On Tuesday 21 March 2006 15:17, Eric Dumazet wrote:

> This is a new point of failure for x86 machines that use lot of tcp sockets, 
> I 
> learnt it the bad way and had to revert to 2.6.14 some servers that cannot 
> run 
> stock 2.6.15/2.6.16 for long because of this problem.

x86-64/ppc64/other 64bit with 4k page machines always had this problem but I'm 
not aware of 
people hitting it in practice.

> 
> Of course, we might argue the problem come from linux memory management...
> Oh well...

I'm sure there are ways to put struct sock / tcp_sock to a diet. From a quick 
look a 
lot of ints could be shorts, the function pointer callbacks could be moved
out to a different structure.

Maybe all those new hints recently added could be moved into a separate 
structure
and only be allocated if the socket has a large window.

Some fields only used by LISTEN sockets (e.g. last_synq_overflow) could be 
unioned
with ESTABLISHED data. Probably more opportunities for this.

etc.etc.

-Andi
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to