to git too Apr 30 20:46:06 * khem looks to fix uclibc svn recipes Apr 30 20:46:39 khem ! Apr 30 20:46:58 Any idea why on earth we put a host library (libiberty.a) from binutils-cross into the target staging dir? Apr 30 20:48:15 Tartarus: depends usually libiberty is built in two flavors Apr 30 20:48:24 usually one is for host and another for target Apr 30 20:48:37 may be we use the wrong one Apr 30 20:48:43 intentionally Apr 30 20:48:48 check out do_stage in binutils-cross.inc Apr 30 20:48:55 I dont remember Apr 30 20:48:55 been like that since day 1, afaict Apr 30 20:49:00 * khem looks Apr 30 20:49:05 we move it out of the cross dir into the target dir Apr 30 20:49:22 and at least today it's an x86, or whatever your host is, library Apr 30 20:50:47 ooops Apr 30 20:51:39 Tartarus: it should be the target libiberty Apr 30 20:51:42 is it noy Apr 30 20:51:48 s/noy/not/ Apr 30 20:54:12 It is not Apr 30 20:54:22 go objdump -a one :) Apr 30 20:54:30 2.18 binutils at least Apr 30 20:55:23 * Tartarus hopes his buildbot will go faster with binary locale generation stuff ripped out Apr 30 20:56:35 Tartarus: I see its the host one Apr 30 20:59:40 03Graeme Gregory  07org.openembedded.dev * r0de958f8a3 10openembedded.git/recipes/icu/ (files/elif-to-else.patch icu-3.6.inc): icu/icu-3.6.inc : fix bad #elif that more modern compilers dont accept. Apr 30 21:00:06 Tartarus: it should be in STAGING_LIBDIR_NATIVE probably Apr 30 21:00:27 03Stanislav Brabec  07xora/angstrom-srcpv * rc7ca3bf3ef 10openembedded.git/recipes/opkg/ (opkg-nogpg-nocurl_svn.bb opkg-nogpg_svn.bb opkg_svn.bb): opkg_svn.bb: Explicitly package datadir/opkg/intercept (not happens automatically for opkg-nogpg). Apr 30 21:00:28 03Graeme Gregory  07xora/angstrom-srcpv * r41ae2d366b 10openembedded.git/recipes/icu/ (files/elif-to-else.patch icu-3.6.inc): icu/icu-3.6.inc : fix bad #elif that more modern compilers dont accept. Apr 30 21:00:28 03Graeme Gregory  07xora/angstrom-srcpv * r0de958f8a3 10openembedded.git/recipes/icu/ (files/elif-to-else.patch icu-3.6.inc): icu/icu-3.6.inc : fix bad #elif that more modern compilers dont accept. Apr 30 21:00:30 03Graeme Gregory  07xora/angstrom-srcpv * rd4b172489b 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git@git.openembedded.net/openembedded into xora/angstrom-srcpv Apr 30 21:01:03 or not at all? Apr 30 21:01:08 It's been broken for forever :) Apr 30 21:01:23 I think it should be Apr 30 21:01:31 because some cross tools depend on it Apr 30 21:01:39 and its better we use the one we provide Apr 30 21:01:54 So you're thinking some cross tools are building their own or using the host one today then? Apr 30 21:02:15 I remember someone reported issues compiling cross binutils on 64bit machines Apr 30 21:02:30 yeh Apr 30 21:03:04 k Apr 30 21:03:21 btw, I pushed the LDFLAGS fix for gcc this am Apr 30 21:03:38 figured it'd be fine since we've been partially doing that for a long time and upstream just does it now Apr 30 21:06:27 I hate this ldflags thing Apr 30 21:06:37 We should configure the toolchain correctly Apr 30 21:06:49 Well, I think we do it right now Apr 30 21:06:55 We're doing what gcc 4.4.0 does Apr 30 21:07:07 When we are using GNU_HASH by default whats the point of compiling toolchain with SYSV hash as defauly Apr 30 21:07:25 khem, ah. Yes, we should also take care of that Apr 30 21:07:27 it'd help a lot Apr 30 21:07:28 This should be implicit Apr 30 21:07:33 builds not obeying ldflags still need fixing, though. Apr 30 21:07:39 and this at least gives a way of spotting them Apr 30 21:07:40 I have patch here never got to finish it Apr 30 21:08:14 Well yes Apr 