=== bjf is now known as bjf_afk [08:59] ogra: hi [08:59] ogra: i'm playing with rootstock [08:59] I: Setting up serial tty in image [08:59] ./rootstock: 555: cannot create /tmp/tmp.QxQsDRVgaU/tmpmount/etc/init/ttyS0.conf: Directory nonexistent [09:00] i may have a patch for it soon [11:25] ogra: ping [11:25] hey goshawk, thanks for the fix :) [11:25] ogra: thx to you, it was just a mkdir -p :P [11:26] it wont save you though [11:26] as i wrote in the branch review [11:26] yes :) [11:26] i saw [11:26] and i'm fixing thsi now [11:26] /bin/installer: line 60: reboot: command not found [11:26] it's in upstart-compat-sysv [11:26] yeah [11:26] which seems to be not installed [11:26] right, which is dead [11:26] so i forcedo to install it [11:26] (Reading database ... 10802 files and directories currently installed.) [11:26] Unpacking upstart-compat-sysv (from .../upstart-compat-sysv_0.3.10-2_armel.deb) ... [11:26] it's 30 mins [11:26] at 100% [11:26] in that position [11:27] i just added a apt-get install -y (Reading database ... 10802 files and directories currently installed.) [11:27] Unpacking upstart-compat-sysv (from .../upstart-compat-sysv_0.3.10-2_armel.deb) ... [11:27] well, i'm not sure what this package needs from the old upstart [11:27] oops sorry [11:27] apt-get install -y upstart-compat-sysv [11:27] the former upstart had tools and binaries scattered across multiple packages [11:27] well the problem is that reboot is not found [11:27] ah... understood [11:27] the new one has everything in a single package [11:28] startup-tasks system-services upstart-compat-sysv and upstart-logd will likely be needed additionally [11:28] though ... [11:29] it might be that udev gets a prob alongside if certain sockets from new upstart arent available [11:29] uhm.. let me see where is /sbin/reboot in karmic [11:29] likely in upstart 0.6 [11:30] yeah, it is ... (on x86) [11:30] yep it's there [11:30] https://edge.launchpad.net/ubuntu/+source/upstart/0.6.0-2/+build/1112005/+files/buildlog_ubuntu-karmic-armel.upstart_0.6.0-2_FAILEDTOBUILD.txt.gz [11:30] thats the prob [11:30] and by the looks of it its caused by a gcc bug [11:30] so.. the question is... when will upstart will be built? [11:30] ah! [11:30] wow! [11:30] so it wil take a while to identify and fix i assume [11:31] ogra: it's nice to have the build daemon :) [11:31] ah, well ... its not very helpful with its error messages in this case [11:32] i only found the actual error by building upstart at home [11:32] uhm... it seems that test_node.c [11:32] did it kill cuz it hangS? [11:33] no [11:33] Session terminated, killing shell...make: *** wait: No child processesmake[1]: *** [check-recursive] Terminated [11:33] Build killed with signal 15 after 150 minutes of inactivity [11:33] gcc spills an error the log doesnt have [11:33] ah [11:33] ok, so i've to build by myself to see it [11:34] tests/test_watch.c: In function ‘test_new’: [11:34] tests/test_watch.c:590: internal compiler error: in df_ref_record, at df-scan.c:2845 [11:34] depending on what board i build t here i get different errors :/ [11:34] :( that's very weird [11:34] yes [11:35] fun is though that the package is built in a way that it should finish building even if the tests fail [11:35] and that actually works in a local build [11:35] can it be a problem related to gcc? [11:36] the buildd lock up though and kills the process after 150min [11:36] yes, very likely [11:36] i will need to talk with our gcc maintainer tomorrow [11:36] and with Keybuk who wrote upstart [11:37] ogra: it seems that gcc goes in loop for me [11:37] and it never exits [11:37] from compiling [11:37] (for the builder) [11:37] at home it gives - internal compiler error - [11:37] right [11:37] ogra: does it hang after "internal compiler error" [11:37] ? [11:38] no idea, a buildd admin might know [11:38] no i mean [11:38] at home [11:38] it might hang before [11:38] no, it doesnt hang