=== desrt_ is now known as desrt | ||
=== rickytaylor26 is now known as ricky26 | ||
=== ericm|ubuntu is now known as ericm-afk | ||
Xase | lilstevie, any thoughts? | 08:11 |
---|---|---|
=== ericm-afk is now known as ericm | ||
=== highvolt1ge is now known as highvoltage | ||
jamespage | good morning: should we be enabling neon support for libraries that support it for ARM on Ubuntu? | 09:30 |
infinity | jamespage: If they do runtime detection yes, if not, no. | 09:31 |
infinity | jamespage: (Not all our targets support neon, so we can't use it by default) | 09:31 |
jamespage | infinity, great - thats what I thought but wanted to check first - thanks! | 09:31 |
infinity | (We have lots of code in the archive that does runtime neon detection, and it's a huge win when you can) | 09:32 |
infinity | libjpeg-turbo being an obvious one. Pretty sure there's handcrafted neon in libssl. Etc, etc. | 09:32 |
infinity | Well, it's a huge win if it's well-written neon. :P | 09:32 |
ppisati | infinity: do you know if any babbage builder is using the kernel with the "large mmap() support"? | 09:42 |
ppisati | bug 861296 | 09:42 |
ubot2 | Launchpad bug 861296 in linux "mmap fails to allocate 2030Mb heap on ARM" [High,Confirmed] https://launchpad.net/bugs/861296 | 09:42 |
ppisati | infinity: or do you know to whom shall i talk for the arm builders? | 09:50 |
ogra_ | lamont | 09:51 |
ogra_ | but i think he switched tham already | 09:51 |
ogra_ | or at least he tested that kernel on one | 09:51 |
ppisati | OlivierN: mail sent | 09:59 |
ppisati | ops | 09:59 |
ppisati | ogra_: ^ | 09:59 |
ogra_ | :) | 10:01 |
lool | Hey | 10:20 |
lool | would someone know why "step" isn't built on ARM? | 10:20 |
lool | it seems to be purposedly disabled on Ubuntu only | 10:20 |
lool | (and this means kdeedu is uninstallable on armel) | 10:20 |
ogra_ | lool, i dont think anyone fo ubuntu-arm touched it ... probably the kde people did | 10:36 |
ogra_ | *of | 10:36 |
infinity | lool: Arch is specified in the source, no idea why. If it builds and runs, by all means, change that. | 11:16 |
lool | infinity: It seems to be an intentional change from Riddell, see latest changelog entry, that's why I was digging a bit | 11:26 |
lool | it's built in experimental (on armel) | 11:27 |
infinity | lool: Looks like it's been like that since it was first uploaded to oneiric. No clue why. | 11:42 |
lool | since I couldn't commit to its Vcs-Bzr which doens't exist, I've pinged the kubuntu team | 13:08 |
infinity | jcrigby: Are you going to have time to do some judicious s/armel/armel armhf/ (and other required tweaks) mojo on uboot-linaro and assorted kernels soon? | 13:34 |
ogra_ | oh, surprise, banshee failed on armhf | 13:37 |
ogra_ | infinity, do we still need the linux-libc-dev entry in the topic ? looks like it built | 13:39 |
=== infinity changed the topic of #ubuntu-arm to: Ubuntu ARMv7 Discussion & Development | https://wiki.ubuntu.com/ARM | Submit a Bug? https://bugs.launchpad.net/ubuntu/+filebug | Get oneiric while it's hot ! http://cdimage.ubuntu.com/releases/11.10/release/ | Logs at http://irclogs.ubuntu.com/ | armhf builds are running | ||
infinity | Nope. | 13:41 |
infinity | adconrad@cocoplum:~$ suite-diff.py /srv/launchpad.net/ubuntu-archive/ubuntu/dists/precise/main/binary-arm*/Packages.gz gt | wc -l | 13:41 |
infinity | 238 | 13:41 |
infinity | Only 238 binary packages in main different (or missing) compared to armel. | 13:41 |
ogra_ | yay | 13:41 |
