[12:25] <zul> damn it
[12:26] <ajmitch> zul: ?
[12:26] <zul> still didnt boot
[06:34] <fabbione> BenC: ping?
[03:31] <BenC__> fabbione: pong
[04:45] <fabbione> BenC: hey dude
[04:45] <fabbione> BenC: bmc driver needs dma-mapping.h on sparc otherwise the latest git is FTBFS
[04:46] <BenC> yeah, pushed a fix
[04:46] <BenC> current git should build, I'm trying it, plus dave's central patch on the e3k now
[04:57] <fabbione> my Niagara doesn't fart anymore on unalligned access with current git
[04:57] <fabbione> tested today
[05:10] <BenC> I hate waiting
[05:10] <fabbione> waiting for what?
[05:11] <fabbione> play with glibc-2.5 in the meantime :)
[05:11] <BenC> spent yesterday extending my d-mark to the office, and wiring in the router...now I'm w..a..i..t..i..n..g
[05:11] <fabbione> i tested them here .. they work fine
[05:12] <zul> new conncection?
[05:12] <BenC> it's the T1...verizon cleared the line, and I got all the equipment hooked up on my end
[05:12] <zul> sweet
[05:13] <BenC> I just had to get an ISP that's -3 hours from me
[05:17] <zul> so i guess you wont be dropping as much
[05:17] <BenC> let's hop
[05:17] <BenC> e
[08:07] <jbailey> BenC: Around?
[08:07] <BenC> jbailey: yep
[08:07] <jbailey> BenC: fabbione mentioned that linux-libc-dev is broken in some sort of interesting way in 2.6.19-rc#
[08:07] <jbailey> Do you need an extra set of eyes on it?
[08:07] <BenC> I haven't had a chance to even look at it, but I will today
[08:08] <BenC> actually I can start right now
[08:08] <jbailey> Ah, cool.
[08:10] <BenC> jbailey: I'm inclined to think it's fixed with latest git, I am building mesa right now without problems
[08:10] <BenC> jbailey: Do you want me to upload i386/amd64/ppc/sparc/ia64/hppa linux-libc-dev packages somewhere today for you to use?
[08:11] <jbailey> BenC: That'd be lovely for a test run.  As long as it bootstraps fine, you also might want to just upload to the feisty repo right away.
[08:11] <jbailey> Since it would be good to make sure the toolchain can build with it.
[08:11] <BenC> jbailey: I have a bit of work still to do for the feisty kernel
[08:11] <BenC> I can upload, but I don't really want people using it
[08:12] <BenC> so I guess I can upload it without linux-meta
[08:12] <jbailey> Right.  I wouldn't change the meta packages to point to the newly uploaded kernel, but it would be good to have the headers.
[08:12] <jbailey> Anyone who's already pointed their sources.list to feisty is *asking* for this grade of headache.  Let'em have it ;)
[08:25] <BenC> jbailey: I think I can make an initial upload today
[08:25] <BenC> jbailey: You guys bootstrapping hppa this time around?
[08:27] <kylem> heh
[08:27] <jbailey> BenC: Yeah, should be.
[08:28] <jbailey> BenC: I checked on the parisc cabal channel, everything needed should be upstream for now to get us going.
[08:28] <jbailey> We have to twiddle binutils a bit for hppa, and then we'll have everything for toolchain at least.
[08:30] <kylem> there is no cabal.
[08:33] <zul> uh huh
[08:42] <fabbione> BenC: it's enough to install the headers and rebuild the kernel. stuff in scripts will fail or build the redhat-cluster-suite (known to break on i386)
[08:44] <BenC> fabbione: it's building fine for me, latest git
[08:45] <fabbione> BenC: ok. i will test tomorrow too just to make uber sure.
[08:45] <fabbione> jbailey: eheh
[08:45] <fabbione> BenC: otherwise i have a very simple test case
[08:46] <fabbione> #include <error.h>
[08:46] <fabbione> #include <sys/socket.h>
[08:46] <fabbione> int main() {
[08:46] <BenC> shoot me the test case and I'll let you know
[08:46] <fabbione>  return -EINVAL;
[08:46] <fabbione> return -ESOL_SOCKET;
[08:46] <fabbione> }
[08:46] <fabbione> that's it
[08:46] <fabbione> in my tests either EINVAL or SOL_SOCKET were undefined. no matter what include order
[08:47] <fabbione> the first one was always winning and the latter not included
[08:49] <BenC> (long paste)
[08:49] <BenC> #include <errno.h>
[08:49] <BenC> #include <sys/socket.h>
[08:49] <BenC> int main() {
[08:49] <BenC>  int i = EINVAL;
[08:49] <BenC>  int t = SOL_SOCKET;
[08:49] <BenC>  return t + i;
[08:49] <BenC> }
[08:49] <BenC> bcollins@gullible:~$ make test
[08:49] <BenC> cc     test.c   -o test
[08:49] <BenC> bcollins@gullible:~$ ./test 
[08:49] <BenC> bcollins@gullible:~$ echo $?
[08:49] <BenC> 23
[08:49] <BenC> seems to be working now
[08:49] <fabbione> ok that's fine
[08:49] <fabbione> it does build
[08:49] <fabbione> that's enough
[08:49] <fabbione> it was not even building before
[08:50] <fabbione> BenC: if you can put the linux-libc-dev somewhere on people, i can redo some build tests.. like rebuild glibc with new headers and stuff like that
[08:55] <BenC> fabbione: On chinstrap now
[08:56] <fabbione> BenC: ok thanks. i am not going to play today.. it's 9pm :) but i will test tomorrow
[08:56] <fabbione> movie time
[08:56] <fabbione> bbl
[08:57] <zul> oooh there is a hockey game on tonight...sweet
[08:58] <fabbione> hmm choccolate biscuits :)
[09:04] <jbailey> BenC: I don't suppose you have the ppc linux-libc-dev handy too? =)
[09:07] <BenC> jbailey: uploading
[09:07] <jbailey> BenC: Thanks.
[10:08] <jbailey> BenC, fabbione: confirmed on ppc32 & ppc64 as well with your linux-libc-dev, no problem.
[10:08] <BenC> excellent
[10:38] <jbailey> Bugger
[10:38] <jbailey> #define IFLA_RTA(r)  ((struct rtattr*)(((char*)(r)) + NLMSG_ALIGN(sizeof(struct
[10:38] <jbailey> ifinfomsg))))
[10:38] <jbailey> Got removed from the headers, but glibc expects it.
[11:32] <jbailey> BenC: http://marc.theaimsgroup.com/?l=linux-netdev&m=115960131615553&w=2 is a patch to re-add them.  These are needed in the headers to build glibc.
[11:33] <jbailey> I'm just trying to track if they're upstream yet, or if not why not.
[11:35] <BenC> jbailey: Ok, let me know what you find
[11:51] <jbailey> Oh, meh.  I misread the patch.
[11:51] <jbailey> Looks like they're patching iproute2 there, rather than the kernel headers.  Suck.