Re: something to experiment with: LIFO and accept()

David S. Miller (
Mon, 7 Apr 1997 01:59:06 -0400

Date: Sun, 6 Apr 1997 22:37:35 -0700 (PDT)
From: Dean Gaudet <>

It's my suspicion that if accept() woke up the last process to
block on it then there'd be a performance improvement for servers
like Apache. Essentially, the last process into accept() is the
most recent in the CPU's caches, and so waking it first should
result, on average, in fewer cache misses.

Seems like a good idea, only two things concern me:

1) What if anything do the "standards" say about this?

2) There could be the possibility of starvation under some
corener cases, this needs to be investigated.

Yow! 11.26 MB/s remote host TCP bandwidth & ////
199 usec remote TCP latency over 100Mb/s ////
ethernet. Beat that! ////
-----------------------------------------////__________ o
David S. Miller, /_____________/ / // /_/ ><