<font face="arial" size="2"><p style="margin:0;padding:0;font-family: arial; font-size: 10pt; word-wrap: break-word;">I just read the first page of the paper so far, but it sounds like it is heading in a good direction.</p>
<p style="margin:0;padding:0;font-family: arial; font-size: 10pt; word-wrap: break-word;">It would be interesting to apply also to home access-point/switches, especially since they are now pushing 1 Gb/sec over the air.</p>
<p style="margin:0;padding:0;font-family: arial; font-size: 10pt; word-wrap: break-word;"> </p>
<p style="margin:0;padding:0;font-family: arial; font-size: 10pt; word-wrap: break-word;">I will put it on my very interesting stack.</p>
<p style="margin:0;padding:0;font-family: arial; font-size: 10pt; word-wrap: break-word;"> </p>
<p style="margin:0;padding:0;font-family: arial; font-size: 10pt; word-wrap: break-word;"> </p>
<!--WM_COMPOSE_SIGNATURE_START--><!--WM_COMPOSE_SIGNATURE_END-->
<p style="margin:0;padding:0;font-family: arial; font-size: 10pt; word-wrap: break-word;"><br /><br />On Wednesday, October 15, 2014 3:49pm, "Wes Felter" <wmf@felter.org> said:<br /><br /></p>
<div id="SafeStyles1413412741">
<p style="margin:0;padding:0;font-family: arial; font-size: 10pt; word-wrap: break-word;">> On 10/10/14, 7:52 PM, dpreed@reed.com wrote:<br />> > The best approach to dealing with "locking overhead" is to stop thinking<br />> > that if locks are good, more locking (finer grained locking) is better.<br />> > OS designers (and Linux designers in particular) are still putting in<br />> > way too much locking. I deal with this in my day job (we support<br />> > systems with very large numbers of cpus and because of the "fine<br />> > grained" locking obsession, the parallelized capacity is limited). If<br />> > you do a thoughtful design of your network code, you don't need lots of<br />> > locking - because TCP/IP streams don't have to interact much - they are<br />> > quite independent. But instead OS designers spend all their time<br />> > thinking about doing "one thing at a time".<br />> <br />> The IX project looks like a promising step in that direction, although<br />> it still doesn't support sub-core granularity like Linux does.<br />> <br />> https://www.usenix.org/conference/osdi14/technical-sessions/presentation/belay<br />> <br />> --<br />> Wes Felter<br />> IBM Research - Austin<br />> <br />> _______________________________________________<br />> Cerowrt-devel mailing list<br />> Cerowrt-devel@lists.bufferbloat.net<br />> https://lists.bufferbloat.net/listinfo/cerowrt-devel<br />> </p>
</div></font>