[PATCH 3/3] splice: fix misuse of SPLICE_F_NONBLOCK

[Miklos Szeredi <miklos@szeredi.hu>]

From: Miklos Szeredi <mszeredi@suse.cz>

SPLICE_F_NONBLOCK is clearly documented to only affect blocking on the
destination pipe and not on the source file.

In __generic_file_splice_read(), however, it returns EAGAIN if the
page is currently being read.

This makes it impossible to write an application that only wants
failure if the pipe buffer is full.  For example if the same process
is handling both ends of a pipe and isn't otherwise able to determine
whether a splice to the pipe will fit or not.

We could make the read non-blocking on O_NONBLOCK or some other splice
flag, but for now this is the simplest fix.

Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
---
 fs/splice.c |   12 +-----------
 1 file changed, 1 insertion(+), 11 deletions(-)

Index: linux-2.6/fs/splice.c
===================================================================
--- linux-2.6.orig/fs/splice.c	2010-06-29 12:46:30.000000000 +0200
+++ linux-2.6/fs/splice.c	2010-06-29 12:46:32.000000000 +0200
@@ -399,17 +399,7 @@ __generic_file_splice_read(struct file *
 		 * If the page isn't uptodate, we may need to start io on it
 		 */
 		if (!PageUptodate(page)) {
-			/*
-			 * If in nonblock mode then dont block on waiting
-			 * for an in-flight io page
-			 */
-			if (flags & SPLICE_F_NONBLOCK) {
-				if (!trylock_page(page)) {
-					error = -EAGAIN;
-					break;
-				}
-			} else
-				lock_page(page);
+			lock_page(page);
 
 			/*
 			 * Page was truncated, or invalidated by the