[00:41] hello [00:41] does any one know where there is some instructions on how to install ubuntu arm on n800? === bjf is now known as bjf-afk [08:31] hi [08:31] is asm("bkpt"); a good way to cause breakpoint in code? === ogra_ is now known as ogra === asac_ is now known as asac [10:59] o/ [10:59] cant believe that my auto joining works again ;) [10:59] even on +R channels [10:59] seems i sent "ident" to nickserv which it doesnt like anymore [11:05] well, dont reconnect in a way that you get an underscore on your nick [11:05] thats overly annoying [12:07] attached filtered debdiff for gvfs update (re bug 512959) [12:07] Launchpad bug 512959 in gvfs "nautilus assert failure: *** stack smashing detected ***: nautilus terminated" [High,Confirmed] https://launchpad.net/bugs/512959 [12:19] asac: Hi there; unfortunately I had an overrunning meeting, but I'm available now. [12:19] dmart: hi [12:19] cool [12:19] so i started to file bugs ;) [12:19] for main [12:19] added the numbers to the status col [12:19] https://wiki.ubuntu.com/ARM/Thumb2PackageReviewList [12:20] i am done to gmp [12:20] Looks good [12:20] * asac wonders if there is a shorthand syntax for lp bugs in the wiki [12:20] i didnt subscribe you to those bugs as you are member of ubuntu-armel i figured [12:22] asac, i think there is, but i cant remember the syntax, i know persia usually does :) [12:23] universe are so many packages :( [12:23] dunno... I shoved it though sed... refresh the page [12:23] yeah [12:23] anyway. getting the logs now [12:23] I was seeing a lot of false-positives, so it may be relatively quick [12:23] great [12:25] ok got the logs ... (thx wiki) ... now geting a new coffee and then we can lift off ;) [12:31] ok after not visiting that page lets get started ;) [12:32] cacao [12:32] needs a check (by doko) [12:33] Yes. I'm not sure why this appears separately from cacao-source; I guess it's basically the same issues [12:33] uses atomics and mov in various places [12:33] cdecl [12:33] false+ [12:34] cell-binutils [12:36] wonders what that is [12:36] guess needs to be checked [12:36] various hits ... not obvious what is used where and how [12:36] hmm. where is that? [12:36] i cant even find the source [12:36] !info cell-binutils [12:37] thats for PS3 [12:37] asac: Package cell-binutils does not exist in karmic [12:37] nothing for arm anyway [12:37] !source cell-binutils [12:37] asac: Error: I am only a bot, please don't think I'm intelligent :) [12:37] ogra: there are files name "arm" in the source at least [12:37] but ok [12:37] if you say its ps3 then its ok [12:37] well, its for the cell CPU, but i bet derived from nowmal binutils [12:38] i guess same is true for other cell-* [12:38] like cell-gcc cell-gdb etc. [12:38] likely [12:38] ogra: ? [12:38] "info cell-gcc [12:38] !info cell-gcc [12:38] asac: Package cell-gcc does not exist in karmic [12:38] heh [12:38] !source cell-gcc [12:38] Factoid 'source cell-gcc' not found [12:38] damn bot ;) [12:38] i think the bot only knows x86 anyway [12:38] !apt cell-gcc [12:38] Factoid 'apt cell-gcc' not found [12:38] !info uboot-imx [12:38] ogra: Package uboot-imx does not exist in karmic [12:38] see [12:38] !madison cell-gcc [12:38] asac: Error: I am only a bot, please don't think I'm intelligent :) [12:38] oh, karmic [12:39] !info redboot-imx [12:39] ogra: Package redboot-imx does not exist in karmic [12:39] !info uboot-imx lucid [12:39] right, there you go [12:39] asac: Package uboot-imx does not exist in lucid [12:39] isnt there a version= [12:39] in the binary? [12:39] i think it just knows binaries [12:39] not for the source package [12:39] ah [12:39] !info uboot-imx51-to3 [12:39] !info uboot-imx51-to3 lucid [12:39] asac: Package uboot-imx51-to3 does not exist in karmic [12:39] asac: Package uboot-imx51-to3 does not exist in lucid [12:39] ok [12:39] ubot4: please tell your owner to improve you. thx [12:39] asac: Error: I am only a bot, please don't think I'm intelligent :) [12:40] heh [12:40] ok marked those as PS3 [12:40] cpqarrayd [12:40] !info cpqarrayd [12:40] asac: cpqarrayd (source: cpqarrayd): monitoring tool for HP (Compaq) SmartArray controllers. In component universe, is extra. Version 2.3-1 (karmic), package size 17 kB, installed size 124 kB [12:40] aha [12:40] ;) [12:41] hmm. that one has parts of the kernel tree in the debian/ directory :( [12:41] cool ! [12:41] bah [12:41] we should that more often ! [12:42] is that normal? [12:42] definately not [12:42] ... sorry, distracted ... [12:42] for that specific package probably ... [12:42] no problem ;) ... we did some bot training ;) [12:42] it might need headers that the kernel doesnt ship by default or some such [12:43] thats why all the .S files are in there? [12:43] I'm pretty sure cell-{gcc,binutils,etc.} will be Cell arch specific tools packages; punt those to doko? [12:43] if he even cares [12:43] grep cpqarrayd * | pastebinit [12:43] http://pastebin.com/f1f093e3a [12:43] not sure the sell port is still alive [12:43] dmart: i marked those as PS3 only for now [12:43] *cell [12:43] OK [12:43] thats what ogra said [12:43] aka nothing to do [12:44] wow, thats ugly [12:44] ok saying that cpqarrayd needs to be checked ... has kernel tree in debian/ [12:44] !info ctypes [12:44] asac: Package ctypes does not exist in karmic [12:44] grep ctypes * | pastebinit [12:44] http://pastebin.com/f15603dcf [12:45] needs investigation i would say ... [12:45] uses call_reg macro so might be good for mov [12:46] dmart: the ldr parts are a false-+ for that? [12:47] see paste above [12:47] ok next ... [12:47] grep cxxtools * | pastebinit [12:47] http://pastebin.com/f281dc3f0 [12:48] seems to use atomics [12:48] ctypes - that looks like x87 assembler, so probaby not relevant. The mov pc, lr stuff seems to be in an embedded libffi, so can we link those two up? [12:49] dmart: so i will comment that hits are probably libffi dupes [12:49] ok. [12:49] ok dd is false-+ [12:49] ddd [12:49] cxxtools - I agree [12:50] ddd - agreed [12:50] deal.ii seems to be atomics again [12:50] deal.ii - agreed [12:50] desmume [12:50] hmm. disassembler [12:50] migt be right. [12:51] weird, guess we don't care about that [12:51] yes [12:51] dietlibc [12:51] desmume - Nintendo DS emulator, apparently [12:52] dietlibc - generally looks like it needs some porting (atomics, low-level asm) [12:52] ack [12:53] disktype [12:53] amiga ;) [12:53] false+ [12:53] yep [12:53] dosbox [12:53] surely false-+... checking... [12:53] disasm again [12:53] e3 [12:54] great package name ;) [12:54] grep e3- * | pastebinit [12:54] http://pastebin.com/feafed1 [12:54] what's this package? [12:54] checkig [12:54] dmart: some arm eabi support was added to dietlibc but I still didnt manage to get it to compile [12:54] !info r3 [12:54] asac: Package r3 does not exist