=== fabbione [i=fabbione@gordian.fabbione.net] has joined #ubuntu-toolchain | ||
=== Keybuk [n=scott@quest.netsplit.com] has joined #ubuntu-toolchain | ||
=== rodarvus [n=rodarvus@ubuntu/member/rodarvus] has joined #ubuntu-toolchain | ||
=== mdz [n=mdz@studiocity-motorola-bsr1-70-36-194-85.vnnyca.adelphia.net] has joined #ubuntu-toolchain | ||
=== doko [n=doko@dslb-088-073-103-221.pools.arcor-ip.net] has joined #ubuntu-toolchain | ||
=== kintaro [n=ad0lf@203.177.212.164] has joined #ubuntu-toolchain | ||
=== kintaro [n=ad0lf@203.177.212.164] has joined #ubuntu-toolchain | ||
=== rodarvus [n=rodarvus@ubuntu/member/rodarvus] has joined #ubuntu-toolchain | ||
doko | jbailey: please could you temporarily revert the linux-kernel-headers change, moving ia64 and hppa headers to the multiarch location. gcc currently doesn't include that directory on non biarch architectures, but I cannot bootstrap it anymore. | 03:28 |
---|---|---|
jbailey | doko: Sure, np. How long do you need it for? | 03:57 |
doko | jbailey: until it's enabled in all compilers ... so may take a while | 03:58 |
jbailey | Hmm. Is it hard to do? I would like this in before feature freeze - it's part of what we spec'd in the roadmap as an optional component. | 04:04 |
doko | jbailey: hard, or not, we need to get at least the default gcc bootstrapping again. | 04:38 |
jbailey | Eh? I think you misunderstood my question. How hard is it to multiarch those architectures to look in those paths? | 04:48 |
jbailey | Like, if I upload lkh today, and you upload a new gcc today, can I upload the old lkh tomorrow? | 04:48 |
doko | jbailey: ohh, yes, it's easy, except that we're frozen for the CD's, and powerpc should be built first. I have to look at gcc-3.4 and gcc-4.0 as well. need to do something else done than the toolchain as well | 04:50 |
jbailey | 'k | 04:50 |
doko | or have infinity build the gcc's with an old lkh | 04:51 |
infinity | That could be done, but eww. | 04:51 |
infinity | Best to just have Jeff revert the change and do it right. | 04:51 |
jbailey | Your call. Either way, I won't be looking at this for a couple hours. | 04:51 |
infinity | gcc-4.1 on PPC will be fine by tomorrow. | 04:52 |
jbailey | You're welcome to do the revert yourself now if you need. | 04:52 |
jbailey | Just change the installed path in the rules file to go to /usr/include for linux/ and the main asm/ directory. | 04:52 |
jbailey | The biarch systems should still prefer the biarch asm directories. | 04:52 |
jbailey | (linux is already in the common one, I think.. Just the asm/ needs to be moved) | 04:52 |
jbailey | fabbione: sparc64 headers bug in lkh is amusing, looks like the problem is a Kbuild one: | 05:40 |
jbailey | <Clint> -unifdef-y := fbio.h perfctr.h | 05:40 |
jbailey | <Clint> +unifdef-y += fbio.h perfctr.h\ | 05:40 |
jbailey | =) | 05:40 |
fabbione | whops... | 05:45 |
fabbione | same as ppc? | 05:45 |
jbailey | Hmm? | 06:24 |
jbailey | this isn't your mesa failure. | 06:24 |
jbailey | Oh wait, you didn't point that one out to me. | 06:24 |
jbailey | Keybuk: ^^ | 06:24 |
jbailey | =) | 06:24 |
Keybuk | right | 06:28 |
doko | jbailey, fabbione: hmm, but the gcc-4.1 build failure on sparc is the same? | 06:28 |
Keybuk | sparc64 was one failure | 06:28 |
Keybuk | mesa's failure is still the same symptom (missing asm/errno.h) but a different problem | 06:29 |
Keybuk | ergo gcc not looking in /usr/include/ia64-linux-gnu on ia64 | 06:29 |
Keybuk | (on ia64, at least) | 06:29 |
doko | Keybuk: on all non-biarchs, so it should fail on hppa as well | 06:30 |
Keybuk | doko: yes, but we don't care about hppa | 06:31 |
Keybuk | because hppa isn't on our build system | 06:31 |
doko | why not, I thought we had the machines in the data center? | 06:31 |
Keybuk | they don't get builds | 06:31 |
Keybuk | (I don't know _why_ they don't get builds, they just don't) | 06:32 |
fabbione | jbailey: sorry you lose me than... | 07:00 |
fabbione | doko: edgy/hppa is waiting glibc ntpl or the world will chease to exist in one ms | 07:01 |
jbailey | Keybuk: They'renot building atm because they didn't meet the initial upload requirements. They have no functioning glibc 2.4 with NPTL. | 07:46 |
jbailey | Keybuk: Architectures that can't meet basic toolchain requirements (current kernel, gcc, gdb, glibc) get a fairly serious no-sympathy treatment. | 07:47 |
jbailey | I'm working with the parisc folks to get a workign nptl for parisc, though. | 07:47 |
jbailey | I'm trying very hard to get that arch up and running and working by Feature Freeze. | 07:47 |
Keybuk | jbailey: right, I just mean I don't understand how they're disabled from an LP point of viedw | 07:49 |
Keybuk | I knew they were deliberately down | 07:50 |
jbailey | Oh, yeah. No idea. =) | 08:03 |
jbailey | I think that they just don't have chroots registered for edgy. | 08:03 |
Keybuk | then they'd be waitingchroot, no? | 08:40 |
Keybuk | though they do not have a chroot, you're right | 08:40 |
=== jbailey_ [n=jbailey@modemcable139.249-203-24.mc.videotron.ca] has joined #ubuntu-toolchain |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!