On Wed, 2006-08-16 at 19:40 +0400, Kirill Korotaev wrote:When I discussed this with Hugh Dickins on summit we agreed
--- ./include/linux/mm.h.kmemcore 2006-08-16 19:10:38.000000000
+0400
+++ ./include/linux/mm.h 2006-08-16 19:10:51.000000000 +0400
@@ -274,8 +274,14 @@ struct page {
unsigned int gfp_mask;
unsigned long trace[8];
#endif
+#ifdef CONFIG_USER_RESOURCE
+ union {
+ struct user_beancounter *page_ub;
+ } bc;
+#endif
};
Is everybody OK with adding this accounting to the 'struct page'? Is
there any kind of noticeable performance penalty for this? I thought
that we had this aligned pretty well on cacheline boundaries.
How many things actually use this? Can we have the slab ubcs withoutslab doesn't use this pointer on the page.
the struct page pointer?