Re: [PATCH] bnx2x: fix panic when TX ring is full

["Dmitry Kravkov" <dmitry@broadcom.com>]

On Thu, 2012-06-21 at 18:25 +0200, Eric Dumazet wrote:
> On Thu, 2012-06-21 at 18:56 +0300, Dmitry Kravkov wrote:
> > On Thu, 2012-06-21 at 17:12 +0200, Eric Dumazet wrote:
> > > On Thu, 2012-06-21 at 15:19 +0300, Dmitry Kravkov wrote:
> > > 
> > > > The crash happens with default configuration since
> > > > [4acb41903b2f99f3dffd4c3df9acc84ca5942cb2] "net/tcp: Fix tcp memory
> > > > limits initialization when !CONFIG_SYSCTL", but it can be hit by
> > > > increasing values of tcp_wmem even earlier.
> > > 
> > > This makes no sense.
> > Bisected to this commit and reproduced before the commit only after:
> > echo "4096 16384 4194304" > /proc/sys/net/ipv4/tcp_wmem
> > by this max nr_frags raised from 8 to 17, when running 40 netperfs
> > 
> > i've decreased rx queue to 200, during the test
> 
> I repeat, this bug can be triggered anytime with a skb not provided by
> local tcp stack.
Got it
> 
> By the way, looking at your fix, its pretty obvious the fix has nothing
> to do with TCP stack.
It just describes the "4" number

> 
> commit changelog must be accurate, so please remove this wrong
> information. This will confuse future readers.
will do so
> 
> 
> 
>