30 21:08:30 it clutters quite a lot Apr 30 21:10:32 * kergoth thinks we need to do a much better job of pushing our patches upstream, and working iwth upstreams in general Apr 30 21:15:17 Tartarus: I cant think of a reason why we would need this libiberty in wrong place Apr 30 21:15:45 cross-gcc builds both target and host versions Apr 30 21:15:53 Yeah Apr 30 21:16:03 I think it's an ancient goofup Apr 30 21:16:58 could be Apr 30 21:17:04 care to patch ? Apr 30 21:17:27 we may want to install it in correct place Apr 30 21:17:30 bit overloaded atm Apr 30 21:17:39 part of why i just pushed the gcc fix Apr 30 21:17:39 ok let me test it out Apr 30 21:17:42 thanks Apr 30 21:18:59 03Mike Westerhof  07org.openembedded.dev * r32810dfbe2 10openembedded.git/recipes/arm-kernel-shim/arm-kernel-shim_1.5.bb: Apr 30 21:18:59 SlugOS: arm-kernel-shim - ensure that the rootfs is mounted with "noatime" option Apr 30 21:18:59 by passing that option in on the command line. Apr 30 21:20:38 03Mike Westerhof  07org.openembedded.dev * r861dedd490 10openembedded.git/conf/distro/include/preferred-slugos-versions.inc: SlugOS: preferred-slugos-versions - get rid of overrides for openssh, fixed upstream. Apr 30 21:22:15 it might break some old binutils reciped Apr 30 21:24:46 03Mike Westerhof  07org.openembedded.dev * r9823dcf3aa 10openembedded.git/conf/distro/include/slugos.inc: Apr 30 21:24:46 SlugOS: slugos.inc - ensure that the rootfs is mounted with the noatime option Apr 30 21:24:46 (this is for documentation purposes, and to future-proof, since the real Apr 30 21:24:46 command line is set in the apex recipe.) Apr 30 21:29:43 ant__: hello Apr 30 21:29:58 hi Jay7 Apr 30 21:30:46 I have a general OE use type question... Apr 30 21:30:48 how do I tell oe to rebuild all kernel modules into the rootfs and image? Apr 30 21:31:52 which image are you using Apr 30 21:32:15 03Mike Westerhof  07org.openembedded.dev * r74efae1b1a 10openembedded.git/recipes/ (images/slugos-image.bb initscripts/initscripts-slugos_1.0.bb): Apr 30 21:32:15 SlugOS: slugos-image.bb, initscripts-slugos_1.0.bb - remove devfs support Apr 30 21:32:15 Remove devfs startup scripts, and revert to the standard OE means to Apr 30 21:32:15 initially populate the /dev directory. This has the side-effect of Apr 30 21:32:15 fixing the empty /dev dir problem when extracting the tar.gz image. Apr 30 21:32:39 well a custom image that wraps the gumstix-basic-image Apr 30 21:34:18 03Stanislav Brabec  07org.openembedded.dev * r87fbf88f7e 10openembedded.git/recipes/subversion/ (files/libtool.patch subversion_1.4.5.bb): subversion: Build with the latest libtool. Apr 30 21:36:59 03Mike Westerhof  07org.openembedded.dev * re40bd5bec5 10openembedded.git/recipes/ (5 files in 2 dirs): SlugOS: refactor slugos images and tasks to support a standard and extended image. Apr 30 22:13:26 any idea why we are using old tzdata? Apr 30 22:13:48 nobody has updated the recipe? Apr 30 22:13:56 :-) Apr 30 22:14:09 I emerged already 2 news version since Jan Apr 30 22:14:25 actual is 2009.something Apr 30 22:14:30 let see Apr 30 22:15:58 well, kergoth...help Apr 30 22:16:31 how is it so many FILES_ Apr 30 22:17:10 ? Apr 30 22:17:26 tzdata Apr 30 22:18:34 Because *somebody* (who shall remain nameless) made a HUGE fuss about only having a very few packages, claiming that almost every blasted timezone needed its own blasted package. So I complied when I wrote it. Apr 30 22:19:01 I've not touched tzdata once in its entire existance, so why are you asking me, exactly? Apr 30 22:19:26 I've read earlier you were talking about FILESDIR etc... Apr 30 22:19:29 If I