From: Benjamin LaHaise <[EMAIL PROTECTED]>
Date: Sat, 10 Dec 2005 10:58:13 -0500
> On Fri, Dec 09, 2005 at 03:50:52PM -0800, David S. Miller wrote:
> > Can you toss together such a patch or would you like me to do it? :-)
>
> Sure, it looks like there were 2 places where the test on sk_err was
>
On Fri, Dec 09, 2005 at 03:50:52PM -0800, David S. Miller wrote:
> Can you toss together such a patch or would you like me to do it? :-)
Sure, it looks like there were 2 places where the test on sk_err was
missing from the event wait logic (in sk_stream_wait_connect and
sk_stream_wait_memory), w
From: Benjamin LaHaise <[EMAIL PROTECTED]>
Date: Fri, 9 Dec 2005 17:53:02 -0500
> Similar to the previous patch, the atomic xchg in sock_error() results in
> extra pipeline flushes due to the need to perform full synchronization as
> a memory barrier. Avoid calling sock_error() in a couple of p
Similar to the previous patch, the atomic xchg in sock_error() results in
extra pipeline flushes due to the need to perform full synchronization as
a memory barrier. Avoid calling sock_error() in a couple of places where
it is safe to do a racy version of the test. Combined with the refcountin
