On Wed, 7 Jul 2010 14:26:18 +0200, Piotr Jaroszyński wrote > 2010/7/7 Michael Brown <[email protected]>: > > On Tuesday 06 Jul 2010 22:09:01 Piotr Jaroszyński wrote: > >> 2010/7/6 Piotr Jaroszyński <[email protected]>: > >> > While we are looking at tcp, there is also an access after free, which > >> > is maybe fixed in [1]. Maybe because I am lazy and didn't look at the > >> > RFC yet, so not sure whether moving the timestamp update is safe. > >> > > >> > [1] - > >> > http://git.etherboot.org/?p=people/peper/gpxe.git;a=commitdiff;h=979a414a > >> >37dd11f155933238fc5b6cb25a0646d8 > >> > >> Heh ok, it's obviously wrong because the seq changes, but at least > >> shows where the problem is :) > > > > Good catch! > > > > http://git.ipxe.org/ipxe.git/commitdiff/68c2f07 > > That was actually caught by valgrind ;) Running gpxe in usermode under > valgrind is part of my GSoC project, the code is at [1]. Just putting > it out there before I get around to writing a proper announcement :) > > P.S. What do you think about the FIN issue? Are you planning on > working on it or maybe have a general idea on how to best approach > fixing it? > > [1] - http://git.etherboot.org/?p=people/peper/gpxe.git;a=shortlog;h=refs/heads/valgrind > > -- > Best Regards > Piotr Jaroszyński
I've just done some work with "[tcp] Several TCP fixes" patch series. Testing and suggestions would be very appreciate. :) Guo-Fu Tseng _______________________________________________ gPXE-devel mailing list [email protected] http://etherboot.org/mailman/listinfo/gpxe-devel
