ksm: no debug in page_dup_rmap()
Hugh Dickins [Tue, 22 Sep 2009 00:01:59 +0000 (17:01 -0700)]
page_dup_rmap(), used on each mapped page when forking, was originally
just an inline atomic_inc of mapcount.  2.6.22 added CONFIG_DEBUG_VM
out-of-line checks to it, which would need to be ever-so-slightly
complicated to allow for the PageKsm() we're about to define.

But I think these checks never caught anything.  And if it's coding errors
we're worried about, such checks should be in page_remove_rmap() too, not
just when forking; whereas if it's pagetable corruption we're worried
about, then they shouldn't be limited to CONFIG_DEBUG_VM.

Oh, just revert page_dup_rmap() to an inline atomic_inc of mapcount.

Signed-off-by: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Signed-off-by: Chris Wright <chrisw@redhat.com>
Signed-off-by: Izik Eidus <ieidus@redhat.com>
Cc: Nick Piggin <npiggin@suse.de>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Avi Kivity <avi@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

include/linux/rmap.h
mm/memory.c
mm/rmap.c

index bf116d0..477841d 100644 (file)
@@ -71,14 +71,10 @@ void page_add_new_anon_rmap(struct page *, struct vm_area_struct *, unsigned lon
 void page_add_file_rmap(struct page *);
 void page_remove_rmap(struct page *);
 
-#ifdef CONFIG_DEBUG_VM
-void page_dup_rmap(struct page *page, struct vm_area_struct *vma, unsigned long address);
-#else
-static inline void page_dup_rmap(struct page *page, struct vm_area_struct *vma, unsigned long address)
+static inline void page_dup_rmap(struct page *page)
 {
        atomic_inc(&page->_mapcount);
 }
-#endif
 
 /*
  * Called from mm/vmscan.c to handle paging out
index 368561f..7a61a11 100644 (file)
@@ -597,7 +597,7 @@ copy_one_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm,
        page = vm_normal_page(vma, addr, pte);
        if (page) {
                get_page(page);
-               page_dup_rmap(page, vma, addr);
+               page_dup_rmap(page);
                rss[!!PageAnon(page)]++;
        }
 
index 1406e67..720fc03 100644 (file)
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -710,27 +710,6 @@ void page_add_file_rmap(struct page *page)
        }
 }
 
-#ifdef CONFIG_DEBUG_VM
-/**
- * page_dup_rmap - duplicate pte mapping to a page
- * @page:      the page to add the mapping to
- * @vma:       the vm area being duplicated
- * @address:   the user virtual address mapped
- *
- * For copy_page_range only: minimal extract from page_add_file_rmap /
- * page_add_anon_rmap, avoiding unnecessary tests (already checked) so it's
- * quicker.
- *
- * The caller needs to hold the pte lock.
- */
-void page_dup_rmap(struct page *page, struct vm_area_struct *vma, unsigned long address)
-{
-       if (PageAnon(page))
-               __page_check_anon_rmap(page, vma, address);
-       atomic_inc(&page->_mapcount);
-}
-#endif
-
 /**
  * page_remove_rmap - take down pte mapping from a page
  * @page: page to remove mapping from