is_locked optimization

[Nadav Danieli]

Oops, sorry, forgot to attach the diff :-)

On Fri, 2004-12-03 at 00:09, Nadav Danieli wrote:

> Hello all,
> 
> When reading/writing file with locking and strict locking (default)
> enabled, is_locked is called. I have very little knowledge about tdb
> overhead, but it seems that if oplock is granted, unnecessary (to me,)
> to call brl_locktest.
> Attached is a suggested change for is_locked in cases an oplock is
> granted.
> I would like to hear any opinion regarding this change, and of course,
> if it is wrong.
> 
> Regards,
> 
> Nadav Danieli
-------------- next part --------------
--- source/locking/locking.c	2004-11-16 05:03:30.000000000 +0200
+++ source/locking/locking.new.c	2004-12-02 22:08:49.000000000 +0200
@@ -80,9 +80,23 @@
 	if (!lp_locking(snum) || !lp_strict_locking(snum))
 		return(False);
 
-	ret = !brl_locktest(fsp->dev, fsp->inode, fsp->fnum,
-			     global_smbpid, sys_getpid(), conn->cnum, 
-			     offset, count, lock_type, check_self);
+	do {
+		if (!check_self) {
+			if (EXCLUSIVE_OPLOCK_TYPE(fsp->oplock_type) &&
+					(lock_type == READ_LOCK || lock_type == WRITE_LOCK)) {
+				ret = 0;
+				break;
+			}
+			if (LEVEL_II_OPLOCK_TYPE(fsp->oplock_type) &&
+					(lock_type == READ_LOCK)) {
+				ret = 0;
+				break;
+			}
+		}
+		ret = !brl_locktest(fsp->dev, fsp->inode, fsp->fnum,
+					 global_smbpid, sys_getpid(), conn->cnum, 
+					 offset, count, lock_type, check_self);
+	} while ( 0 );
 
 	DEBUG(10,("is_locked: brl start=%.0f len=%.0f %s for file %s\n",
 			(double)offset, (double)count, ret ? "locked" : "unlocked",