svn commit: r255245 - head/sys/kern

[Konstantin Belousov]

Author: kib
Date: Thu Sep  5 12:56:08 2013
New Revision: 255245
URL: http://svnweb.freebsd.org/changeset/base/255245

Log:
  The vm_pageout_flush() functions sbusies pages in the passed pages
  run.  After that, the pager put method is called, usually translated
  to VOP_WRITE().  For the filesystems which use buffer cache,
  bufwrite() sbusies the buffer pages again, waiting for the xbusy state
  to drain.  The later is done in vfs_drain_busy_pages(), which is
  called with the buffer pages already sbusied (by vm_pageout_flush()).
  
  Since vfs_drain_busy_pages() can only wait for one page at the time,
  and during the wait, the object lock is dropped, previous pages in the
  buffer must be protected from other threads busying them.  Up to the
  moment, it was done by xbusying the pages, that is incompatible with
  the sbusy state in the new implementation of busy.  Switch to sbusy.
  
  Reported and tested by:	pho
  Sponsored by:	The FreeBSD Foundation

Modified:
  head/sys/kern/vfs_bio.c

Modified: head/sys/kern/vfs_bio.c
==============================================================================
--- head/sys/kern/vfs_bio.c	Thu Sep  5 12:54:40 2013	(r255244)
+++ head/sys/kern/vfs_bio.c	Thu Sep  5 12:56:08 2013	(r255245)
@@ -3994,7 +3994,7 @@ vfs_drain_busy_pages(struct buf *bp)
 		m = bp->b_pages[i];
 		if (vm_page_xbusied(m)) {
 			for (; last_busied < i; last_busied++)
-				vm_page_xbusy(bp->b_pages[last_busied]);
+				vm_page_sbusy(bp->b_pages[last_busied]);
 			while (vm_page_xbusied(m)) {
 				vm_page_lock(m);
 				VM_OBJECT_WUNLOCK(bp->b_bufobj->bo_object);
@@ -4004,7 +4004,7 @@ vfs_drain_busy_pages(struct buf *bp)
 		}
 	}
 	for (i = 0; i < last_busied; i++)
-		vm_page_xunbusy(bp->b_pages[i]);
+		vm_page_sunbusy(bp->b_pages[i]);
 }
 
 /*