| 00:35.01 | CIA-5 | 03kergoth 07 * r1.1338.1.1 10packages/linux/ (8 files in 8 dirs): Add a 2.6.5-omap1 kernel, with configs for various MACHINEs (helen2, perseus2, innovator 1510, innovator 1610, & generic omap1510 and omap1610). |
| 00:39.17 | CosmicPenguin | heh |
| 00:40.40 | CosmicPenguin | are we still suffering from the same X startup issues we had under OZ? |
| 00:40.54 | kergoth | dunno, ask pb :) |
| 00:41.22 | CosmicPenguin | man, I'm looking forward to the day when I start pushing freedesktop.org X on my co-worker |
| 00:41.39 | CosmicPenguin | He's so set in his ways that its going to make his head explode |
| 00:43.12 | kergoth | hehe |
| 00:46.56 | kergoth | bleh |
| 00:51.03 | CosmicPenguin | hmmm... ImUtil.h no longer lives in include/X11 |
| 00:52.51 | CosmicPenguin | heh 1.7 gigs of lovin' |
| 00:57.51 | CIA-5 | 03kergoth 07 * r1.1341 10packages/linux/ (3 files in 2 dirs): Add linux-2.6.5-gnalm1-gum0, a kernel for Gumstix devices. |
| 01:01.18 | treke|wherever | putty is making my eyes bleed |
| 01:08.09 | reenoo_ | kergoth: got one of those gumstix thingies? |
| 01:08.59 | reenoo_ | would be fun to play with I guess :) |
| 01:12.08 | reenoo_ | 'night all |
| 01:48.45 | kergoth | okay, i got sick of the current patcher sucking |
| 01:58.56 | kergoth | technically, i'm rewriting it in python, and adapting oe to just import and use the damn thing directly |
| 01:59.01 | kergoth | heh, oh well |
| 01:59.08 | kergoth | cpenguin: get anywhere with oe? |
| 01:59.33 | cpenguin | kergoth: yesish |
| 02:00.04 | cpenguin | It still needs some time |
| 02:00.37 | kergoth | anything i can help with? |
| 02:01.51 | cpenguin | If you get a chance to fire up the x86_64, we've got that libtool-native issue, and you can see if your ipkg segfaults too.. :) |
| 02:02.01 | kergoth | hehe, k |
| 02:03.07 | cpenguin | I started to bring in rpm & friends too |
| 02:03.44 | kergoth | man, i'm so behind on this driver.. dunno whats up with it.. probably shouldnt leave until its done, but i cant think |
| 02:03.48 | kergoth | oh well |
| 02:04.33 | cpenguin | Heh - I was supposed to be working on my kernel FB driver the last few days, but I was OEing instead |
| 02:04.44 | cpenguin | This is why I pad my estimates so much - so I can play and still be done on time |
| 02:06.19 | kergoth | heh, yeah, i'm still pretty new to doing this as a job, need to get better at that |
| 02:06.19 | cpenguin | My boss hasn't yet figured out if I'm bullshitting him or not |
| 02:06.19 | kergoth | hehe |
| 02:06.23 | kergoth | my boss told me to make sure i pad it well, but i didnt do so enough |
| 02:06.32 | kergoth | oops :) |
| 02:06.47 | kergoth | k, got a package_ipk build running on an opteron box |
| 02:06.53 | kergoth | task-bootstrap |
| 02:07.04 | cpenguin | good luck with all of that... :) |
| 02:07.32 | cpenguin | I think the pacgkage_ipk will work, its the install (image_ipk) that waas breakin |
| 02:07.41 | cpenguin | When it gets done, try to install base-files |
| 02:14.08 | kergoth | k |
| 02:14.54 | kergoth | cpenguin: libtool-native still fails |
| 02:15.04 | kergoth | i'll see if i can get that fixed before moving on to repro'ing the other issue |
| 02:16.04 | cpenguin | kergoth: indeed |
| 02:16.49 | kergoth | doesnt look like a fun one to debug |