at all [11:38] ah.. so it goes in the loop [11:38] »upstart« in »../upstart_0.6.0-3_armel.deb«. [11:38] it rols a package [11:40] uhm [11:43] ok... time to go for me [11:43] exam tomorrow [11:43] see you [11:48] * ogra files bug 398403 [11:48] Launchpad bug 398403 in gcc-4.4 "gcc-4.4 fails to build upstart 0.6 on armel due to an internal compiler error" [High,New] https://launchpad.net/bugs/398403 === Omegamoon is now known as Omegamoon|away [18:37] hmm, did you make the armv6+vfp switch in karmic yet? [18:37] as in, for gcc [19:34] hello [19:34] would someone be so kind and build a kernel for me to be used with qemu?) [19:36] there's a kernel that works with qemu.. try the -versatile kernel [19:36] well [19:36] i'd need one with preempt [19:37] im using the versatile kernel already, but I have some slightly tricky application which would benefit from preempt kernel [19:39] I see [19:41] or else, i wonder what all i need to be able to build the kernel myself? [19:41] can i just get the latest kernel and build it myself, or do I need to apply some patches? [19:43] uh, good question.. it's been a while since I had to build a kernel myself [19:44] but building for armel-versatile shouldn't be different than the normal procedure for building a kernel on ubuntu [19:44] well i need to crossbuild [19:45] i found http://www.emdebian.org/tools/crosstools.html [19:45] i wonder if apt-get install libc6-armel-cross libc6-dev-armel-cross binutils-arm-linux-gnueabi gcc-4.3-arm-linux-gnueabi g++-4.3-arm-linux-gnueabi gets me the right toolchain [19:45] i'll try :) [19:46] oh dear W: GPG error: http://www.emdebian.org lenny Release: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY B5B7720097BB3B58 [19:46] that looked too easy [19:46] can i get around that? [19:46] does ubuntu have crosstool packages?) [20:22] well i guess im going to try my luck with codesourcery toolchain then [20:23] http://www.nas-central.org/wiki/Setting_up_the_codesourcery_toolchain_for_X86_to_ARM9_cross_compiling [20:23] looks like pretty close what im doing [20:23] except im using vmware instead of virtualbox, but anyway :D [20:23] virtualbox bluescreened my macbook :( [20:25] I've had a very good run with the CodeSourcery toolchain [20:25] just remember you don't have NEON or good vfp support in the Lite version [20:25] well [20:26] im targeting for qemu [20:26] nod [20:27] there's also a good 4.2 and 4.4 toolchain set that supports armv5/6/7 [20:27] I'm using CodeSourcery only because I'm developing for the Cortex-A9 [20:30] uuh [20:31] i downloaded arm-2009q1-203-arm-none-linux-gnueabi.bin [20:31] it needs X11 huh [20:31] i dont have X11 :D [20:31] i dont event want one [20:32] Wait, needs x11? [20:32] why? [20:32] Wierd. [20:32] i suppose it has some graphical installer [20:34] Did you download that from silver.arm.com? [20:34] I'm trying to figure out which toolchain it's part of [20:35] from codesourcey [20:35] now i'll try this script http://www.nas-central.org/wiki/Setting_up_the_codesourcery_toolchain_for_X86_to_ARM9_cross_compiling [21:04] i wonder which kernel configuration to use for qemu [21:09] There's a standard one [21:09] make qemuconfig [21:09] or something along those lines [21:09] you'll find it in the kernel tree [21:10] how? [21:10] make qemuconfig didn't work [21:11] http://kernel.xc.net/search.cgi?string=qemu&version=2.6.30&arch=arm doesnt find much about qemu [22:33] oh dear [22:33] the kernel i compiled didn't start in qemu :/ [22:35] would anyone have their kernel config? [22:48] hmm [22:48] :1097:2: warning: #warning syscall fadvise64 not implemented [22:48] :1265:2: warning: #warning syscall migrate_pages not implemented [22:48] :1321:2: warning: #warning syscall pselect6 not implemented [22:48] :1325:2: warning: #warning syscall ppoll not implemented [22:48] :1365:2: warning: #warning syscall epoll_pwait not implemented [22:49] i wonder if that is anything to worry about