![]() | |||||||||||||
Messages in this thread |
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote: > > On Sat, 25 Feb 2006 15:22:18 +0900 > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote: > > > On Fri, 24 Feb 2006 22:16:51 -0800 > > Andrew Morton <akpm@osdl.org> wrote: > > > > > > Some of these things must generate a large amount of code - would you have > > > time to look into uninlining them, see what impact that has on .text size? > > > > Okay, I'll do soon on ia64. > > > I compared inlined and out-of-lined vmlinux on ia64 NUMA config kernel. > > inline out-of-line > .text 005c0680 005bf6a0 > > 005c0680 - 005bf6a0 = FE0 = 4Kbytes. > > Considering the usage of this loop, above size looks big ;) Yes, we inline too many things. Still. > I attaches a patch which makes pgdat_walk funcs out-of-line. Looks fine to me. > I'll rewrite this if necessary. > (make this patch depends on some config or move the place of funcs...) We wouldn't want a config option for it. And the new mmzone.c probably makes sense too - I expect there are a few related things (page_alloc.c) which could be moved there. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ | ||||||||||||
| Last update: 2006-02-25 07:33 [from the cache] ©2003-2008 | |||||||||||||