[ofa-general] IPOIB checksum offload support

[Eli Cohen]

On Sun, 2008-03-16 at 08:43 -0700, Roland Dreier wrote:
> > 1. It seems that there is a memory leak that the following fixes:
> 
>  > 2. I see that you added an else clause here:
> 
> Thanks... I fixed the memory leak and removed the else, and pushed out
> a new tree.
> 
I believe we still have a problem if don't set the pointer to NULL. If
one of the subsequent operations fails we will end up freeing the same
pointer twice.