[llvmlinux] Linux-v3.10 is bootable here
dl9pf at gmx.de
Fri Jul 12 12:22:52 UTC 2013
True - was half-way through and got side-tracked.
So here's the full info:
git show c208c43
Author: Michael Kuperstein <michael.m.kuperstein at intel.com>
Date: Tue May 7 14:05:33 2013 +0000
Re-enable AVX detection on x64 platforms.
git show acccd2edc
Author: Preston Gurd <preston.gurd at intel.com>
Date: Tue May 7 19:57:34 2013 +0000
Corrected Atom latencies for SSE SQRT instructions.
jansimon.moeller at gmx.de
On Friday 12 July 2013 14:17:21 Sedat Dilek wrote:
> On Fri, Jul 12, 2013 at 2:11 PM, Jan-Simon Möller <dl9pf at gmx.de> wrote:
> > I went through the log and think these are candidates:
> > c208c43a8ac5bfd3ee09b048cbc3f760691cefad - !!!
> > acccd2edc8d4c078aa03c4dd43ef815087176ef9
> From my experiences, it is very helpful for interested people when you
> always add the "subject" of the commit together with the commit-id.
> Are those commits in llvm-git or clang-git?
> - Sedat -
> > --
> > Dipl.-Ing.
> > Jan-Simon Möller
> > jansimon.moeller at gmx.de
> > On Friday 12 July 2013 12:50:01 Jan-Simon Möller wrote:
> >> > You helped me a lot with giving a "working startpoint", otherwise my
> >> > experiments have been "endless".
> >> >
> >> > Hope I could give sth. back.
> >> You could help us to track down why llvm/clang HEAD does not boot but 3.3
> >> does. If you have a spare box, you could bisect it.
> >> So far we know it happens during bootup using sse/mmx (which should be be
> >> used at this point). So changes to the x86 sse/mmx handling can be the
> >> culprit.
> >> Best,
> >> JS
> >> _______________________________________________
> >> LLVMLinux mailing list
> >> LLVMLinux at lists.linuxfoundation.org
> >> https://lists.linuxfoundation.org/mailman/listinfo/llvmlinux
> LLVMLinux mailing list
> LLVMLinux at lists.linuxfoundation.org
More information about the LLVMLinux