ogra_ | and only 4800 jobs in the queue | 13:48 |
ogra_ | not bad | 13:48 |
=== cmagina_ is now known as cmagina | ||
doko_ | infinity, is this suite-diff command complete? | 14:31 |
infinity | doko_: The copy in ~adconrad/bin is. | 14:32 |
infinity | Well, FSVO "complete". It's a hackish fix of a hackish branch of Colin's. | 14:32 |
doko_ | good enough | 14:35 |
doko_ | infinity, why is apr rescored to -1? | 14:38 |
infinity | doko_: Testsuite that hangs on the buildds, if I recall. Was saving it for "later". | 14:39 |
infinity | (Or until I had a chance to look at it) | 14:39 |
doko_ | ok, starting a local build | 14:40 |
=== ericm is now known as ericm-Zzz | ||
jcrigby | infinity, tell me about armhf and u-boot, u-boot does no floating point | 15:03 |
jcrigby | infinity, I think I just answered my own question. U-boot should be built for both and go into both archives? | 15:04 |
infinity | jcrigby: Yup. | 15:06 |
infinity | jcrigby: Same story with kernel packages. | 15:06 |
infinity | jcrigby: But u-boot's our current blocker. | 15:06 |
doko_ | infinity, apr tests run fine here | 15:07 |
infinity | doko_: Whacky. Which kernel and hardware? | 15:08 |
infinity | doko_: But I'm happy to potentially eat a buildd to find out. | 15:08 |
doko_ | infinity, oneiric panda | 15:10 |
infinity | Kay. The buildds are natty panda, but we can try. I'm game. | 15:10 |
infinity | Score it up. | 15:10 |
kenws | Hi there, I noticed that the Packages.gz for armel on Oneiric has gcc-avr 4.3.5 while on x86 it's 4.5.3. The funny thing is that 1) avr-libc depends on gcc-avr 4.5.3 and 2) the gcc-avr 4.5.3 deb is actually available. | 15:16 |
kenws | http://paste.ubuntu.com/760527/ | 15:16 |
infinity | gcc-avr | 1:4.3.5-1 | oneiric/universe | armel | 15:17 |
infinity | gcc-avr | 1:4.5.3-2 | oneiric/universe | source, amd64, i386, powerpc | 15:17 |
ogra_ | looks like NBS | 15:17 |
kenws | infinity: yeah, but the avr-libc for armel depends on gcc-avr 4.5.3 | 15:18 |
infinity | kenws: The reason the deb is there is because it was built in precise. | 15:18 |
kenws | infinity: ok, I see. | 15:18 |
infinity | avr-libc is arch:all, hence the weird dep. | 15:18 |
kenws | aah, ok - interesting | 15:19 |
kenws | I haven't tried it but it looks like there is no easy way (without apt pinning) to get avr-libc installed on armel oneiric. | 15:20 |
infinity | That should have been tidied and removed before release. :/ | 15:20 |
infinity | Would be just as broken, but less confusing. | 15:20 |
ogra_ | yep | 15:20 |
infinity | You could always try running precise. ;) | 15:20 |
kenws | : ) | 15:20 |
ogra_ | armhf indeed :) | 15:20 |
kenws | dOh, wdo we have armhf packages for precise already? | 15:21 |
ogra_ | sure | 15:21 |
kenws | s/dOh/Oh/ | 15:21 |
kenws | cool | 15:21 |
infinity | No installers yet, and missing some desktop bits (and lots of universe), but we're getting there. | 15:21 |
ogra_ | send flowers to infinity | 15:21 |
ogra_ | looks good for FF if we dont hit a bad disaster now | 15:21 |
infinity | Yeah, I think we're going to be fine. | 15:23 |
infinity | It might actually be in better shape than armel soon. :P | 15:23 |
ogra_ | well, lets see how the build system behaves | 15:23 |
infinity | (Which won't hurt my feelings at all) | 15:23 |
ogra_ | i expect a bunch of changes to the subarch bits | 15:23 |
infinity | Which build system? | 15:23 |
ogra_ | (image builds i mean( | 15:23 |
kenws | so, the way for armhf is to deboostrap a rootfs or are there better ways? | 15:23 |
infinity | d-i's done, livecd-rootfs is done, live-build is done. | 15:23 |
infinity | All we need is uboot and kernels. | 15:24 |
ogra_ | kenws, use ubuntu-core | 15:24 |
ogra_ | infinity, cdimage and debian-cd | 15:24 |
infinity | I did cdimage. | 15:24 |
infinity | debain-cd is cake. | 15:24 |
infinity | I'll do it tomorrow. | 15:24 |
ogra_ | there are plenty areas weher we check for armel | 15:24 |
ogra_ | *where | 15:24 |
infinity | Like i said, cdimage is all done. :P | 15:24 |
kenws | ogra_: ok, thanks | 15:24 |
infinity | debian-cd won't take long. | 15:24 |
ogra_ | k | 15:24 |
xranby | infinity: debootstrap of armhf are ticking on nicely... thanks! | 15:25 |
kenws | will there be a precise release for the softfp abi as well? | 15:31 |
infinity | kenws: Yes. | 15:31 |
ogra_ | might be an unsupported one though | 15:31 |
infinity | kenws: But only one of them (probably armhf) will get official LTS support. | 15:31 |
ogra_ | which will get really funny wrt binary drivers | 15:32 |
infinity | Binary drivers are going to be fun. | 15:32 |
kenws | ok, so lets that the proprietaryGL libs will be available for hf as well ; ) | 15:32 |
kenws | yep | 15:32 |
* kenws is looking for the ubuntu-core root fs | 15:32 | |
infinity | We have some good relationships with some vendors who will (hopefully) DTRT there. | 15:32 |
infinity | But not all vendors. | 15:32 |
ogra_ | well, i think TI might be with us here if we can actually provide them a build and test env early enough | 15:32 |
ogra_ | but for others ... | 15:33 |
kenws | which abi a the binaries in the precise-core-armel.tar.gz? (http://cdimage.ubuntu.com/ubuntu-core/releases/precise/alpha-1/) | 15:33 |
ogra_ | dont take A1 | 15:33 |
ogra_ | take the last daily that built | 15:33 |
ogra_ | (and that should have the most recent abi) | 15:33 |
kenws | ok, so http://cdimage.ubuntu.com/ubuntu-core/daily/current/ | 15:33 |
ogra_ | right | 15:34 |
kenws | thanks | 15:34 |
infinity | ubuntu-core will actually be producing dailies again shortly. Just fixing the buildd right now. :P | 15:34 |
infinity | kenws: As for "which abi", armel is always softfp, armhf is always hard. That's the point of having two. ;) | 15:35 |
kenws | infinity: I agree : ). when I looked a the alpha-1 there was only one tarball and I didn't know how the binaries were built | 15:39 |
infinity | kenws: Yeah, I landed all this post-A1. ;) | 15:39 |
lilstevie | also, nvidia do have hf binary drivers for tegra | 15:39 |
infinity | lilstevie: \o/ | 15:39 |
ogra_ | yay, where ? | 15:39 |
lilstevie | trimslice developer forums | 15:40 |
infinity | lilstevie: Of course, if they also have executables with those, they won't run, I bet. | 15:40 |
infinity | lilstevie: But the libraries and drivers will work. | 15:40 |
lilstevie | they were built post a12r1 | 15:40 |
lilstevie | but they do have those daemon bits that aren't strictly needed anymore | 15:40 |
ogra_ | nice ! | 15:40 |
ogra_ | bah, sigh | 15:41 |
ogra_ | so not like the linux4tegra drop then | 15:41 |
infinity | Yeah, the daemon bits won't work. | 15:41 |
lilstevie | not quite | 15:41 |
infinity | We really need to push our linker changes upstream and get every toolchain on the same page. | 15:41 |
infinity | Like, yesterday. | 15:41 |
lilstevie | the daemons weren't running on Mer | 15:42 |
lilstevie | but using the drivers | 15:42 |
lilstevie | but in any case | 15:42 |