in karmic [12:54] !info e3 [12:54] asac: e3 (source: e3): A very small editor. In component universe, is optional. Version 1:2.71-1 (karmic), package size 37 kB, installed size 120 kB (Only available for i386 kfreebsd-i386 amd64 kfreebsd-amd64) [12:55] hmm ... seems not relevant [12:55] (not supported arch) [12:55] ecl [12:55] Weird that there is ARM code in there... it looks like someone tried to support it... [12:56] dupe of libatomic-ops i guess [12:56] yeah [12:56] but given what the bot said, the debian maintainer has explicitly marked it for just that arch? [12:56] those archs [12:56] available for i386 kfreebsd-i386 amd64 kfreebsd-amd64) [12:57] okadded to comment that arm code exists [12:57] and someone might want to finish ;) [12:57] suihkulokki: ftbfs just for lucid, or generally? [12:57] asac: ok [12:57] so ok ... i said that ecl depends on libatomic-ops (which we have in main i think) [12:58] hmm ... and gmp ;) [12:58] Yes, those both look like dupes, can we link them? [12:59] yes [12:59] I expect we can dismiss the emacs packages (emacs23 in main was false-+) [12:59] well. i added that as comment [12:59] so i can add a task to the other bugs [12:59] yes [12:59] dmart: debian :) [12:59] (emacs) [13:00] suihkulokki: so we can judge dietlibc as not Ubuntu-ready then? (Ubuntu is EABI) [13:00] !info etherboot [13:00] asac: etherboot (source: etherboot): Bootstrapping for various network adapters. In component universe, is optional. Version 5.4.3+dfsg-0.2ubuntu2 (karmic), package size 20515 kB, installed size 30644 kB [13:00] dfsg/src/arch/armnommu/core/start.S: [13:00] dmart, yeah, dietlibc is dead for ARM. There are patches to make it EABI complient, but upstream was stalled last I checked. [13:01] NCommander, suihkulokki: I guess we don't have specific concerns for Thumb-2 right now then [13:01] etherboot - bootloader, so we can probably ignore this one [13:01] ok ... i add acomment [13:01] dmart: unless you want to work on it.. [13:01] (to dietlibc) [13:02] falconpl [13:02] !info falconpl [13:02] asac: falconpl (source: falconpl): The Falcon P. L. - command line tools. In component universe, is optional. Version 0.9.4-0ubuntu1 (karmic), package size 93 kB, installed size 252 kB [13:02] suihkulokki: I'll see whether anyone else on my side knows about it :) [13:02] false-+ [13:02] faumachine [13:02] !info faumachine [13:02] asac: faumachine (source: faumachine): Virtual machine running in user mode. In component universe, is optional. Version 20090512-1ubuntu1 (karmic), package size 1204 kB, installed size 4368 kB [13:02] false positive too [13:03] !info ffcall [13:03] asac: Package ffcall does not exist in karmic [13:03] http://pastebin.com/fe0b29f1 [13:03] faumachine - the match appears to be part of a comment, so OK [13:03] needs to be checked for mov's i guess [13:04] Yes; doesn't look T2-compatible right now [13:04] ffmpeg-php -> false-+ [13:04] fpc [13:04] mov and swp [13:04] !info fpc [13:04] dmart: fpc (source: fpc): Free Pascal Compiler - Meta Package. In component universe, is optional. Version 2.2.4-3 (karmic), package size 10 kB, installed size 40 kB (Only available for all i386 powerpc sparc amd64 arm) [13:05] Hmmm, I guess that needs looking at [13:05] yes. noted as mov and swp porting ) [13:05] !info freesci [13:05] dmart: freesci (source: freesci): a portable interpreter for SCI games like Space Quest 3. In component universe, is extra. Version 0.6.4-3 (karmic), package size 595 kB, installed size 1312 kB [13:05] onh [13:06] asac, fpc is going to be painful to port [13:06] * NCommander did some looking at it [13:06] freesci needs to be looked at [13:06] not sure if that applies [13:06] NCommander: looked into fpc 2.4 too ? [13:06] freesci - agreed [13:06] NCommander: right. its universe, so it comes last :) [13:06] fte -> false- [13:06] + [13:07] gamera [13:07] suihkulokki, nope [13:07] false-+, not arm asm [13:07] gamera - false-+ (x86 asm) [13:08] gauche-c-wrapper - embedded libffi? (same implications as ctypes) [13:08] yes [13:08] has macro ... might be safe [13:08] gcc-4.x -> doko (do we know which ones are actually in use to build stuff in universe?) [13:09] ok gcc stuff is already dealt elsewhere ... yes [13:09] gcj is next that doesnt have a comment [13:09] guess its doko too [13:09] gauche-c-wrapper - yes, looks like a simple review [13:09] gcj is gcc [13:09] gcl [13:09] so doko again [13:09] yes [13:09] marked gcj for doko [13:09] !gcl [13:09] Factoid 'gcl' not found [13:09] !info gcl [13:09] asac: gcl (source: gcl): GNU Common Lisp compiler. In component universe, is optional. Version 2.6.7-45ubuntu1 (karmic), package size 45862 kB, installed size 152848 kB [13:10] gcl is evil in several ways [13:10] It has an embedded copy of binutils and gcc :-/ [13:10] hmm. plenty of stuff i guess [13:10] ok so [13:10] grep gcl- * | grep -v binutils [13:10] search-arm-mov.filt: ./gcl-2.6.7/gmp3/mpn/arm/invert_limb.asm: add r2, pc, #invtab-.-8 [13:10] gcl - just embedded binutils, or is there anything else? [13:10] search-arm-mov.filt: ./gcl-2.6.7/gmp3/mpn/arm/udiv.asm: mov pc, lr [13:11] the rest should be covered by binutils [13:11] embedded gmp3 might be the same as gmp? [13:11] Probably, looking at our previous review [13:11] hmm [13:11] asac, I recommend you check the size of gcl's diff.gz if you want to see why I think gcl is crack [13:12] !info gcl [13:12] dmart: gcl (source: gcl): GNU Common Lisp compiler. In component universe, is optional. Version 2.6.7-45ubuntu1 (karmic), package size 45862 kB, installed size 152848 kB [13:12] !info gclcvs [13:12] asac: gclcvs (source: gclcvs): GNU Common Lisp compiler, CVS snapshot. In component universe, is optional. Version 2.7.0-64 (karmic), package size 48061 kB, installed size 175136 kB [13:12] ugh [13:12] ;) [13:12] lol [13:12] anyway, it looks to me like the hits are in imported packages [13:12] NCommander: i didnt say that gcl isnt crack ;) ... i dont know the code [13:13] binutils again etc. [13:13] damn [13:13] same as gcl [13:13] !info gdal [13:13] asac: Package gdal does not exist in karmic [13:13] ok, needs a look in principle, but the problems are probably answered by addressing gmp and binutils (binutils may already be done anyway, depending on the age of the snapshot) [13:13] x86 assembbler [13:14] For tools packages in general, we should try to find out if they're used to build anything [13:14] hmm. good point [13:16] Can we tag such packages with a magic word, or collect them somewhere? Sounds like someone could do a 1-pass review for build-deps later [13:17] not sure ... added that as a comment [13:17] we can decide when filing that bug [13:17] etc. [13:17] gdal - "Geospatial Data Abstraction Library" [13:17] thats x86 asm [13:17] !gdb-avr [13:17] Factoid 'gdb-avr' not found [13:18] hmm. some