Messages in this thread | | | Date | Thu, 10 Jul 2008 12:11:38 -0700 | From | Mike Travis <> | Subject | Re: [RFC 00/15] x86_64: Optimize percpu accesses |
| |
Christoph Lameter wrote: ... > The patches are reasonably small. The problem that Mike seems to have is early boot debugging.
Note that the early boot debugging problems go away with gcc-4.2.4. My problem now (what I [maybe incorrectly] believe) is stack overflow with NR_CPUS=4096 and a specific random config file.
Btw, I've completed the first half of splitting the zero_based_fold into zero_based_only and fold_pda_into_percpu and am testing that now.
| |