| 02:16.51 | kergoth | bleh |
| 02:17.24 | cpenguin | Whats wierd is that it works manually |
| 02:17.34 | cpenguin | I have the same problem with rpm |
| 02:20.36 | cpenguin | fucking Greg Haerr |
| 02:20.43 | cpenguin | Cool - nxlib works for tck/tk with no prior testing! |
| 02:20.54 | cpenguin | Thats because I did support for tk/tcl before he canned my ass |
| 02:21.09 | kergoth | heh |
| 02:52.02 | *** join/#oe mithro (~tim@dsl1-83.gw1.adl1.airnet.com.au) |
| 02:53.49 | *** join/#oe jhonen (~jhonen@cdm-66-76-82-127.fayt.cox-internet.com) |
| 02:56.36 | *** part/#oe jhonen (~jhonen@cdm-66-76-82-127.fayt.cox-internet.com) |
| 03:29.44 | *** join/#oe SubNuke (~subnuke@nh-londndry-c41-bg1a-32-219.lndnnh.adelphia.net) |
| 04:17.26 | CIA-5 | 03kergoth 07 * r1.1342 10packages/elfutils/ (elfutils_0.89.oe elfutils-0.89/warnings.patch): Apply patch to elfutils to allow us to override -Wall -Werror. This is necessary due to new deprecated behavior warnings from gcc 3.4.0. |
| 04:25.43 | kergoth | cpenguin: well, i seem to have found a possible source of the problem with libtool |
| 04:25.45 | kergoth | cpenguin: confirming now |
| 04:26.13 | kergoth | if its what i think it is, its specific to machines that are 64 bit. |
| 04:29.38 | kergoth | hmmmm |
| 04:38.49 | cpenguin | That makes sense |
| 04:39.02 | kergoth | damnit |
| 04:39.18 | kergoth | cpenguin: it has this logic in its macros, it checks which binary format we're using, 32 or 64 |
| 04:39.27 | kergoth | then it adds an additional flag to the LD variable based on it |
| 04:39.38 | cpenguin | config? |
| 04:39.42 | kergoth | now, it seems that the shared lib check only fails when this option is appended to LD -more than once- |
| 04:39.50 | cpenguin | I saw that too, actually |
| 04:39.54 | cpenguin | the -m flag, right? |
| 04:39.55 | kergoth | e.g. ld -m elf_x86_64 -m elf_x86_64 -m elf_x86_64 |
| 04:40.20 | kergoth | hacking the macro to make it only append once, on the off chance its a quirk with the ld syntax, but i dunno |
| 04:40.23 | kergoth | its odd |
| 04:40.31 | cpenguin | Well, its 2 times as manhy bits, then 2 times as many flags, right... :) |
| 04:40.31 | *** join/#oe _stigger (~jeff@203.41.111.90) |
| 04:41.08 | kergoth | hehe |
| 04:43.22 | treke|wherever | it's 9:45 on Friday night. Do you know where your scotch is? |
| 04:43.31 | kergoth | i'm out! :( |
| 04:45.34 | kergoth | cpenguin: by default LD isnt in the MAKEFLAGS to be passed into submakes, so each time it does the check, it defaults back to 'ld' instead of picking up the previous LD modification |
| 04:45.43 | kergoth | cpenguin: thats why it only fails under OE. we set LD in the env |
| 04:46.24 | cpenguin | THats probably whats happening with rpm too |
| 04:46.38 | cpenguin | It fails because it looks at cpp, which is set to gcc instead of gcc -E |
| 04:46.49 | cpenguin | and thats in a subconfigure, which is the same concept |
| 04:46.49 | cpenguin | damn |
| 04:47.17 | kergoth | we should be setting it to gcc -E in oe.. |
| 04:48.23 | cpenguin | We are, but its getting modified somewhere along the line |
| 04:48.42 | kergoth | ahh, and because we're setting it, its being passed down into submakes when it normally wouldnt have been |
| 04:48.45 | kergoth | oi |