disasm [13:18] gdal - agreed [13:18] guess is related to gdb [13:18] will put that on doko plate [13:18] I guess it's gdb for the avr arch [13:18] ? [13:18] hmm [13:18] gdm-m68blah similarly [13:19] there is arm code inside the -avr [13:19] not sure its because its a full copy of something [13:19] ok lets treat that as arch dependent [13:19] That's probably just be a binutils dump [13:19] geshi [13:19] !info geshi [13:19] asac: Package geshi does not exist in karmic [13:19] false-+ [13:19] php code [13:20] "generic syntax highlighter" [13:20] agreed [13:20] !info ghdl [13:20] asac: ghdl (source: ghdl): VHDL compiler/simulator using GCC technology. In component universe, is optional. Version 0.27+svn110+gcc4.3.3+dfsg-1 (karmic), package size 11429 kB, installed size 32804 kB [13:20] hmm. gcc copy [13:20] needs to be checked i guess [13:20] esp. because its 4.3.3 [13:20] ->doko? [13:21] lets say its doko ... yes. marked [13:21] !gimp-gap [13:21] Factoid 'gimp-gap' not found [13:21] !info gimp-gap [13:21] asac: gimp-gap (source: gimp-gap): The GIMP Animation Package. In component universe, is optional. Version 2.4.0-2ubuntu1 (karmic), package size 2526 kB, installed size 7896 kB [13:21] false-+ [13:21] yes [13:21] !info gnat-4.3 [13:21] asac: gnat-4.3 (source: gnat-4.3): The GNU Ada compiler. In component universe, is optional. Version 4.3.4-1ubuntu1 (karmic), package size 12284 kB, installed size 55668 kB [13:21] gnat-3.4 -> doko [13:21] i think we already have gnat [13:21] yes [13:22] I think it comes out of the GCC tree anyway [13:22] !info gnu-smalltalk [13:22] asac: gnu-smalltalk (source: gnu-smalltalk): GNU Smalltalk interpreter and image. In component universe, is extra. Version 3.0.3-2 (karmic), package size 641 kB, installed size 2208 kB [13:22] uses call_reg macro [13:22] libffi dupe again [13:22] probably ok but needs verified [13:22] yeah [13:22] !info gnuplot [13:22] asac: gnuplot (source: gnuplot): A command-line driven interactive plotting program. In component universe, is optional. Version 4.2.5-2 (karmic), package size 1 kB, installed size 20 kB [13:23] os2 ;) [13:23] false-+ [13:23] !info google-perftools [13:23] asac: Package google-perftools does not exist in karmic [13:23] hmm linux syscall [13:24] might need a look [13:24] my guess is that it may not be critical [13:24] yes [13:25] !info gromacs [13:25] asac: gromacs (source: gromacs): Molecular dynamics simulator, with building and analysis tools. In component universe, is extra. Version 4.0.5-4 (karmic), package size 3684 kB, installed size 13188 kB [13:25] false-+ [13:25] false [13:25] y [13:25] ok guess gstreamer0.10-ffmpeg is dealt by ffmpeg [13:25] probably... is that a separate source package? [13:25] hmm. maybe not [13:26] seems so [13:26] suihkulokki, i get a lot of "unsupported syscall 335" in recent qemu-arm-static, that appears to be utrace, any idea about an existing fix ? [13:26] embedded ffmpeg - someone should check it. It may be one of those cases where the upstream project embeds a lib but we use the installed one (ie from the ffmpeg source package) [13:27] gxemul - emulator - false-+ [13:27] yes [13:28] helix-player - a mov pc, lr; and some swp which might need porting [13:28] helix-player -> uses atomics [13:28] yes [13:28] hol88 [13:28] !info hol88 [13:28] asac: hol88 (source: hol88): Higher Order Logic, system image. In component universe, is optional. Version 2.02.19940316-8 (karmic), package size 9530 kB, installed size 44532 kB [13:28] scary [13:28] false [13:28] + [13:28] yes [13:28] !info ht [13:28] asac: ht (source: ht): Viewer/editor/analyser (mostly) for executables. In component universe, is optional. Version 2.0.17-1 (karmic), package size 619 kB, installed size 1668 kB [13:29] arm-dis [13:29] looks like a disassembler [13:29] seems to not apply [13:29] hurd - also scary :P has this been ported to arm at all? [13:29] !info hurd [13:29] asac: Package hurd does not exist in karmic [13:30] heh [13:30] thats out [13:30] doesnt apply ;) [13:30] !info imview [13:30] asac: imview (source: imview): Image viewing and analysis application. In component universe, is optional. Version 1.1.9c-4ubuntu2 (karmic), package size 608 kB, installed size 1560 kB [13:30] hurd - guess not (or someone else's problem) [13:30] false [13:31] (imview) [13:31] yes [13:31] !info insight [13:31] asac: insight (source: insight): Graphical debugger based on GDB. In component universe, is optional. Version 6.7.1.dfsg.1-10.1ubuntu2 (karmic), package size 1804 kB, installed size 5284 kB [13:31] hmm [13:31] smells like something was copied ;) [13:31] i saw the sim/testsuite at least ;) [13:32] also ./insight-6.7.1.dfsg.1/bfd/elf32-arm.c: [13:32] thats some toolchain stuff? [13:32] embedded gdb, bfd and stuff ... yeah, bfd comes from binutils [13:32] ogra: thats pselect6, feel free to implement :) [13:32] one for doko? I don't know what the /sim/ stuff is or how must we care [13:33] suihkulokki, ogra: was pselect6 implemented for arm yet? [13:34] asac: insighttoolkit - false-+ (x86 asm) [13:34] ironpython - false-+ [13:35] jamvm - swp-based atomics and native calls with mov pc, lr [13:36] * dmart grabs a sandwich [13:36] * asac too [13:37] i doubt pselect was implemented [13:37] well pselect6 [13:38] ok did a checkpoint save on wiki [13:38] getting some bread too [13:39] * NCommander grumbles [13:45] ok seems we did a big chunk ;) [13:45] still only half way through :( [13:45] hmm [13:45] kaya [13:46] * ogra doesnt get why he still sees qemu: Unsupported syscall: 242 [13:46] kino [13:46] ogra: i dont get why it doesnt work at all for me ;) [13:46] i am on lucid and plain lucid kerenel [13:46] according to linux-user/arm/syscall_nr.h thats supported now [13:46] still get the binfmt problem :( [13:46] weird [13:46] works flawless here [13:46] even after the upgrade [13:47] can you try to run build-arm-choort standalone ? [13:47] i will try to just use qemu-arm-static chroot manually [13:47] *chroot [13:47] right. thats the idea [13:47] thats what i just did here [13:47] ruinning [13:47] to find out if mono started to behave [13:47] but apparently its not [13:48] ogra: that doesn't list supported syscalls. it lists syscall numbers. [13:48] oh, i thought they are only added there when someone implemented it actually [13:48] read linux-user/syscall.c to find what is really supported [13:48] ok, that expleins it then [13:49] asac: did you get my comments up to jamwm? [13:49] i know there exists a patch for pselect and ppoll somewhere [13:49] yes [13:49] ok [13:49] dmart: i thought i saved that alradcy [13:49] but no idea if that was ever applied [13:49] so lets continue [13:49] kaya and kino ;) [13:50] I didn't read the scrollback very carefully... [13:50] kaya false-+ [13:50] jocaml? [13:51] jocaml - some assembler