were to do it again, I would do it differently -- because i don't get that persons emails anymore, so his complaint would not have been heard. Apr 30 22:19:41 perhaps you could suggest a better way Apr 30 22:19:47 what? Apr 30 22:19:52 FILESDIR/FILESPATH have NOTHING to do with FILES_* Apr 30 22:19:55 FILES_* is packaging Apr 30 22:20:02 FILESDIR/FILESPATH is file:// urls Apr 30 22:20:28 I lost a bit the actual state about * Apr 30 22:20:53 was banned then valid again iirc Apr 30 22:20:57 what? Apr 30 22:21:14 once again, FILES is for packaging, it has absolutely nothing to do with what i was talking about earlier Apr 30 22:21:46 kergoth: I ask *you* because you happen to know matrix in the details Apr 30 22:21:54 or you shows Apr 30 22:22:01 like you do :-) Apr 30 22:22:39 if you have a question, ask it in the channel and see who has an answer. directly addressing someone because they *might*, *possibly* have an answer isn't the way to go Apr 30 22:22:47 sorry Apr 30 22:22:59 if I've bothered you Apr 30 22:23:06 np Apr 30 22:23:14 I ask somebody else Apr 30 22:23:16 thx Apr 30 23:06:36 03Andrea Adami  07org.openembedded.dev * rebfefe4f9a 10openembedded.git/recipes/nandlogical/nandlogical-static_1.0.0.bb: nandlogical-static: remove bogus PR Apr 30 23:09:48 yay Apr 30 23:10:00 at91sam9263ek built meta-toolchain in dev Apr 30 23:10:05 on to canadian Apr 30 23:32:15 03Andrea Adami  07org.openembedded.dev * re34ffc1bf7 10openembedded.git/conf/machine/include/zaurus-2.6.inc: Apr 30 23:32:15 zaurus-2.6.inc: force creation of jffs2 images Apr 30 23:32:15 - tar.gz is (weak) assigned in bitbake.conf Apr 30 23:56:58 canadian is good too, yay May 01 00:17:40 03Rod Whitby  07org.openembedded.dev * re8410422c0 10openembedded.git/ (103 files in 18 dirs): Unslung: Removed all trace of the unslung distro from OE May 01 00:19:49 rwhitby, well, that's gonna conflict with my LIBC changes ;) May 01 00:21:14 Tartarus: should be an easy merge - in fact I would expect that support for glibc 2.2.5 can now be removed completely, but will need to RFC that separately. May 01 00:21:40 rwhitby, yeah, probably. Kinda sad to see it go really tho May 01 00:21:42 every line modified had 'unslung' in it, and can be removed from any version you have to resolve the merge. May 01 00:21:52 since I'm the crazy guy trying to build every gcc version we have a recipe for May 01 00:22:03 it lives on in http://svn.nslu2-linux.org/svnroot/unslung/trunk/ May 01 00:22:38 rwhitby, what host do you use to build on? May 01 00:23:10 Ubuntu 8.04.2, 32 bit. May 01 00:23:49 although I doubt that the nslu2-linux autobuilder will have rebuild glibc for unslung in a very long time May 01 00:26:30 Me too May 01 00:26:40 But I'll have to give my hardy chroot a shot next time May 01 00:35:31 Unslung hasn't built in OE for many many months. May 01 00:35:57 Tartarus: SlugOS is still there, though! May 01 00:50:25 heh May 01 00:50:46 What glibc are you on again? I fixed up celinux-test the other night which is 2.3.5+cvs May 01 00:50:50 03Rod Whitby  07org.openembedded.dev * rc82032fe18 10openembedded.git/contrib/oe-stylize.py: oe-stylize: Remove Unslung distro reference May 01 00:50:55 But ran into problems even in my chroot when linux32'd May 01 00:52:32 glibc-2.6.1 May 01 00:53:47 and a recent'ish gcc too i bet May 01 00:54:14 gcc older than 4.1.0 might as well be deleted, except for perhaps the gcc-native-3.4.6 or whatever it is we use on qemu May 01 00:54:18 gcc-cross-4.2.4 May 01 00:54:36 We're happy to not be the mostest-behindest distro in the batch. May 01 00:54:37 :) **** ENDING LOGGING AT Fri May 01 02:59:57 2009