| 04:48.47 | cpenguin | Probably the same concept that we pass it in, so it gets modified in ways the subconfigures don't expect |
| 04:48.49 | kergoth | buildsystems making bad assumptions |
| 04:49.03 | cpenguin | this is automake we're talking about here |
| 04:49.05 | cpenguin | you are surprised? |
| 04:49.24 | kergoth | hehe |
| 04:50.23 | kergoth | hmm |
| 04:54.51 | kergoth | hehe |
| 04:54.52 | kergoth | i cheated |
| 04:54.55 | kergoth | hmmmm |
| 05:06.20 | kergoth | god this bug is ugly |
| 05:06.23 | kergoth | damn crap buildsystems |
| 05:06.40 | kergoth | i'm tempted to tell the toplevel configure not to call out the subconfigures, and fucking doing it myself. |
| 05:06.51 | cpenguin | I've done that before |
| 05:21.34 | kergoth | okay fuck this, i'll pick this back up again tomorrow |
| 05:21.35 | kergoth | blech |
| 05:21.36 | kergoth | ugly |
| 05:25.09 | cpenguin | heh - yeah, take a break |
| 05:28.51 | *** join/#oe _Bala_ (~Bala@hsdbsk142-165-117-156.sasknet.sk.ca) |
| 07:12.35 | *** join/#oe dirk_ (~dirk@do13.in-dsl.de) |
| 08:35.41 | *** join/#oe fido (~hl@z2421-034-174.Raab-Heim.Uni-Linz.AC.AT) |
| 08:43.43 | *** join/#oe pb_ (~pb@2002:5160:4550:1:240:95ff:fe30:aa84) |
| 08:44.37 | pb_ | morning |
| 09:01.24 | CIA-5 | 03mickeyl 07 * r1.1339.1.1 10packages/gcc/gcc-cross_3.4.0.oe: depend on libc-for-gcc, not glibc - patch courtesy pb_.\ |
| 09:01.34 | CIA-5 | 03mickeyl 07 * r1.1343 10packages/: |
| 09:01.34 | CIA-5 | Merge bk://openembedded@openembedded.bkbits.net/packages\ |
| 09:01.34 | CIA-5 | into gandalf.tm.informatik.uni-frankfurt.de:/usr/local/projects/packages\ |
| 09:01.56 | CIA-5 | 03mickeyl 07 * r1.1344 10packages/udev/udev_024.oe: fix typo - patch courtesy pb_.\ |
| 09:09.17 | mithro | pb: mickey's gateway? |
| 09:14.18 | pb_ | yup |
| 09:48.23 | *** join/#oe reenoo_ (~r@pD9528BE6.dip.t-dialin.net) |
| 09:48.46 | reenoo_ | morning |
| 09:48.52 | pb_ | hi reenoo_ |
| 09:49.25 | reenoo_ | hey pb_ |
| 09:50.20 | pb_ | brb |
| 10:07.52 | *** join/#oe mjc (mjc@sartre.ispvip.biz) |
| 10:13.49 | pb_ | re |
| 10:16.26 | *** join/#oe ar_oszine (~ar@port-ip-213-211-233-136.reverse.mdcc-fun.de) |
| 10:16.59 | mjc | hey pb_ |
| 10:20.12 | pb_ | hi mjc |
| 10:20.30 | mjc | what's up |
| 10:20.57 | *** join/#oe beiss (~beiss@srb9-d9bbf165.pool.mediaWays.net) |
| 10:21.21 | pb_ | just poking at some packaging code at the moment. |
| 10:23.31 | *** join/#oe __stigger (~jeff@203.41.111.147) |
| 10:28.12 | CIA-5 | 03mickeyl 07 * r1.541 10oe/classes/package.oeclass: add do_split_packages() to package.oeclass - patch courtesy pb_.\ |
| 10:29.54 | CIA-5 | 03mickeyl 07 * r1.542 10oe/classes/kernel.oeclass: abolish KERNEL_ARCH; use ARCH for packaging. use do_split_packages() from package.oeclass and delete local copy - patch courtesy pb_.\ |
| 10:47.49 | CIA-5 | 03mickeyl 07 * r1.1345 10packages/xtrans/xtrans_0.1.oe: add xtrans_0.1 tarball release - patch courtesy pb_.\ |
| 10:53.26 | beiss | hello together |
| 10:53.28 | mjc | hows life guys |
| 10:53.31 | pb_ | hi beiss |
| 10:54.01 | beiss | reenoo_: I heard you worked out kernel.oeclass ;-) |
| 10:55.52 | reenoo_ | I've done the modules splitting code, together with pb_ |
| 10:57.41 | stigger | hi de hi what does "./gentbl: ./gentbl: cannot execute binary file" when complieing a kernel |