with mov pc stuff, probably needs a look [13:51] yes [13:51] thats what is already in there [13:51] skipped it [13:51] oh yes [13:51] and no, getaffinity will not help you get mono work in qemu linux-user if that is why you are after 242 [13:51] kino seems armv4l [13:51] should be checked [13:52] cross-ref kino to ffmpeg - it's a embedded ffmpeg (may not be a problem if the system ffmpeg libs are used) [13:52] hmm seems to be fffmpeg [13:53] kompozer -> is a mess [13:53] really old mozilla code [13:53] dont think we can port that [13:53] maybe it should finally get removed from archive [13:53] !info kompozer [13:53] dmart: kompozer (source: kompozer): complete Web Authoring System. In component universe, is optional. Version 1:0.8~alpha4+dfsg+svn163-2 (karmic), package size 7328 kB, installed size 19580 kB [13:53] hmmm, maybe make a note, but otherwise not too important I guess [13:54] yes [13:54] done [13:54] !info kvirc [13:54] asac: kvirc (source: kvirc): KDE based next generation IRC client with module support. In component universe, is optional. Version 4:4.0.0~svn3240-1 (karmic), package size 2976 kB, installed size 9276 kB [13:54] kvirc - false-+ [13:54] ack [13:55] !info libgd-gd2-noxpm-perl [13:55] asac: libgd-gd2-noxpm-perl (source: libgd-gd2-noxpm-perl): Perl module wrapper for libgd - gd2 variant without XPM support. In component universe, is extra. Version 1:2.39-2 (karmic), package size 216 kB, installed size 644 kB [13:55] false [13:55] + [13:55] agreed [13:55] !info libgii [13:55] asac: Package libgii does not exist in karmic [13:55] atomics [13:55] !info libnet-dri-perl [13:55] asac: Package libnet-dri-perl does not exist in karmic [13:55] yes, may need a look [13:55] (libggi) [13:55] false-+ [13:56] !info libopenspc [13:56] asac: Package libopenspc does not exist in karmic [13:56] other arch it seems [13:56] !info libtk-img [13:56] asac: libtk-img (source: libtk-img): Extended image format support for Tcl/Tk (runtime). In component universe, is optional. Version 1:1.3-release-8 (karmic), package size 118 kB, installed size 464 kB [13:56] libopenspc - might be a snes music chip emulator lib iirc [13:56] also other arch false-+ [13:57] yes, last two are other arch afaics [13:57] !info libv8 ? [13:57] asac: '?' is not a valid distribution: hardy, intrepid, jaunty, karmic, lucid [13:57] !info libv8 [13:57] asac: Package libv8 does not exist in karmic [13:57] how did that get in that list? isnt libv8 the v8 js engine of chromium? [13:57] afaik thats only packaged outside of archive [13:58] anyway, needs a check ... generates arm code with mov [13:59] !info lifelines [13:59] asac: lifelines (source: lifelines): text-based genealogy software. In component universe, is optional. Version 3.0.61-1 (karmic), package size 967 kB, installed size 2392 kB [13:59] false+ [13:59] * dmart acks recent packages [13:59] !info lightning-sunbird [13:59] asac: Package lightning-sunbird does not exist in karmic [13:59] ok that one is mozilla code base [13:59] we will move to something based on 1.9.2 branch [14:00] so should be ok once we have the thumb patch landed [14:00] (otherwise we have the patch) [14:00] so: needs bump to 1.9.2 codebase + thumb2 patch [14:00] I guess we should keep track of all these x has embedded y somehow [14:01] yeah [14:01] linux-rt - linux snapshow [14:01] well. we open bugs and since we do it from top down we can add new tasks to the other bug [14:01] yes, linux-rt -> dealt elsewhere? [14:01] agreed [14:01] not sure where ;) [14:01] we can assign a bug to cooloney or someone to ack that [14:02] !info llvm-gcc-4.2 [14:02] asac: llvm-gcc-4.2 (source: llvm-gcc-4.2): C/C++ front end for LLVM compiler. In component universe, is optional. Version 2.6~pre1-0ubuntu1 (karmic), package size 23726 kB, installed size 107388 kB [14:02] doko gcc [14:02] ? [14:02] I think this is llvm, not gcc; for main we just said to review; we can probably link those together [14:02] ok [14:03] !info ltp [14:03] asac: ltp (source: ltp): The Linux Test Project test suite. In component universe, is optional. Version 20090531+dfsg-4ubuntu1 (karmic), package size 51 kB, installed size 360 kB [14:03] false-+ [14:03] !info lwp [14:03] asac: Package lwp does not exist in karmic [14:03] needes a check [14:03] mov [14:03] hmm [14:03] ltp - agreed [14:04] lwp - agreed (dunno what this is) [14:04] !info lyx [14:04] asac: lyx (source: lyx): Document Processor. In component universe, is optional. Version 1.6.4-1ubuntu1 (karmic), package size 3093 kB, installed size 7924 kB [14:04] atomics check [14:04] esems to be boost [14:04] suihkulokki, hmm, there seems to be anold suse patch for 242 http://paste.ubuntu.com/364585/ [14:04] marking [14:04] *an old [14:05] asac: lyx - agreed [14:05] !info mapivi [14:05] asac: mapivi (source: mapivi): Photo viewer and organizer with emphasis on IPTC fields. In component universe, is optional. Version 0.9.7-1 (karmic), package size 426 kB, installed size 1676 kB [14:05] false-ü [14:05] yes [14:05] !info mapserver [14:05] asac: Package mapserver does not exist in karmic [14:05] as well [14:05] !info metacity-themes [14:05] asac: metacity-themes (source: metacity-themes): Themes for the Gtk2 metacity window manager. In component universe, is optional. Version 1.0.11 (karmic), package size 384 kB, installed size 1956 kB [14:05] dito [14:06] mapserver, metachtiy-themes - agreed [14:06] * asac fighting with wiki ;) [14:07] ok [14:07] !info mit-scheme [14:07] asac: mit-scheme (source: mit-scheme): MIT/GNU Scheme development environment. In component universe, is optional. Version 7.7.90+20090107-1ubuntu1 (karmic), package size 6662 kB, installed size 18380 kB (Only available for i386) [14:07] mit-scheme - x86 asm for scheme compiler... false-+ [14:07] ack [14:07] !info mlton [14:07] asac: mlton (source: mlton): Optimizing compiler for Standard ML. In component universe, is optional. Version 20070826-1 (karmic), package size 10647 kB, installed size 47156 kB (Only available for amd64 hppa i386 powerpc sparc) [14:07] same [14:07] looks like it [14:07] !info mupen64plus [14:07] asac: mupen64plus (source: mupen64plus): plugin-based Nintendo 64 emulator. In component universe, is optional. Version 1.5+dfsg1-7 (karmic), package size 1308 kB, installed size 4288 kB [14:07] hope thats too ;) [14:08] !info mysql-dfsg-5.0 [14:08] asac: Package mysql-dfsg-5.0 does not exist in karmic [14:08] mupen64plus - yes [14:08] atomics ... do we already have a mysql one in main? [14:08] hmmm, I can't find it [14:09] guess that may need a look [14:09] !info netsurf [14:09] dmart: netsurf (source: netsurf): Small portable web browser with CSS and Unicode support. In component universe, is extra. Version 1.2-1build1 (karmic), package size 416 kB, installed size 1232 kB [14:09] ...but the matches