Ocfs2 devel - Aug 2008 - ocfs2-test: fix flock_unit_test for fcntl locks

We were expecting a failure for the same-node same-fd trylock tests when in
fcntl mode, but fcntl locks will actually succeed for this, while flock
locks won't. The solution is to check which locking mode we're using and
adjust our expected return accordingly.

Signed-off-by: Mark Fasheh <mfasheh at suse.com>

Index: ocfs2-test/programs/flock_tests/flock_unit_test.c
==================================================================---
ocfs2-test/programs/flock_tests/flock_unit_test.c	(revision 183)
+++ ocfs2-test/programs/flock_tests/flock_unit_test.c	(working copy)
@@ -329,7 +329,19 @@
 static void test_1node_1file(char *fname)
 {
 	int fd1, fd2;
+	int trylock_return = 0;
 
+	/*
+	 * flock allows one lock per fd, so a trylock on the 2nd file
+	 * descriptor will return EWOULDBLOCK.
+	 *
+	 * fcntl on the other hand, only allows one lock per process,
+	 * regardless of the number of fds used. So our trylocks here
+	 * should succeed as it'll be considered an EX->EX convert.
+	 */
+	if (use_flock)
+		trylock_return = EWOULDBLOCK;
+
 	fd1 = get_fd(fname);
 	fd2 = get_fd(fname);
 
@@ -344,13 +356,13 @@
 
 	info("Two exclusive trylocks\n");
 	lock_abort(fname, fd1, 0, 1, 1);
-	lock_abort(fname, fd2, EWOULDBLOCK, 1, 1);
+	lock_abort(fname, fd2, trylock_return, 1, 1);
 	unlock_abort(fname, fd1, 0);
 
 
 	info("One exclusive, one shared trylock\n");
 	lock_abort(fname, fd1, 0, 1, 0);
-	lock_abort(fname, fd2, EWOULDBLOCK, 0, 1);
+	lock_abort(fname, fd2, trylock_return, 0, 1);
 	unlock_abort(fname, fd1, 0);

On Wed, Aug 13, 2008 at 09:06:12PM -0700, Mark Fasheh
wrote:
> We were expecting a failure for the same-node same-fd trylock tests when in
> fcntl mode, but fcntl locks will actually succeed for this, while flock
> locks won't. The solution is to check which locking mode we're
using and
> adjust our expected return accordingly.
> 
> Signed-off-by: Mark Fasheh <mfasheh at suse.com>
Signed-off-by: Joel Becker <joel.becker at oracle.com>

> 
> Index: ocfs2-test/programs/flock_tests/flock_unit_test.c
> ==================================================================> ---
ocfs2-test/programs/flock_tests/flock_unit_test.c	(revision 183)
> +++ ocfs2-test/programs/flock_tests/flock_unit_test.c	(working copy)
> @@ -329,7 +329,19 @@
>  static void test_1node_1file(char *fname)
>  {
>  	int fd1, fd2;
> +	int trylock_return = 0;
>  
> +	/*
> +	 * flock allows one lock per fd, so a trylock on the 2nd file
> +	 * descriptor will return EWOULDBLOCK.
> +	 *
> +	 * fcntl on the other hand, only allows one lock per process,
> +	 * regardless of the number of fds used. So our trylocks here
> +	 * should succeed as it'll be considered an EX->EX convert.
> +	 */
> +	if (use_flock)
> +		trylock_return = EWOULDBLOCK;
> +
>  	fd1 = get_fd(fname);
>  	fd2 = get_fd(fname);
>  
> @@ -344,13 +356,13 @@
>  
>  	info("Two exclusive trylocks\n");
>  	lock_abort(fname, fd1, 0, 1, 1);
> -	lock_abort(fname, fd2, EWOULDBLOCK, 1, 1);
> +	lock_abort(fname, fd2, trylock_return, 1, 1);
>  	unlock_abort(fname, fd1, 0);
>  
>  
>  	info("One exclusive, one shared trylock\n");
>  	lock_abort(fname, fd1, 0, 1, 0);
> -	lock_abort(fname, fd2, EWOULDBLOCK, 0, 1);
> +	lock_abort(fname, fd2, trylock_return, 0, 1);
>  	unlock_abort(fname, fd1, 0);
>  
>  
> 
> _______________________________________________
> Ocfs2-tools-devel mailing list
> Ocfs2-tools-devel at oss.oracle.com
> http://oss.oracle.com/mailman/listinfo/ocfs2-tools-devel
-- 

"You must remember this:
 A kiss is just a kiss,
 A sigh is just a sigh.
 The fundamental rules apply
 As time goes by."

Joel Becker
Principal Software Developer
Oracle
E-mail: joel.becker at oracle.com
Phone: (650) 506-8127

Mark Fasheh wrote:
> We were expecting a failure for the same-node same-fd trylock tests when in
> fcntl mode, but fcntl locks will actually succeed for this, while flock
> locks won't. The solution is to check which locking mode we're
using and
> adjust our expected return accordingly.
>
> Signed-off-by: Mark Fasheh <mfasheh at suse.com>
>   
Signed-off-by: Sunil Mushran <sunil.mushran at oracle.com>