From: Larry Woodman <lwoodman@redhat.com>
Date: Thu, 24 Apr 2008 14:11:00 -0400
Subject: [mm] fix debug printks in page_remove_rmap()
Message-id: 1209060660.13895.14.camel@dhcp83-220.boston.redhat.com
O-Subject: [HEL5-U3 patch] Fix debug printk()s in page_remove_rmap()
Bugzilla: 457458
RH-Acked-by: Neil Horman <nhorman@redhat.com>
RH-Acked-by: Jeff Moyer <jmoyer@redhat.com>

We recently got a BZ for a BUG_ON() in page_remove_rmap(), the problem
is the page->_mapcount went negative.  I dont know what happened yet but
the code only prints out critical debug info if CONFIG_DEBUG_VM is yet
it unconditionally does "BUG_ON(page_mapcount(page) < 0);" !!!

-------------------------------------------------------------------
        if (atomic_add_negative(-1, &page->_mapcount)) {
#ifdef CONFIG_DEBUG_VM
                if (unlikely(page_mapcount(page) < 0)) {
                        printk (KERN_EMERG "Eeek! page_mapcount(page)
                        printk (KERN_EMERG "  page->flags = %lx\n",
                        printk (KERN_EMERG "  page->count = %x\n",
                        printk (KERN_EMERG "  page->mapping = %p\n",
                }
#endif
                BUG_ON(page_mapcount(page) < 0);
--------------------------------------------------------------------

This is ridiculous, the attached patch removes the
"#ifdef CONFIG_DEBUG_VM" and prints out the debug info since we are
going to BUG_ON() anyway.

Doesnt fix but helps debug BZ 443651.

diff --git a/mm/rmap.c b/mm/rmap.c
index b44fbdb..887c3df 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -579,15 +579,13 @@ void page_add_file_rmap(struct page *page)
 void page_remove_rmap(struct page *page)
 {
 	if (atomic_add_negative(-1, &page->_mapcount)) {
-#ifdef CONFIG_DEBUG_VM
 		if (unlikely(page_mapcount(page) < 0)) {
 			printk (KERN_EMERG "Eeek! page_mapcount(page) went negative! (%d)\n", page_mapcount(page));
 			printk (KERN_EMERG "  page->flags = %lx\n", page->flags);
 			printk (KERN_EMERG "  page->count = %x\n", page_count(page));
 			printk (KERN_EMERG "  page->mapping = %p\n", page->mapping);
+			BUG();
 		}
-#endif
-		BUG_ON(page_mapcount(page) < 0);
 		/*
 		 * It would be tidy to reset the PageAnon mapping here,
 		 * but that might overwrite a racing page_add_anon_rmap