are in /riscos/, so I don't expect it applies for us [14:10] openbios-ppc - surely not relevant [14:11] ok netsurf is false+ then with riscos [14:11] !info openocd [14:11] asac: openocd (source: openocd): Open on-chip JTAG debug solution for ARM and MIPS systems. In component universe, is extra. Version 0.0+r2403-1 (karmic), package size 1882 kB, installed size 4960 kB (Only available for alpha amd64 arm armel armeb hppa hurd-i386 i386 ia64 kfreebsd-amd64 kfreebsd-i386 m32r m68k mips mipsel netbsd-alpha netbsd-i386 powerpc sh sparc) [14:11] ogra: which is buggy. and as I said, wont help with mono. [14:11] openocd - some kind of embedded hardware debugger? Probably not relevant for us. [14:11] interesting [14:11] guess that needs a check [14:11] hmm [14:12] maybe take a quick look, but my guess is that the asm doesn't run on the host [14:12] i dont know how that is run. guess in worst case it can be built with -marm? [14:12] hmm code be [14:12] could [14:12] Hmmm, yes, I guess the code will get built by something. Maybe you're expected to install cross tools packages or something. [14:13] !info openser [14:13] asac: openser (source: openser): very fast and configurable SIP proxy. In component universe, is optional. Version 1.3.2-3 (karmic), package size 1357 kB, installed size 3856 kB [14:13] i like those "very, very" software ;) [14:13] atomics candidate [14:13] !info oprofile [14:13] asac: oprofile (source: oprofile): system-wide profiler for Linux systems. In component universe, is optional. Version 0.9.4+cvs20090629-2.1ubuntu2 (karmic), package size 1365 kB, installed size 5152 kB (Only available for i386 ia64 alpha hppa powerpc sparc amd64 arm armel mips mipsel s390 lpia) [14:13] openocd - Depends: libc6 (>= 2.8), libftdi1 (>= 0.17), libgcc1 (>= 1:4.4.0), libusb-0.1-4 (>= 2:0.1.12), dpkg (>= 1.15.4) | install-info (not very informative) [14:13] false-+ [14:14] oprofile seems to work (-ish) on lucid, though I do have to unload and reinitialise it between profiling runs. [14:14] dmart: yeah. i marked openocd package for manual review [14:14] yeah. but there is no real code in the filter logs [14:15] so i think its false-positive for this list [14:15] !info osiris [14:15] asac: osiris (source: osiris): network-wide system integrity monitor control interface. In component universe, is optional. Version 4.2.3-3 (karmic), package size 392 kB, installed size 916 kB [14:15] oprofile - agreed [14:15] atomics ... db4.2 copy [14:15] dupe [14:15] yes [14:15] !info palbart [14:15] asac: palbart (source: palbart): An enhanced version of the PAL PDP8 assembler. In component universe, is extra. Version 2.4-5 (karmic), package size 17 kB, installed size 80 kB [14:15] false-+ [14:16] paraview [14:16] !info paraview [14:16] hmmm, 60-70% of the way through the list [14:16] asac: paraview (source: paraview): Parallel Visualization Application. In component universe, is extra. Version 3.4.0-4ubuntu4 (karmic), package size 43766 kB, installed size 154432 kB [14:16] yes. we can stop soon ;=) [14:16] paraview -> x86 [14:16] :) [14:16] palbart - agreed [14:16] !info parrot [14:16] asac: parrot (source: parrot): A virtual machine for dynamic languages. In component universe, is optional. Version 1.4.0-1ubuntu1 (karmic), package size 205 kB, installed size 728 kB [14:16] paraview - agreed [14:17] has a mov pc, ip [14:17] Also in /jit/... guess it needs a look [14:17] yes [14:17] !info pdl [14:17] asac: pdl (source: pdl): perl data language: Perl extensions for numerics. In component universe, is optional. Version 1:2.4.3-8ubuntu1 (karmic), package size 4781 kB, installed size 15648 kB [14:17] false-+ [14:17] good [14:17] !info perl-tk [14:17] asac: perl-tk (source: perl-tk): Perl module providing the Tk graphics library. In component universe, is optional. Version 1:804.028-5 (karmic), package size 2455 kB, installed size 8596 kB [14:17] pdl - agreed [14:17] x86 [14:18] perl-tk - agreed [14:18] !info petsc [14:18] asac: Package petsc does not exist in karmic [14:18] alse false-+ [14:18] agreed [14:18] !info php-apc [14:18] asac: php-apc (source: php-apc): APC (Alternative PHP Cache) module for PHP 5. In component universe, is optional. Version 3.0.19-2 (karmic), package size 56 kB, installed size 172 kB [14:18] swp-based spinlock [14:18] yes [14:18] needs a look [14:18] !info plt-scheme [14:18] asac: plt-scheme (source: plt-scheme): PLT Scheme Programming Environment. In component universe, is optional. Version 4.1.5-1 (karmic), package size 32256 kB, installed size 147924 kB [14:19] similar [14:19] libffi [14:19] yes [14:19] and mzscheme [14:19] never heard of that [14:19] mzscheme? [14:19] ok comment is: "libffi call_reg + mzscheme swp/atomics" [14:19] well ;) the source file is: [14:19] ok [14:20] ./plt-scheme-4.2.1/src/mzscheme/gc/include/private/gc_locks.h: [14:20] no clue if that is a third party [14:20] !info postgresql-8.3 [14:20] asac: postgresql-8.3 (source: postgresql-8.3): object-relational SQL database, version 8.3 server. In component universe, is optional. Version 8.3.8-1 (karmic), package size 5011 kB, installed size 14132 kB [14:20] yeah, dunno [14:20] guess that needs atomics [14:20] right... same fix we have for 8.4 [14:20] ./postgresql-8.4-8.4.1/src/include/storage/s_lock.h: " swpb %0, %0, [%2] \n" [14:20] yes, link it to that [14:20] done [14:20] !info ptop [14:20] asac: ptop (source: ptop): PostgreSQL performance monitoring tool akin to top. In component universe, is optional. Version 3.6.2-1 (karmic), package size 43 kB, installed size 144 kB [14:20] false-+ [14:21] yes [14:21] python* -> doko? [14:21] python is doko [14:21] !info r-cran-eco [14:21] asac: r-cran-eco (source: r-cran-eco): GNU R routines for Bayesian ecological inference. In component universe, is optional. Version 3.1-4-1 (karmic), package size 248 kB, installed size 744 kB [14:21] r-cran-* - looks like false-+ [14:21] false-+ [14:21] yes [14:21] (both) [14:21] !info radare [14:21] dmart: radare (source: radare): free advanced command line hexadecimal editor. In component universe, is optional. Version 1:1.4-1 (karmic), package size 596 kB, installed size 1332 kB [14:21] !info radare [14:21] asac: radare (source: radare): free advanced command line hexadecimal editor. In component universe, is optional. Version 1:1.4-1 (karmic), package size 596 kB, installed size 1332 kB [14:21] ! [14:22] disassembler [14:22] guess can be ignored [14:22] a hex editor with an arm disassembler in it? [14:22] weird [14:22] hehe [14:22] ;) [14:22] sounds like emacs ;) [14:22] maybe i should try that :) [14:22] !info rlplot [14:22] asac: rlplot (source: rlplot): Generate publication quality graphs. In component universe, is optional. Version 1.4-1 (karmic), package size 955 kB, installed size 2440 kB [14:22] can see how that might