[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: Bug: main.c:344: connection_stop_writing



On Thu, Oct 11, 2007 at 05:59:20AM +0200, Olaf Selke wrote:
> Roger Dingledine wrote:
> > On Wed, Oct 10, 2007 at 09:41:05PM +0200, Olaf Selke wrote:
> >>
> >> Is this a known bug?
> >>
> > Yes, but it isn't fixed:
> > http://bugs.noreply.org/flyspray/index.php?do=details&id=451
> > 
> > If you can help fix it, that would be grand. :)

FWIW, I think I've just fixed this.  If the bug recurs in
0.2.0.8-alpha or later, please let us know.  For the gory details, see
the last two comments on bug 451 (link above).  The patch was r11882.

> oooh, better I don't since my coding capabilities are very limited
> :-) On my wishlist is a fix for the memory leak which lets the tor
> process grow to about 2GB workspace after two weeks

Wow, that's a lot!  We've fixed a few memory leaks recently; let us
know if 0.2.0.8-alpha is better.

> and better multiprocessor support.

That will probably have to wait for 0.2.1.x or later; right now, we
use extra CPUs for RSA, and not much else.  The logical next step is
to use them for AES too, but that will require a bit of refactoring in
how we do our circuit crypto.

(Anybody who feels like hacking on this, let me know.)

yrs,
-- 
Nick Mathewson

Attachment: pgpLz8RYf37ez.pgp
Description: PGP signature