lilstevie | they use them for the trimslice on Mer, and Archlinux | 15:43 |
lilstevie | hf drivers that is | 15:43 |
Xase | lilstevie: any abstract thoughts on my problem? | 15:47 |
lilstevie | Xase: not yet, been a little busy | 15:47 |
Xase | No sweat, any knowledge anyone could provide is a great help. | 15:48 |
infinity | And we have Unity on armhf. Welcome to a (probably partially broken) desktop. | 15:51 |
lilstevie | infinity: heh | 15:52 |
lilstevie | nice | 15:52 |
lilstevie | I love comments on forums like "-it might be a good idea to test the binaries before running the script to make sure they don't segfault, since this seems to be a relatively common problem" | 15:56 |
lilstevie | :/ | 15:56 |
lilstevie | 1) I have had 0 reports of this, and have no idea where this 'common problem' is. 2) I give an email address for people to report bugs to; nobody has | 15:57 |
lilstevie | oh, and my favorite... 3) I use the stuff I do myself | 15:57 |
doko_ | ppisati, could you build linux-ti-omap4 for armhf please? | 16:02 |
ppisati | doko_: is the next item in my todo list | 16:03 |
infinity | \o/ | 16:03 |
doko_ | ppisati, thanks, including the mmap patch? | 16:10 |
ppisati | yep | 16:12 |
=== Guest76640 is now known as NCommander | ||
ndec | ogra_: hi. do you think i could get another PPA on tiomap-dev? something for 'experimental' stuff that we want ubuntu users to get access to easily? | 16:35 |
=== NCommander is now known as Guest67564 | ||
ogra_ | ndec, will have to talk to davidm about that, but he is traveling, how urgent is that ? | 16:36 |
=== Guest67564 is now known as NCommander | ||
ndec | not so many things in terms of build load, but i just need to make sure that we make a clear separation between what should work (release PPA) and our sandbox | 16:36 |
ndec | not too much urgent. | 16:36 |
ogra_ | ok | 16:36 |
ogra_ | i'll get an answer before the friday meeting | 16:36 |
ndec | good! | 16:37 |
GrueMaster | ppisati: Does this round of SRU kernels have the mmap fix? | 16:59 |
ppisati | GrueMaster: nope | 17:11 |
ppisati | GrueMaster: it entered the trees today | 17:11 |
GrueMaster | sigh. Why so late? It was tested a couple of weeks ago? | 17:11 |
GrueMaster | Now we have to wait for another round of SRU updates. | 17:12 |
infinity | Hey look, d-i succeeded. | 17:14 |
ogra_ | wohoo | 17:14 |
ogra_ | quick, lets build alternates :P | 17:14 |
GrueMaster | For? armhf??? | 17:15 |
infinity | I haven't done the debian-cd mangling yet. ;) | 17:15 |
infinity | GrueMaster: Yeah. | 17:15 |
ogra_ | yep :P | 17:15 |
GrueMaster | If only we had a kernel. | 17:15 |
ogra_ | well, netinst should work | 17:15 |
ogra_ | we do (if it finished yet) | 17:15 |
ogra_ | for omap at least | 17:15 |
infinity | We have omap. | 17:16 |
infinity | d-i wouldn't have built without it. | 17:16 |
ogra_ | ah, right | 17:16 |
ogra_ | u-boot is missing | 17:16 |
infinity | Is not. | 17:16 |
GrueMaster | omap? Cool. I can start testing it later today once my mirror has pulled it in. | 17:16 |
infinity | Again, d-i won't build without it. :P | 17:16 |
ogra_ | you fixed it ? | 17:16 |
ogra_ | oookay | 17:16 |
infinity | Just needed armhf added to the arch list. | 17:16 |
* ogra_ wants ac100 :P | 17:17 | |
infinity | doko did uboot and x-loader. | 17:17 |
ogra_ | i saw x-loader | 17:17 |
ogra_ | missed u-boot somehow | 17:17 |
infinity | I think he's just been looking at suite-diff.py output and adding armhf everywhere it wasn't. ;) | 17:17 |
ogra_ | heh | 17:17 |
