On Fri, Mar 14, 2008 at 04:56:26PM -0700, Luck, Tony wrote: > Perhaps we should leave mca.c using alloc_pages_node() > for cpu!=0 and just have the mm/{contig.c,discontig.c} > do a pre-allocation for cpu0 only. In contig.c you'd > just use the !SMP version of alloc_mca_data() for all > cases. Changes to discontig.c are left as an exercise > to the reader. Patch for mca.c looks like this: I fixed up the -V7 patch as, what I intended to be, -V8 RFC, but I rushed the send. Could you test that patch instead? It does keep the mca stacks being allocated by the early arch init instead of mca.c. Thanks, Robin -- To unsubscribe from this list: send the line "unsubscribe linux-ia64" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
![]() |