[00:00] <dtchen> Darxus: not strictly necessary
[00:09] <JanC> Darxus: patching the binary package name & kernel/image file name would be nice though  ;)
[00:10] <JanC> s/patching/changing/
=== bjf is now known as bjf-afk
=== lamont` is now known as lamont
[01:33] <Darxus> It seems like changing the source package name to linux-bfs and the ABI to 12bfs1 is the cleanest way to go.
[02:56] <Darxus> Ugh, trying to figure out every place I need to change "linux" to "linux-bfs" is... hard.
[02:57] <Darxus> So tell me how it works for two source packages to have the same source package name in the official archives (one in main, and mine in universe)?
[03:02] <JanC> Darxus: that won't work
[03:02] <JanC> AFAIK
[03:03] <Darxus> JanC: That's what I thought.
[03:03] <Darxus> 05:29PM < Darxus> When patching the linux kernel package with bfs, should I  change the source package name from linux to linux-bfs?
[03:03] <Darxus> 07:00PM < dtchen> Darxus: not strictly necessary
[03:04] <JanC> Darxus: packaging for personal use vs. packaging for the Ubuntu repositories
[03:06] <JanC> it seems like they have a separate source package for the -rt kernels too
[03:07] <JanC> Maybe you should have a look at how this is done
[03:08] <Darxus> JanC: Thanks, I did look at it some, but... it's not a full patch of the linux source package.
[03:08] <Darxus> I've gotten it to build and install.  It works.  My only remaining problem is renaming it.
[03:09] <JanC> well, you could maybe ask the -rt maintainers
[03:09] <Darxus> 1-maintainer.mk:  @git diff-tree -p refs/remotes/linux-2.6/master..HEAD $(shell ls | grep -vE '^(ubuntu|$(DEBIAN)|\.git.*)')
[03:09] <Darxus> See, I don't even know if I need to change "linux" to "linux-bfs" on that line...
[03:09] <Darxus> JanC: Yeah, thanks.
[03:10] <Darxus> Looks like no.
[03:11] <Darxus> It's so frustrating to wait hours just to be informed that I missed another one.
[03:12] <JanC> Darxus: you might want to document your findings  :-)
[03:13]  * JanC goes to sleep now
[03:16] <Darxus> JanC: It has certainly occurred to me that the kernel package could use a patch to un-hardcode the source package name.
[03:19] <Darxus> It also doesn't help that it doesn't clean up the files it regenerates for every build, so I'm not sure which files I need to edit.
[03:30] <Darxus>          dpkg-source: error: source package has two conflicting values - linux-bfs-bfs and linux-bfs
[03:30] <Darxus> Hah.
[03:32] <Darxus> Fixed.
[03:32] <Darxus>                                                            
[03:32] <Darxus> Rejected:                                                                       
[03:32] <Darxus> linux_2.6.31-12bfs1.41.dsc: Version older than that in the archive.             
[03:32] <Darxus> 2.6.31-12bfs1.41 <= 2.6.31-12.41bfs1     
[03:32] <Darxus> Why must everything be designed for my misery?
[03:33] <Darxus> Do you think it would accept 2.6.31-12bfs1.42?
[03:37] <Darxus> I should go to sleep now, but I can't, because am too anxious for this build to... fail again.  In three hours.
=== Whoopie_ is now known as Whoopie
=== ericm_ is now known as ericm-afk
=== tjaalton_ is now known as tjaalton
=== ericm-afk is now known as ericm
[08:42] <Darxus> Wow.  It didn't fail.
[08:53] <darthanubis> https://bugs.launchpad.net/ubuntu/+source/lm-sensors-3/+bug/447837
[08:53] <ubot3> Malone bug 447837 in lm-sensors-3 "asus_atk0110 driver not enabled in kernel configuration by default" [Undecided,New] 
[09:46] <Darxus> The kernel package's debian/rules clean is missing "rm -r ignore-dups debian/d-i-i386/".
[10:43] <phenompanda> ericm, hi, have you seen this error?
[10:43] <ericm> phenompanda, which error?
[10:43] <phenompanda> kernel/time.c #error ICSA specification requires the logging of time changes. This architecture will not log changes.
[10:44] <phenompanda> this error will happen after i comments off the CONFIG_OABI_COMPAT
[10:45] <phenompanda> ericm, what is the ICSA specification BTW ?
