[09:59] <columbobaas> Anyone a perfect tuturial for compiling to the latest kernel? Always black screen after reboot now. 
[10:00] <columbobaas> (after compiling)
[10:41] <apw> columbobaas, what do you mean by the latest kernel, and what are you compiling for
[10:42] <columbobaas> The latest kernel has a good AMD support
[10:42] <columbobaas> apw: so I want to try it. 
[10:42] <apw> yes but where are you getting it, which latest kernel source are you referring to
[10:43] <apw> and what h/w platform is it
[10:43] <columbobaas> x86_64
[10:44] <ohsix> 'perfect' 'tutorial'
[10:44] <ohsix> a tutorial is a list of things to do, and that falls short of documentation which still isn't perfect
[10:44] <columbobaas> ohsix, well documentation about fixing x then
[10:44] <columbobaas> after compiling
[10:45] <apw> columbobaas, i am trying to find out where the source you are trying to compile when you say "latest"
[10:45] <apw> is this latest ubuntu kernel, latest upstream, latest random vendor kernel, what
[10:45] <apw> what level of that are you trying to find, as we have kernels at different levels
[10:45] <apw> what release are you trying to make it work on, etc etc
[10:45] <apw> "how make latest work" isn't enough to help really
[10:46] <ohsix> nor is 'black screen'
[10:46] <columbobaas> ohsix, x not loading
[10:46] <ohsix> literally hundreds of kconfig options can lead to that, you should probably figure out how to build the ubuntu kernel first and diverge slowly
[10:47] <apw> columbobaas, for example is what you really want is a 4.2 based kernel, there is a 4.2 based ubuntu modified and configured kernel in the CKT PPA
[10:47] <apw> if you want a pure upstream kerenl, we produce builds of those in the mainline build archive
[10:47] <apw> which are configured with the ubuntu config
[10:47] <columbobaas> thank you for the ckt ppa tip
[10:48] <apw> columbobaas, that is an early preview of the next wily kernel
[10:48] <apw> which if you are after the very latest support might be an option, trying a wily daily CD and seeing if it works
[10:48] <apw> if so upgrading there might help
[10:48] <apw> but ... without knowing more about what you are referring to, its all guess work
[10:49] <columbobaas> yes, but I learn
[10:50] <columbobaas> I tried a lot of distros, but ubuntu is the most solid. I compiled the kernel in Debian and now i want to try it in Ubuntu
[10:52] <apw> the most important thing is to make sure you have installed linux-image-* and linux-image-extra-* if you are building ubuntu kernels
[10:52] <apw> else you have only half the kernel (the bit a virtual machine needs) not the bit a machine with displays etc need
[10:52] <columbobaas> but is it really with .deb files? I found it really strange. 
[11:03] <apw> columbobaas, "it" ?  debian builds produce .deb's, ubuntu is a debian/apt based distro so our builds produce debs
[11:04] <columbobaas> apw: on Debian i downloaded the latest kernel from kernel.org and used to make command
[11:04] <columbobaas> the make command*
[11:05] <apw> you can do that, but it dumps files into your system that you cannot easily find to remove
[11:05] <apw> which is why you normally use .debs to contain them
[11:05] <apw> most likely your issue is configuration then, and i'd use the one in /boot for the nearest version of an ubuntu-kernel
[11:06] <columbobaas> Thanks for that explenation
[11:06] <apw> or you could use the mainline builds which do the same thing for you
[13:19] <tseliot> apw: problem solved (with flush_workqueue): https://lkml.org/lkml/2015/8/4/775
[13:34] <rtg> tseliot, not until 4.3 - in the meantime I've committed it as a SAUCE patch on Wily master-next
[13:34] <tseliot> rtg: thanks
[13:34] <tseliot> that will make my life much easier
[13:36] <rtg> tseliot, I've verified that linux 4.2-3.3 fixes build problems with bcmwl. I also uploaded bcmwl yesterday to fix a different FTBS
[13:37] <tseliot> rtg: nice, I hadn't noticed the failure
[13:38] <tseliot> I still have to fix one more thing for fglrx I think, but I'm busy with other work. I promise to do that soon though
[14:13] <apw> tseliot, heh yeah, that was tim :)
[14:17] <ricotz> apw, hi, is 4.2.0-3 already scheduled to land in the archive?
[14:18] <apw> ricotz, we're working on getting it to feature paritity with 4.1 before so
[14:18] <ricotz> apw, alright, got a bit scared ;)
[14:18] <apw> ricotz, heh why ?
[14:19] <ricotz> apw, seems a bit early for my taste even it is rc5 already
[14:21] <apw> ricotz, oh devel is always the wild west
[14:21] <apw> we've got gcc5, what could be more wild than that
[14:21] <ricotz> of course but the kernel is another thing
[14:21] <ricotz> right
[14:22] <apw> ricotz, the kernel is a unique snowflake in the sense you get to keep older ones :)
[14:23] <apw> so one could claim that being more caviliere is ok, but anyhow its pending some DKMS fixes for sure
[14:23] <apw> and likley by the time those kinks are worked out we'll be at a much later rc or even release upstream
[14:23] <ricotz> yeah, although while the filesystem gets crumbled into pieces there is no older kernel anymore ;)
[14:26] <ricotz> of course still looking forward for a new shiny kernel
[14:27] <apw> it passes some pretty hard filesystem testing before it gets out of our PPA, and most of us run it for some time before as well
[15:11] <arges> sforshee: hey bug 1476900, did this get uploaded into wily
[15:16] <sforshee> arges: I think so, let me double check
[15:17] <arges> sforshee: just not marked in the bug, wanted to verify before accepting into proposed
[15:17] <sforshee> arges: yeah, the last upload did include it, I didn't think to add the bug number
[15:19] <arges> sforshee: ok i'll mark it fix release then. thakns
[15:20] <sforshee> arges: already did it
[15:20] <arges> sforshee: cool