be useful [14:22] false [14:22] + [14:22] yes [14:23] !info root-system [14:23] asac: root-system (source: root-system): Meta package to install all ROOT packages. In component universe, is optional. Version 5.18.00-2.3ubuntu2.1 (karmic), package size 26 kB, installed size 72 kB [14:23] x86 [14:23] root-system - x86 asm false-+ [14:23] !info rtai [14:23] asac: rtai (source: rtai): Real Time Application Interface. In component universe, is extra. Version 3.6.1-1 (karmic), package size 34 kB, installed size 68 kB [14:23] some arm asm in there... probably needs a look [14:23] whats that? with imx patches? [14:24] (atomics, pc arith) [14:24] unless this is a bare-metal thing? [14:24] Probably imx < imx51 [14:24] yeah [14:24] !info scummvm [14:24] asac: scummvm (source: scummvm): free implementation of LucasArts' SCUMM interpreter. In component universe, is optional. Version 1.0.0~rc1-1-1 (karmic), package size 3465 kB, installed size 9152 kB [14:24] (there are some earlier architectures) [14:24] seems it has mov [14:25] search-arm-mov.filt: ./scummvm-1.0.0~rc1-1/backends/platform/ds/arm9/source/interrupt.s: mov pc,lr [14:25] ? [14:25] or isnt arm9 what we are doing? [14:25] hmm .. .gues snot [14:25] needs to be checked ;) [14:25] might need a look... it might be in sound chip cmulation code or something though [14:25] have to have scummvm working on netbooks though :) [14:25] yes [14:25] !info sdparm [14:25] asac: sdparm (source: sdparm): Output and modify SCSI device parameters. In component universe, is optional. Version 1.02-1 (karmic), package size 113 kB, installed size 372 kB [14:25] false [14:26] yes [14:26] ok seamonkey is mozilla gain [14:26] needs 1.9.2 [14:26] etc. [14:26] ok [14:26] see - false-+ [14:26] !info skyeye [14:26] asac: skyeye (source: skyeye): Embedded Hardware Simulation. In component universe, is extra. Version 1.2.5-2ubuntu1 (karmic), package size 239 kB, installed size 700 kB [14:26] Did we alreay have ser? [14:27] oh skipped ;) [14:27] sorry [14:27] !info ser [14:27] asac: ser (source: ser): high-performance SIP server. In component universe, is optional. Version 2.0.0-2 (karmic), package size 1488 kB, installed size 3876 kB [14:27] some atomics, it does use lrdex/strex but we need to check that is actually used on v7, and that it's smp-safe [14:28] skyeye - a emulator of embedded arm platforms, guess it doesn't apply [14:29] * dmart scrolls again [14:29] I can see the end of the table now [14:29] sm - looks like false-+ [14:30] !info softgun [14:30] dmart: softgun (source: softgun): ARM system emulator. In component universe, is optional. Version 0.16-2ubuntu1 (karmic), package size 372 kB, installed size 1096 kB [14:30] softgun - false-+ (arm instruction decode emulation...) [14:30] sorry got pulled out ;) ... back now [14:31] I think we had skyeye, softgun since [14:31] Plus ser if you didn't note that yet [14:32] spu-newlib - code looks similar / the same as newlib from main. I guess we can link the two. [14:32] !info survex [14:32] dmart: survex (source: survex): cave surveying and mapping software. In component universe, is extra. Version 1.0.39.1-2.1 (karmic), package size 656 kB, installed size 1700 kB [14:32] some asm code in /riscos/ ... probably doesn't apply [14:33] softgun, yes. [14:34] synopsis - embedded libatomic_ops [14:35] (or very similar code anyway). Link those? [14:36] eah [14:36] yeah [14:36] !info systemtap [14:36] dmart: systemtap (source: systemtap): instrumentation system for Linux 2.6. In component universe, is optional. Version 0.0.20090613-1 (karmic), package size 851 kB, installed size 3440 kB (Only available for i386 amd64 ia64 s390 powerpc arm armel armeb all) [14:36] I guess it needs a look... sub , pc, 4 [14:36] yes [14:36] What does ubot mean by (Only available for [...] all)? [14:36] uboot-imx -> skip [14:37] uboot-imx - yes [14:37] no clue [14:37] thats a bug [14:37] Well, I guess it's logically true ;) [14:37] !info asac [14:37] armin76: Package asac does not exist in karmic [14:38] we are really almost done [14:38] like 12 to go ;) [14:38] !info uclmmbase [14:38] asac: Package uclmmbase does not exist in karmic [14:38] looks like false-+ to me [14:39] ack [14:39] virtualbox-ose - what's this doing with embedded mozilla code? [14:39] dont ask me :( [14:39] needs to be check [14:39] ed [14:40] !info visualboyadvance [14:40] asac: visualboyadvance (source: visualboyadvance): full featured Game Boy Advance emulator. In component universe, is extra. Version 1.8.0-5 (karmic), package size 308 kB, installed size 1360 kB [14:40] visualboy advance - emulator, not relevant I guess [14:40] haha [14:40] yeah [14:40] !info vtk [14:40] asac: Package vtk does not exist in karmic [14:41] x86 [14:41] seems to be some x86 asm [14:41] asac, did your test with build-arm-chroot work ? [14:41] ogra: failed the same way :( [14:41] I: Extracting sysvinit-utils... [14:41] I: Extracting tar... [14:41] I: Extracting tzdata... [14:41] I: Extracting upstart... [14:41] I: Extracting util-linux... [14:41] I: Extracting zlib1g... [14:41] chroot: cannot run command `debootstrap/debootstrap': Exec format error [14:41] meh [14:41] websvn -> false-+ [14:41] try to reinstall qemu-arm-static [14:41] wrong arch? [14:41] !info wine [14:41] asac: wine (source: wine): Microsoft Windows Compatibility Layer (Binary Emulator and Library). In component universe, is optional. Version 1.0.1-0ubuntu8 (karmic), package size 7359 kB, installed size 54436 kB [14:41] ogra: will do. does that compile something on install? [14:42] false positive (wine) [14:42] it sets the sysctl stuff for procps [14:42] I don't expect wine works on arm? [14:42] same for wine1.2 [14:42] heh [14:42] and dumps the binfmt hgandler in place [14:42] would be funny if it did ;) [14:42] *handler [14:42] wireshark -+ [14:42] ogra: let me finish this ;) ... [14:42] probably wine-ce :) [14:42] almost dying ;) [14:42] wxwidgets* -+ (same x86 code as vtk) [14:42] !info wireshark [14:42] asac: wireshark (source: wireshark): network traffic analyzer - GTK+ version. In component universe, is optional. Version 1.2.2-2 (karmic), package size 716 kB, installed size 1824 kB [14:42] (only 6 more) [14:43] asac, yeah, i see W :) [14:43] ah [14:43] yeah thats fals-+ [14:43] and the ex thing too [14:43] !info xenomai [14:43] dmart: Package xenomai does not exist in karmic [14:43] bah [14:43] !info xemacs21-packages [14:43] asac: Package xemacs21-packages does not exist in karmic [14:43] guess emacs is covered [14:43] yes [14:43] xenomai has swp-based atomics and other asm [14:44] probably needs a look [14:44] yeah. lots of stuff in there [14:44] !info xindy [14:44] asac: xindy (source: xindy): index generator for structured documents like LaTeX or SGML. In component universe, is optional. Version 2.3-2 (karmic), package size 2245 kB, installed size 3656 kB [14:44] I predict -+ [14:44] it has mov etc. [14:45] oh [14:45] rte/... [14:45] that looks familliar [14:45] is that a code copy again? [14:45] clisp-2.43/ffcall [14:45] I don't think so (can't find it anyway) [14:45] yes [14:46] !info xmp [14:46] asac: xmp (source: xmp): A module player supporting AWE32, GUS, and software-mixing. In component universe, is extra. Version 2.7.0-0ubuntu1 (karmic), package size 235 kB, installed size 560 kB [14:46] some embedded lisp interpreter anyway [14:46] false-+ [14:46] !info xulrunner [14:46] asac: xulrunner (source: xulrunner): XUL + XPCOM application runner. In component universe, is optional. Version 1.8.1.16+nobinonly-0ubuntu1 (karmic), package size 279 kB, installed size 1020 kB [14:46] -> covered by xul 1.9.2 + patch [14:46] still needed (so a bug) [14:46] !info yabause [14:46] asac: yabause (source: yabause): beautiful and under-rated Saturn emulator. In component universe, is optional. Version 0.9.10-1 (karmic), package size 16 kB, installed size 52 kB [14:46] (LAST ONE) [14:47] x86 false + [14:47] -+ yes [14:47] yessssss [14:47] whats a saturn emulator [14:47] The author is modest in the description ;) [14:47] sega saturn? [14:47] ok saving [14:47] uff [14:47] i think i need a break [14:47] ah, sega [14:48] i was wondering about something kosmic :) [14:48] main() { while(1) be big and round; } [14:48] dmart: did you move the false-positives out to a separte table for main? [14:49] i guess we should do tha tbefore starting to file bugs ;) [14:49] I made a heading with "Disregarded packages (no further action required)" - the junk is under there. We can maybe move all that stuff to the end of the page [14:49] yeah [14:50] or to a separte wiki page even [14:50] For universe, I already made [DISREGARDED], though it is not a proper wiki heading [14:50] to make that page less huge [14:50] Yes, a separate page would be good [14:50] we should surely add a ToC [14:50] How do you do that in the Ubuntu wiki? [14:50] one second there are examples [14:51] like on this page: https://wiki.ubuntu.com/MozillaTeam [14:51] so ||<>|| [14:51] guess just <> wrks [14:51] but looks bad [14:51] let me add it [14:51] okay [14:51] thanks [14:52] ok its on there [14:52] a bit huge ;) [14:52] it works [14:52] hmm. let me remove the width constraint [14:52] thats bad [14:53] ok saved [14:53] its better now [14:53] dont think we need a new page with that [14:53] ok [14:53] just so that the not so releva nt stuff gets to the bottom or osmething [14:54] I should write a wiki page with an overview of what sort of action is needed to fix the different kinds of matches, then we can link to if from the bugs raised [15:46] dmart: yes, if you can do that it would be great [15:46] we could also port a few different cases, and then point to the patches as examples [15:46] if thats easier [15:46] Yes, that would be good. [15:47] I'm writing some stuff for the wiki now... [15:48] The "mov pc" type cases, and cases where "swp" is causing ftbfs, are the highest priority things for luci, since these may break function calls. [15:49] The swp and smp-safety stuff is more for future-proofing and some performance aspects (though we should work through these, time allowing). [15:49] I'll try to explain this on the wiki... [16:14] 17:13 < seb128> asac, can somebody maybe get a stracktrace? [16:14] hmm, so sqid fails reliably at the same line with a segfault [16:14] plars: ^^ [16:14] thats on nautilus crash [16:14] *squid [16:15] plars: we are looking at the diff atm. [16:15] asac: this is on the gvfs thing? [16:15] plars: yes [16:15] asac: sure, let me see what I can do [16:15] ogra: we want the -save-temps .ii ... then give that doko [16:15] ? [16:15] guess needs a local build ;) [16:16] ogra: doko probably will probably want a .ii file for the source that causes the gcc crash [16:16] so build with -save-temps [16:16] in CFLAGS [16:16] oh, ok [16:17] dmart: thanks. [16:17] i will try to adjust priority for the bugs i file accordingly [16:17] making the mov pc and swp with ftbfs release critical [16:17] ogra: so what do you want from binfmt? [16:17] i ? [16:17] you want something from binfmt :) [16:17] check if /usr/share/binfmts/arm exists [16:17] find /sys/ | grep binfmt | pastebinit [16:17] http://pastebin.com/f63420615 [16:18] ok [16:18] and also check for /etc/sysctl.d/qemu-arm-static.sysctl.conf [16:18] these are the two files that enable qemu-arm-static [16:18] i have /usr/share/binfmts/arm [16:18] cat /etc/sysctl.d/qemu-arm-static.sysctl.conf [16:18] vm.vdso_enabled = 0 [16:18] vm.mmap_min_addr = 4097 [16:19] now reinstalling the quem-arm-static [16:19] one second [16:19] well, thats all looking good [16:19] do you have /etc/rc2.d/S90binfmt-support ? [16:19] asac: It's in /proc [16:19] /proc/sys/fs/binfmt_misc/arm usually [16:20] ogra: yes. [16:20] cat /proc/sys/fs/binfmt_misc/arm [16:20] enabled [16:20] interpreter /usr/bin/qemu-arm-static [16:20] flags: [16:20] offset 0 [16:20] magic 7f454c4601010100000000000000000002002800 [16:20] mask ffffffffffffff00fffffffffffffffffeffffff [16:20] looks good too i guess :( [16:20] lool, last time we checked his machine didnt even have binfmt_misc loaded [16:20] asac: Try running /usr/bin/qemu-arm-static [16:20] yeah. thats loaded now [16:20] (i force it in modules) [16:20] asac, by running the initscript ? [16:20] thats wrong [16:20] ogra: no in /etc/modules [16:20] the initscript wraps a lot of other stuff around it [16:21] guess i need some arm binary [16:21] check if you have /etc/rc2.d/S90binfmt-support in place [16:21] to test /usr/bin/qemu-arm-static [16:21] what would be a good one? [16:21] just run build-arm-chroot [16:21] ogra: i have /etc/rc2.d/S90binfmt-support [16:21] second stage execs in qemu-arm-static [16:21] ogra: well. thats what i ran before ;) [16:21] it failed the same way [16:21] then i dont get why you have to forcefully load the module [16:22] and since i didnt even change anything i would think it still fails [16:22] if the initscript runs it should load it [16:22] anyway. i reinstalled the package [16:22] let me try again [16:22] * asac runs [16:22] i suspect its rather binfmt-support [16:22] build-arm-chroot lucid lucid1 [16:22] than qemu-arm-static [16:22] +you should really use a local package proxy :) [16:23] what does /usr/sbin/update-binfmts do? [16:23] worth running? [16:23] it enables the arm binfmt handler [16:23] it runs from the postinst of qemu-arm-static [16:23] ok reinstalling the -support package [16:23] hmm [16:23] update-binfmts: warning: current package is openjdk-6, but binary format [16:23] already installed by sun-java6 * Enabling additional executable binary formats binfmt-support [16:24] guess thats not a prob [16:24] lets wait if the