[PATCH 04/39] USB: fix needless failure under certain conditions

From: Oliver Neukum <[email protected]>

in devices.c we have a piece of code for dealing with losing in a race.
If we indeed lose the race we don't care whether our own memory allocation
worked. The check for that is so early that we return early even if we
don't have to.

Signed-off-by: Oliver Neukum <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
---
 drivers/usb/core/devices.c |    9 +++++----
 1 files changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/usb/core/devices.c b/drivers/usb/core/devices.c
index a47c30b..aefc798 100644
--- a/drivers/usb/core/devices.c
+++ b/drivers/usb/core/devices.c
@@ -604,10 +604,6 @@ static unsigned int usb_device_poll(struct file *file, struct poll_table_struct
 	lock_kernel();
 	if (!st) {
 		st = kmalloc(sizeof(struct usb_device_status), GFP_KERNEL);
-		if (!st) {
-			unlock_kernel();
-			return POLLIN;
-		}
 
 		/* we may have dropped BKL - need to check for having lost the race */
 		if (file->private_data) {
@@ -615,6 +611,11 @@ static unsigned int usb_device_poll(struct file *file, struct poll_table_struct
 			st = file->private_data;
 			goto lost_race;
 		}
+		/* we haven't lost - check for allocation failure now */
+		if (!st) {
+			unlock_kernel();
+			return POLLIN;
+		}
 
 		/*
 		 * need to prevent the module from being unloaded, since
-- 
1.5.0


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel