>It would be a good idea to avoid per-packet computation introduced by
>Andrea's new patch (i think there are a lot more people who are *not* MPI
>users). Could the problem be resolved by setting the sender MSS
My patch is not MPI specific. It only avoid you to break rfc1122. This is
at least my interpretation of rfc1122 and I don't know of more uptodate
specs that allow you to delay the ack for 2*MMS data.
BTW, since I am using my new patch the TCP/IP performances in my slow link
seems doubled. But maybe it's only that it's been increased the bandwith
somewhere else. I didn't tried yet to remove the patch and see if I was
continuing to go so fast. Note: some time ago I added a sysctl to kill
delacks completly and such thing was improving iteractive response too but
it was harming long donwloads since the TX line was busy sending tons of
acks. With my new patch instead the download performances are good as
before.
Andrea Arcangeli
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/