chroot fails again [16:24] yeah [16:24] i dont get why your binfmt module isnt loaded though [16:25] ogra: its loaded atm [16:25] havent checked if it still doesnt get loaded [16:25] no, i mean by the initscript [16:25] will reboot after the chroot fails [16:25] yeah. i will double check ... maybe its fixed now somehow ;) [16:25] heh [16:25] asac: wasn't having much luck getting a backtrace with symbols yesterday, but I'm not seeing any -dbg packages for gvfs stuff [16:25] plars: we have -dbgsym [16:25] http://ddebs.ubuntu.com/ [16:26] https://wiki.ubuntu.com/DebuggingProgramCrash [16:26] and -dbg packages should additionally be on LP [16:26] there are the apt lines you need to get -dbgsym for packages that dont have -dbg [16:26] not for all packages [16:26] usually for all gnome ones [16:28] ogra: none for gvfscommon [16:28] oh [16:28] plars: so yeah. grab the -dbgsym ... also for glib i guess and gio or whatever is involved [16:28] and libc-dbg [16:28] etc. [16:28] but guess you just lacked the ddeb trick [16:29] ogra: [16:29] I: Configuring libselinux1... [16:29] I: Configuring libstdc++6... [16:29] I: Configuring coreutils... [16:29] I: Configuring makedev... [16:29] that looks good, right? [16:29] yes [16:29] * asac thinks it didnt get that far last time [16:29] nice [16:29] so reinstall one of the two packages helped [16:29] so seems something with triggers isnt right [16:29] bad packaging ;) [16:29] i suspect binfmt-support [16:30] guess it doesnt ship any triggers [16:30] but should [16:30] since the module was missing [16:30] too bad i reinstalled both ;) [16:30] otherwise we would have known it now [16:30] or its just that lool broke qemu-arm-static with all his repackaging :P [16:30] * ogra hides [16:30] lol [16:31] did anything happen on that front at all? [16:31] thought its still the same somewhat [16:31] no, we dropped the spec [16:31] it works, why break it [16:31] i'D love to see mono support [16:31] then rootstock wouldnt need a vm [16:32] but given that this seems close to impossible to fix (its very old and many people tried before) i wont waste my time on it [16:38] ok now for rootstock [16:38] Âñåì äîáðûé âå÷åð [16:38] åñòü ðóññêèå? [16:38] Devils0411, english please :) [16:40] lool, did you get anywhere with the versatile kernel ? [16:40] Offtopic. Who knows, what OS in this device http://www.promise.com/product/product_detail_eng.asp?segment=undefined&product_id=211 [16:46] asac: still not having much luck [16:46] plars: how does the backtrace look like? [16:47] asac: http://paste.ubuntu.com/364678/ === bjf-afk is now known as _bjf [16:48] plars: post that to the bug and say thats best you can get [16:48] what is signal 6 ? [16:49] asac: sigabrt [16:49] thats not gvfs [16:49] ogra: the downgrade of libgvfscommon fixes it [16:49] cant you somehow get a gvfs BT [16:50] could be that its nautilus, but for now it feels its gvfs ... especially since that has a recent update [16:50] no, i meant the backtrace [16:50] well. nautilus links against libgvfs? [16:50] i wonder if there is a way to rather get something more diectly from gvfs [16:51] for that we need to find someting else that crashes i guess [16:52] That's what I found! / bin / uname-a [16:52] ---------------------------------- [16:52] Eval command:! / bin / uname-a [16:52] Linux nas_storage 2.6.20.21-4600 # 30 PREEMPT Fri Apr 10 12:06:05 CST 2009 i686 unknown [16:52] but what exactly is OS? [16:52] Devils0411: It could be in-house. Either way, it isn't arm based. [16:53] asac, well, apt-cache rdepends gvfs gets quite a bunch of binaries that depend on it [16:54] How to determine? [16:55] plars: do you see crashes in any of those? [16:55] evince [16:55] ekiga [16:55] guess you need to open something throw gvfs [16:55] i guess deb-gview would be an easy candidate to test from cmdline [16:55] asac: will have to try remotely, as the nautilus issue keeps me from getting into X successfully [16:55] seems a small thing [16:55] plars: cant you comment it out in gnome-session? [16:56] asac: then I just get a blank desktop, no gnome, just a cursor [16:56] so gnome-panel also dies? [16:56] evince works [16:56] see the reverse deps :) [16:56] does not crash [16:56] yeah, here too [16:57] installing ekiga [16:57] i guess until you open something [16:57] using a gvfs path [16:57] plars: did you open something from a gvfs location? [16:57] like ssh? [16:57] or just with gvfs path, yeah [16:57] not sure about the syntax though [16:59] i wonder what gcalctool does with gvfs [16:59] plars: maybe using evince file:///path/to/file.pdf ... would trigger gvfs code paths [16:59] it doesnt even have a file dialog [17:00] evince? [17:00] wow [17:00] o gcalctool [17:00] no, gcalctool [17:00] yea. i think its clutter [17:01] -Wl,--as-needed [17:01] missing [17:02] evince still working fine, even with file:// path [17:02] hmm, opening a file in evince complains a lot about gconfd not running [17:02] ogra: opening with file:// ? [17:02] ys [17:02] yep [17:02] anyway. [17:02] or with the file dialog [17:02] so i have to get the code ;) [17:02] the patch alone hasnt enough context [17:04] brasero works fine as well [17:04] i can add files etc [17:04] so its likely the way nautilus uses gvfs [17:05] oh [17:05] and i can fire off nautilus --no-desktop through ssh [17:06] so its nautilus in interaction with something that actually causes the crash [17:07] Are these problems the reason we don't have daily builds? [17:08] no [17:08] gnome uploads arent finished [17:08] there are out of sync packages [17:08] (the logs should tell) [17:08] Any eta? [17:08] no idea [17:08] its a development release :) [17:08] people upload stuff [17:23] GrueMaster, seems the breaking dove keeps imx51 from building, i manually triggered an imx51 build now [17:23] plars: so you say just gvfscommon fixes it? [17:23] yea. [17:23] libgvfscommon0 [17:24] downgrading by one level fixes [17:25] hmm [17:26] plars: all other gvfs packages are up to date? e.g. gvfs-backends? [17:26] asac: yes [17:26] and have the same versoin as the crasjing common package i guess? [17:27] asac: yes, all of them at 1.5.2-0ubuntu1 === asac_ is now known as asac [17:30] ok, all gstreamer bits (including totem) are gone from the ftbfs list [17:31] * ogra wishes the page would be updated more frequently than once a day [17:39] right [17:40] ogra: do you know who owns that? === Martyn is now known as Guest58882 === Martyn is now known as Guest63396 === mcasadevall is now known as NCommander [20:51] ogra: what was the url i had to use for the lucid kernel? [20:57] asac, which one ? for qemu ? [20:59] ogra: http://paste.ubuntu.com/364787/ [20:59] oh, yeah, thats the jaunty one [21:00] you try to build lucid, right ? [21:00] http://bazaar.launchpad.net/~ogra/project-rootstock/trunk/revision/31