On 12/31/06, Mitchell Blank Jr <[EMAIL PROTECTED]> wrote:
Looking at aligned_kmalloc() it seems to be pretty badly broken -- its fallback
if it gets a non-aligned buffer is to just try a larger size which doesn't
necessarily fix the problem. It looks like explicitly aligning the buffer
is a
From: Amit Choudhary <[EMAIL PROTECTED]>
Date: Sat, 30 Dec 2006 18:26:03 -0800
> Description: Fix infinite recursion when alignment passed is 0 in function
> aligned_kmalloc(), in file drivers/atm/firestream.c. Also, a negative value
> for alignment does not make sense. Check for negative value
From: Amit Choudhary [EMAIL PROTECTED]
Date: Sat, 30 Dec 2006 18:26:03 -0800
Description: Fix infinite recursion when alignment passed is 0 in function
aligned_kmalloc(), in file drivers/atm/firestream.c. Also, a negative value
for alignment does not make sense. Check for negative value too.
First, if you want to get patches merged you should send them to the subsystem
maintained (in this case Chas, who I've cc:'ed) Also if you feel it needs to
be sent to mailing list you should usually use a more specific list first
(like the ATM list or maybe netdev) Please see
Description: Fix infinite recursion when alignment passed is 0 in function
aligned_kmalloc(), in file drivers/atm/firestream.c. Also, a negative value for
alignment does not make sense. Check for negative value too.
The function prototype is:
static void __devinit
Description: Fix infinite recursion when alignment passed is 0 in function
aligned_kmalloc(), in file drivers/atm/firestream.c. Also, a negative value for
alignment does not make sense. Check for negative value too.
The function prototype is:
static void __devinit
First, if you want to get patches merged you should send them to the subsystem
maintained (in this case Chas, who I've cc:'ed) Also if you feel it needs to
be sent to mailing list you should usually use a more specific list first
(like the ATM list or maybe netdev) Please see
7 matches
Mail list logo