[SQU] CPU use stuck at 100%

Wondering if anyone is familiar with this problem. There is a brief
Q+A in the FAQ with a similar problem, but the only answer is to make
sure you have a /dev/null, which I do.

When I run squid, the child it spawns off uses 100%, in a tight
poll/gettimeofday loop. The whole thing works fine, including using
smb_auth for authentication, but its wasting a ton of resources doing
this loop. I commented out the authentication program to make sure it
wasn't caused by smb_auth. The process thats using the cpu is 'squid',
the child of 'squid' (one PID higher). From pstree:

 |-squid
 | `-squid <--- thats the sucker
 | `-unlinkd

Heres the loop, from an strace:

.
.
poll(0xbfffdc64, 0x2, 0, 0x40167520, 0xbfffdc64) = 0
gettimeofday({973735859, 973048}, NULL) = 0
poll(0xbfffdc64, 0x2, 0, 0x40167520, 0xbfffdc64) = 0
gettimeofday({973735859, 975345}, NULL) = 0
gettimeofday({973735859, 976071}, NULL) = 0
poll(0xbfffdb84, 0, 0, 0x40167520, 0xbfffdb84) = 0
poll(0xbfffdc64, 0x2, 0, 0x40167520, 0xbfffdc64) = 0
.
.

I don't remember noticing this behaviour when I installed squid quite
a few months ago, but I may have just missed it. Any clues?

-- 
Jeremy M. Dolan <jmd@turbogeek.org>
OpenPGP key = http://turbogeek.org/openpgp-key
OpenPGP fingerprint = 494C 7A6E 19FB 026A 1F52  E0D5 5C5D 6228 DC43 3DEE
--
To unsubscribe, see http://www.squid-cache.org/mailing-lists.html