[10:45] <ericm> no idea
[10:46] <phenompanda> em, seems report from compiler?
[10:46] <ericm> preprocessor
[10:46] <phenompanda> ericm, do you always build arm kernel only support EABI?
[10:47] <ericm> so you disable CONFIG_OABI_COMPAT?
[10:47] <phenompanda> ericm, ah, i see, preprocessor ... 
[10:47] <ericm> phenompanda, usually EABI + OABI_COMPAT
[10:47] <phenompanda> ericm, yeah, disabled CONFIG_OABI_COMPAT to squeeze couple of bytes from kernel :(
[10:47] <ericm> phenompanda, let me do a test build here
[10:48] <ericm> phenompanda, poor man - struggling for a few bytes
[10:48] <phenompanda> ericm, no idea, mmu less ... :(
[10:49] <ericm> phenompanda, will get the test build result back to you later - just upgraded to latest karmic - something's broken
[10:49] <phenompanda> ericm, ok, talk later ... hope it's a easy fix
[11:48] <ericm> phenompanda, test build has no prob with latest kernel - EABI without OABI_COMPAT
=== cjwatson_ is now known as cjwatson
[13:59] <tj83> g'morning all, I filed this bug automatically during hardware testing at ALF, It was just responded to, and I think it sould probably be a closed bug. Unsure If I should or someone else be the one to make it as invalid? Bug #433100
[13:59] <ubot3> Malone bug 433100 in linux "SD card not recognized and mounted upon insert" [Medium,Triaged] https://launchpad.net/bugs/433100
=== makke_ is now known as makke
=== Whoopie_ is now known as Whoopie
[18:28] <Darxus> Hah, I was trying to build a diff of only the changes involved in renaming the linux source package, and kept wondering why I was getting so much extra crap.  Then I realized I was diffing my modifications of -13 against a pristine -12.
[19:46] <wekt> what is the correct command to pull a kernel source package?    apt-get source linux-????????
[20:03] <Darxus> wekt: Yes.
[20:03] <Womble2> The main source package is just 'linux'
[20:03] <Womble2> whereas in Debian it's 'linux-2.6'
[20:05] <wekt> apt-get source linux         <--downloaded linux-meta package files.
[20:05] <Womble2> oh
[20:05] <Womble2> Ignore me, then
[20:06] <wekt> thanks for trying
[20:06] <wekt> Darxus: do you know what replaces the ???????
[20:08] <Womble2> 'apt-get source' should accept binary package names
[20:08] <Womble2> so, give it the image package name and you should end up with the source (but not necessarily the same version)
[20:10] <wekt> that is working with this attempt
[20:10] <wekt> apt-get source linux-image-2.6.31-7-rt
[20:19] <Darxus> wekt: Just curious, what are you doing with it?
[20:20] <wekt> I'm comparing the RT & Generic versions to see what is different
[20:22] <Darxus> Ah.  Good luck.
[20:23] <wekt> 889 different files.  That is quite a lot.
[20:27] <Darxus> Yeah.  The linux source package you're looking at builds just about every linux kernel related binary package except -rt.
[20:35] <Darxus> wekt: Do you know about the BFS patch?
[20:36] <wekt> no, i don't.  are you seeking info?
[20:36] <wekt> or think i am?
[20:39] <Darxus> wekt: http://ck.kolivas.org/patches/bfs/bfs-faq.txt
[20:39] <Darxus> wekt: I just think people who are playing with kernel packages are likely to find it interesting.
[20:40] <Darxus> I have karmic packages built in my ppa.
[20:40] <wekt> please continue to throw anything that fits that nature my way
[20:40] <Darxus> The history of the author's relationship with the kernel is interesting.
[20:42] <Darxus> Basically, the author, Con, wrote a couple of the kernel schedulers (the bits that decide which process to give cpu time to next) that were in the mainline kernel, and a bunch of other patches to improve desktop experience.  But the mainstream kernel guys wouldn't take his patches for the stuff that improved the desktop experience.  So after much frustration, he swore off it.
[20:42] <Darxus> Two years later, he's back, and he wrote BFS.
[22:11] <wekt> Darxus: you have a PPA of what? linux with the BFS patch?
[22:17] <Darxus> wekt: Yes.
[22:17] <Darxus> https://launchpad.net/~darxus/+archive/bfs
[22:18] <Darxus> Use the linux-bfs packages, not the linux packages.