infinity | I've been focusing on getting the desktop built. | 17:18 |
infinity | unity-2d done. | 17:18 |
infinity | Some ubuntuone stuff still to go. | 17:18 |
ogra_ | yeah, and that lenses thing for 3d | 17:18 |
infinity | Hey, we have a libinfinity. | 17:18 |
infinity | The file lens needs actual fixing. | 17:18 |
ogra_ | well, a bunch hasnt built yet for -desktop | 17:18 |
infinity | The desktop guys are on it. | 17:18 |
ogra_ | yep, saw that | 17:19 |
infinity | I don't know about "a bunch". We're only 153 packages short in main. | 17:19 |
doko_ | ogra_, all packages on armhf should be built, which also exist on armel (or are currently building ...) | 17:19 |
infinity | Binary packages, that is. | 17:19 |
ogra_ | http://qa.ubuntuwire.org/ftbfs/primary-precise-armhf.html shows 10 (one ftbfs) | 17:19 |
infinity | doko_: Redboot failed. Calling the compiler by ABI name. Not that it matters. | 17:19 |
doko_ | infinity, currently fixing | 17:20 |
ogra_ | for the ubuntu-desktop set | 17:20 |
doko_ | had to do libinfinity first ;) | 17:20 |
ogra_ | doko_, i wonder if we shouldnt just remove redboot | 17:20 |
GrueMaster | redboot fails? Oh no! Panic time. :P | 17:20 |
infinity | banshee's just waiting on ubuntuone stuff to sort out. | 17:20 |
ogra_ | nobody maintains it | 17:20 |
infinity | ogra_: We don't boot any subarches with it anymore. | 17:20 |
ogra_ | and nothing uses it anymore | 17:20 |
infinity | ogra_: Freescale moved entirely to u-boot. | 17:20 |
ogra_ | infinity, well, we probably kept it around for oem or something | 17:21 |
infinity | I'm game for removal. | 17:21 |
infinity | Or at least demotion. | 17:21 |
ogra_ | not sure | 17:21 |
ogra_ | huh | 17:21 |
GrueMaster | Only redboot platform we have is Babbage3, which is stuck on Lucid. | 17:21 |
ogra_ | it was demoted ages ago | 17:21 |
infinity | Or is it already in universe | 17:21 |
infinity | ? | 17:21 |
ogra_ | if its in main again thats a bug | 17:21 |
infinity | In that case, no idea why doko uploaded it. :P | 17:21 |
ogra_ | heh | 17:21 |
ogra_ | nostalgy :) | 17:21 |
infinity | redboot-imx | 200952-0ubuntu2 | precise | source | 17:22 |
infinity | ^-- Looks main to me. | 17:22 |
ogra_ | GrueMaster, the vodafone netbook is babbage "like" so it might have helped having around a redboot package to patch it | 17:22 |
infinity | Find where that's seeded and get rid of it. | 17:22 |
ogra_ | infinity, wow | 17:22 |
infinity | Actually, I will. | 17:22 |
ogra_ | it shouldnt be seeded | 17:23 |
ogra_ | weird | 17:23 |
infinity | ... wow. I still don't have the precise seeds locally? I fail at life. | 17:23 |
GrueMaster | ogra_: So is the Efika, but iirc, it is u-boot. | 17:23 |
infinity | platform.oneiric/installer: * redboot-tools-udeb [armel] | 17:23 |
infinity | Yeah, it's seeded. | 17:23 |
infinity | Now to fix it in precise. | 17:23 |
ogra_ | GrueMaster, right, i think the vodabook is actually very close to bbg | 17:23 |
ogra_ | so might use redboot | 17:23 |
GrueMaster | ENOCLUE. | 17:24 |
infinity | Freescale ditched redboot ages ago. | 17:24 |
infinity | Any newer imx51 stuff should be u-boot, from what I understand. | 17:24 |
ogra_ | infinity, yes, thats what i say, the HW is based on a board where no u-boot existed | 17:24 |
GrueMaster | I thought we were looking at switching bbg3 during Lucid also, but couldn't get it working. | 17:24 |
ogra_ | right | 17:24 |
infinity | Well, I'll keep it seeded then. | 17:24 |
