Re: [RFC][PATCH 2/9] deadlock prevention core

From: Peter Zijlstra
Date: Mon Aug 14 2006 - 04:34:21 EST

On Mon, 2006-08-14 at 12:25 +0400, Evgeniy Polyakov wrote:
> On Mon, Aug 14, 2006 at 10:15:52AM +0200, Peter Zijlstra (a.p.zijlstra@xxxxxxxxx) wrote:
> > > If this refers to the socket buffers, they're mostly allocated with
> > > at least __GFP_WAIT, aren't they?
> >
> > Wherever it is that packets go if the local end is tied up and cannot
> > accept them instantly. The simple but prob wrong calculation I made for
> > evgeniy is: suppose we have 64k sockets, each socket can buffer up to
> > 128 packets, and each packet can be up to 16k (roundup for jumboframes)
> > large, that makes for 128G of memory. This calculation is wrong on
> > several points (we can have >64k sockets, and I have no idea on the 128)
> > but the order of things doesn't get better.
> TCP memory is limited for all sockets - it is tcp_*mem parameters.
> tcp_mem max on my amd64 with 1gb of ram is 768 kb for _all_ sockets.

Yes, I've said as much a few emails back, but that does not make the
theoretical limit any lower.

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at
Please read the FAQ at