| 10:58.26 | mjc | don't have hostcc/cc set up correctly |
| 10:59.05 | stigger | ok lost me completely now |
| 11:00.39 | beiss | reenoo_: I was thinking about separating the build and installation of the modules form building the kernel |
| 11:01.20 | beiss | reenoo_: in the moment, when i try to build a kernel without module support it fails. |
| 11:02.13 | beiss | reenoo_: perhaps it would be better to introduce something like modules_do_compile() and modules_do_install() |
| 11:02.23 | beiss | wht you think about that. |
| 11:03.47 | reenoo_ | probably better to fix "make modules" or whatever is breaking |
| 11:05.14 | beiss | reenoo_: make modules does it right, it says that thereis no modul support. so we even should not call make modules when the are not supported |
| 11:06.35 | pb_ | beiss: I don't really understand why you want to make that a new task, rather than just conditionalising the call in the existing do_compile(). |
| 11:06.43 | beiss | reenoo_: we can retriev this information from the .config-file or the user has to define a variable or to overwrite the according function. |
| 11:06.56 | reenoo_ | just grep .config for the appropriate options |
| 11:07.15 | reenoo_ | if they are not present, skip make modules etc. |
| 11:09.03 | beiss | pb_: I wanted to find out what would be the best solution, without breaking other functions related with the kernel and modules e.g. packaging! |
| 11:09.33 | reenoo_ | note that building a kernel without modules support will break other packages that try to build kernel modules... |
| 11:10.18 | reenoo_ | anyways |
| 11:10.44 | beiss | that's it, tha I wanted to be recognized |
| 11:13.46 | *** join/#oe darkschneider (~gab@81-208-36-80.fastres.net) |
| 11:18.30 | CIA-5 | 03mickeyl 07 * r1.1346 10packages/glibc/glibc_cvs.oe: eliminate local do_split in favour of version in package.oeclass - patch courtesy pb_.\ |
| 11:39.21 | mickeyl | treke: ping |
| 11:42.42 | pb_ | hi mickey |
| 11:42.49 | mickeyl | yo pb_ |
| 11:42.55 | mickeyl | up early |
| 11:43.09 | mickeyl | doing admin work in the office or not this weekend? |
| 11:44.58 | CIA-5 | 03mickey 07 * r1.1345.1.1 10packages/scummvm/scummvm_0.6.0.oe: pass -DSCUMM_NEED_ALIGNMENT |
| 11:47.51 | mickeyl | libtool-native causingmore problems? |
| 11:48.04 | mickeyl | didn't i say something like that before you convinced me? :D |
| 11:48.09 | mickeyl | should i revert? |
| 11:54.11 | pb_ | mickeyl: yeah, please do |
| 11:54.27 | pb_ | heh, I should have listened to you in the first place :-) |
| 11:55.05 | mickeyl | now you know |
| 11:55.06 | mickeyl | ;) |
| 11:55.53 | mickeyl | Chicken got bootldr up and running on SIMpad |
| 11:55.55 | *** join/#oe darkschneider (~gab@81-208-36-80.fastres.net) |
| 11:55.56 | mickeyl | this is _great_ news |
| 11:55.59 | pb_ | awesome! |
| 11:56.25 | darkschneider | can i have pasted it again :P |
| 12:00.42 | CIA-5 | 03mickey 07 * r1.1348 10packages/libtool/libtool-native_1.5.oe: revert removing --program-prefix from libtool-native |
| 12:04.07 | pb_ | mickeyl: could you rename some files for me? packages/linux/epia_2.6.5.oe needs renaming to packages/linux/linux-epia_2.6.5.oe, and likewise for the epia-2.6.5 directory. |
| 12:04.28 | pb_ | having PN be the same as MACHINE screws up the overrides, as it turns out. |