ogra_ | code was immature and incomplete | 17:25 |
infinity | ogra_: Care to verify with our new PES overlords? :) | 17:25 |
ogra_ | infinity, no, drop it | 17:25 |
ogra_ | thats a 2009 source that nobody touched afterwards if people need the source package they should pull from lucid | 17:25 |
ogra_ | i dont think it changed since | 17:25 |
GrueMaster | If needed, universe isn't too far to see for the eye of Sauron. | 17:26 |
infinity | Yeah, I won't remove it, just demote it. | 17:26 |
infinity | Committed. | 17:28 |
* ogra_ just asked about the vidabook, if they dont use it i'll file a removal bug | 17:29 | |
ogra_ | *voda | 17:29 |
infinity | It does little harm being in universe. | 17:29 |
infinity | If some keener wants to do a home-brew precise setup on their babbage. | 17:29 |
ogra_ | hm | 17:30 |
ogra_ | as long as nobody bothers me about it (it might have my name all over the changelog) | 17:30 |
infinity | :P | 17:30 |
ogra_ | hmm, looks like the xubuntu seeds would like to know about armhf | 17:32 |
micahg | ogra_: indeed, but I was waiting for universe to be more complety built before doing that | 17:33 |
ogra_ | ah | 17:33 |
infinity | micahg: Hey, I spent some time unsnagging at least the core of xubuntu... | 17:35 |
micahg | infinity: I wasn't blaming anyone :) | 17:35 |
infinity | micahg: libxfcegui4 needs love, though. | 17:35 |
infinity | micahg: (hint, hint) | 17:35 |
micahg | infinity: doko fixed that I thought | 17:35 |
infinity | Oh, so he did. | 17:36 |
infinity | He's been rather nuts with the uploads today. | 17:36 |
ogra_ | argh | 17:36 |
infinity | And banshee's off! | 17:37 |
ogra_ | root@horus:/# exit | 17:37 |
ogra_ | ogra@horus:~/chroot-hf$ ps ax | 17:37 |
ogra_ | Killed | 17:37 |
infinity | ogra_: Reboot. | 17:37 |
ogra_ | no way to umount the /proc inside the chroot now i guess | 17:37 |
ogra_ | grr | 17:37 |
infinity | ogra_: You upgraded your chroot without having a no-op policy-rc.d in place. | 17:37 |
infinity | ogra_: And something broke your init. | 17:37 |
micahg | infinity: I can update the xubuntu seeds tomorrow for armhf if you think everything's close enough for it to work | 17:37 |
infinity | micahg: No huge rush. | 17:38 |
micahg | ok | 17:38 |
ogra_ | infinity, well, i upgraded without /proc or /sys mounted ... and only mounted them for configuring procps ... | 17:38 |
ogra_ | afterwards | 17:38 |
infinity | ogra_: And procps is what killed you, then. | 17:38 |
infinity | I bet. :P | 17:39 |
ogra_ | likely | 17:39 |
ogra_ | not init though | 17:39 |
* ogra_ crosses fingers and reboots | 17:39 | |
infinity | I should go do debian-cd now. | 17:41 |
infinity | We're closer to desktop images than I thought we would be. | 17:41 |
infinity | ogra_: debian-cd done. | 18:00 |
infinity | ogra_: If you want to review the latest revision (but I already merged it into production...) | 18:00 |
ogra_ | well, if its already merged i'll look at the fallout :P | 18:00 |
ogra_ | livecd-rootfs (or rather BuildLiveCD) dont seem happy at, | 18:01 |
ogra_ | *atm | 18:01 |
ogra_ | the failure mail i got looks like a quoting issue in BuildLiveCD | 18:01 |
ogra_ | Subject: LiveFS --f-omap4/ubuntu/armel+omap4 failed to build on 20111205 ... | 18:02 |
ogra_ | -f is usually for the filesystem call | 18:02 |
infinity | ogra_: But does the failure log itself look goofy? | 18:02 |
ogra_ | there is no content | 18:02 |
infinity | ogra_: Or just the subject line? | 18:02 |
infinity | Weird... | 18:02 |
