=== tmpRAOF is now known as RAOF === ara is now known as Guest36049 [08:15] What's the recommended process to bisect on a rebased branch? I'm trying to track down a regression in PCI code between v3.15 and v3.16-rc6 on mainline, using the ubuntu-utopic branch which rebased from 3.15 to 3.16. The problem is the rebase causes a divergence back 578 commits to the common ancestor so bisect can't do anything [08:27] TJ-, we tend to acertain if the issue is present / not present in the corresponding mainline kernels at the same base level [08:27] if it is you can the bisect in th emainline which has no rebases [08:27] if ti doesn't exist then it is a matter of bisecting between the mainline base and the ubuntu tip [08:28] apw: I already know that, it came in with the PCI change added to v3.16; I was hoping to use the Ubuntu branch to do the bisect on rather than mainline [08:29] apw: the reason being I can use the Ubuntu packaging tools to distribute/install the test versions [08:30] well as you know procedurally that doesn't work, which is why indeed we make the mainlnie ones using the ubuntu tooling [08:30] and when bisecting you can do the same, the proceedure is essentially to "git checkout ubuntu/master-next debian debian.master" over the top of your bisect point, update the configs and then build [08:31] apw: Yes... I might try using the mainline repos as the base instead but when I tried that a few hours ago I got lost somehow with no debian/ debian.master after the bisect since those had to be applied from the packaging patches [08:32] Ahhhhh.... I'll play around with that approach === AceLan_ is now known as AceLan === mwenning is now known as mwenning-wfh [14:42] Hi, I'm getting Kernel Panic (Fatal Exception In Interrupt) about 30 seconds after I boot 15.04, can anyone help me? [14:54] smb: http://dep.debian.net/deps/dep5/ [14:54] smb: lib/librte_net/rte_ip.h [15:00] smb: https://lintian.debian.org/tags/package-name-doesnt-match-sonames.html === infinity1 is now known as infinity === shaderslayer is now known as shadeslayer === pgraner is now known as pgraner-afk [16:24] is there any automatic building of Ubuntu's trusty master-next? http://kernel.ubuntu.com/git/ubuntu/ubuntu-trusty.git/?h=master-next [16:32] gQuigs, it is uploaded into a PPA about every 24 hours if changed [16:33] apw: oh, cool.. I just haven't been able to find the PPA [16:34] where is it? [16:34] ppa:ubuntu-kernel-test/ubuntu/stable perhaps [16:34] www.lauchpad.net/~ubuntu-kernel-test [16:36] apw: that's it, thanks! === Trevinho is now known as Trevinho|Holiday === ming is now known as Guest91806 [20:52] i'm compiling kernel 3.16 gcc versions are 5 and 4.8, i've set CC=/usr/bin/gcc-4.8 still i get this error: fatal error: linux/compiler-gcc5.h: No such file or directory. any ideas? [21:02] swair, you'd have to be more specific as to how you are compiling, as that file definatly exists in v3.16 [21:03] apw: its not in the 3.16 tree. It was added in 3.18. [21:04] swair, ok it came in via stable v3.16.7 ... so sort of in 3.16 [21:05] it is therefore in ubuntu 3.16 based kernels [21:05] ok., though i got the tar from kernel.org/pub/linux/kernel/v3.x/linux-3.16.tar.gz. it wasn't there. [21:06] swair, but regardless, if we don't know how you are incanting at it, we are not going to be able to hel [21:06] help [21:07] apw: i had gcc5 installed. so i installed gcc4.8 as /usr/bin/gcc-4.8. Set CC to as that and just make [21:07] it might help most if you pastbin a transcript of your how you did that [21:08] and the errors etc it produces [21:08] as CC=x is one thing, export CC=x is another, and likely make CC=x is a third [21:08] i did: export CC=/usr/bin/gcc-4.8 [21:09] well, right now i fixed the issue by just copying compiler-gcc5.h from the 3.18 tree. [21:09] and it worked [21:10] swair, yep, what bit actually failed? as the kernel uses two compilers the host and target compilers, and CC only overrides one [21:10] should i have set HOSTCC as well? [21:11] maybe, but i keep saying tell me what error you got, and maybe we can tell [21:11] sorry, just a sec. [21:11] or we could ask and answer questions for 40 minutes, oh no i won't be here to do that [21:13] apw: http://pastebin.com/ifu80YdW [21:13] yeah thats HOSTCC ... set that too [21:14] yeah. so why two different compilers? hostcc and cc? [21:14] it is building host tooling to build bits of the build [21:14] the hostcc is building things to run on the building machine to make bits of the build [21:14] the cc is building bits for the machine which run it [21:14] ahh ok. thanks for the help! [21:14] when doing a cross-compile they are not the same [21:15] ok [22:24] is there an easy way to see the socket options on one of my processes? [22:24] i've looked around a while and can't see anything obvious