* GrueMaster has become numb to these emails. | 18:03 | |
ogra_ | seems to have failed in buildLiveCd already | 18:03 |
infinity | Cause I have a log... | 18:03 |
infinity | http://people.canonical.com/~ubuntu-archive/livefs-build-logs/precise/ubuntu-omap4/20111205/livecd-20111205-armel.out | 18:03 |
ogra_ | GrueMaster, i only glance over them if they roll in ... that way such inconsistencies jump in my face | 18:03 |
infinity | ogra_: I've done literally nothing to BuildLiveCD that would lead to that. | 18:04 |
ogra_ | stramge | 18:04 |
infinity | Which host mailed that to you? | 18:05 |
ogra_ | nusakan | 18:05 |
ogra_ | indeed | 18:05 |
infinity | So, it's not from BuildLiveCD at all... | 18:06 |
ogra_ | if it starts with LiveFS it is from the buildlive run | 18:06 |
ogra_ | (the subject i mean) | 18:06 |
infinity | Yes, but not BuildLiveCD. | 18:06 |
ogra_ | well, thatrs just a subsequent call indeed | 18:07 |
ogra_ | buildlive isnt doing much else but running BuildLiveCD remotely | 18:07 |
ogra_ | but yeah, i indeed meant the execution of BuildLiveCD | 18:08 |
infinity | I blame Colin's revision. | 18:08 |
infinity | - mail -s "LiveFS $PROJECT/$DIST/$arch failed to build on $datestamp" \ | 18:08 |
infinity | + mail -s "LiveFS $PROJECT${SUBPROJECT:+-$SUBPROJECT}${subarch:+-$subarch}${UBUNTU_DEFAULTS_LOCALE:+-$UBUNTU_DEFAULTS_LOCALE}/$DIST/$arch failed to build on $datestamp" \ | 18:08 |
ogra_ | oh | 18:08 |
infinity | Though, how we're getting OPTIONS in there, I don't know... | 18:09 |
infinity | Oh, we're not. | 18:10 |
ogra_ | where does --f come from then ? | 18:10 |
* ogra_ doesnt see it | 18:10 | |
ogra_ | must come out of $PROJECT${SUBPROJECT:+-$SUBPROJECT} | 18:11 |
infinity | Did you get that mail more than once? | 18:11 |
ogra_ | nope, not yet | 18:11 |
infinity | I suspect it was a manual build with bad options on the CLI. | 18:11 |
ogra_ | ah | 18:11 |
infinity | Leading to the empty mail and the whacky subject. | 18:11 |
ogra_ | it would really help if it would publish the euid of the executing user :) | 18:12 |
ogra_ | that way we should be able to distinguish between sudoed CLI calls and cron runs :) | 18:12 |
ogra_ | SUDO_USER sshould be unset if it runs from cron, right ? | 18:15 |
ogra_ | that way we should be able to tag it as manual build, i think that would help | 18:17 |
infinity | cron runs as root, so yes. | 18:17 |
=== doko_ is now known as doko | ||
jhobbs | where is the most proper place to check for the status of an ubuntu package for arm? | 21:04 |
micahg | jhobbs: in most cases it should be the same as any other architecture, you can use rmadison in devscripts to query what the latest version per arch is of a package | 21:05 |
GrueMaster | jhobbs: You can also search launchpad.net for the package. | 21:06 |
GrueMaster | Just select the version "in Ubuntu". | 21:06 |
jhobbs | thanks | 21:07 |
infinity | ppisati: So, on pressure, but if we had a ti-omap4 kernel on armhf, we could build desktop CDs in about an hour (without libreoffice, it's still building... But close enough!) | 21:12 |
infinity | ppisati: s/on pressure/no pressure/ ;) | 21:12 |
* GrueMaster increases the pressure. | 21:15 | |
infinity | ubuntu-core/armhf dailies re-enabled, tomorrow we build desktop images. | 22:37 |
infinity | Or maybe tonight after I've napped. :P | 22:37 |
GrueMaster | W00T! | 22:38 |
doko | infinity, exo ftbfs on armhf, succeeds here locally ... | 23:04 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!