More 'resource' problems with "ath0"

[Sam Leffler]

Ross Finlayson wrote:
> The problems are occurring even with a very recently built kernel:
> 
> %cat /etc/motd
> FreeBSD 6.1-STABLE (GENERIC) #0: Tue May 16 15:55:19 PDT 2006
> 
> May 17 07:53:22 ns dhcpd: send_packet: No buffer space available
> May 17 07:53:25 ns dhcpd: send_packet: No buffer space available
> May 17 07:53:35 ns named[382]: client 66.80.62.46#1026: error sending 
> response: not enough free resources
> 
> Note that client 66.80.62.46 is on "ath0".
> 
> %netstat -m
> 188/337/525 mbufs in use (current/cache/total)
> 109/205/314/10688 mbuf clusters in use (current/cache/total/max)
> 109/147 mbuf+clusters out of packet secondary zone in use (current/cache)
> 0/0/0/0 4k (page size) jumbo clusters in use (current/cache/total/max)
> 0/0/0/0 9k jumbo clusters in use (current/cache/total/max)
> 0/0/0/0 16k jumbo clusters in use (current/cache/total/max)
> 265K/494K/759K bytes allocated to network (current/cache/total)
> 0/0/0 requests for mbufs denied (mbufs/clusters/mbuf+clusters)
> 0/0/0 requests for jumbo clusters denied (4k/9k/16k)
> 0/87/2928 sfbufs in use (current/peak/max)
> 0 requests for sfbufs denied
> 0 requests for sfbufs delayed
> 1904 requests for I/O initiated by sendfile
> 246 calls to protocol drain routines
> 
> Any ideas?  Has anyone else run into this problem?
> 
> Is there anything else I can run (other than "netstat -m") to help 
> diagnose this  problem?

Please do not cross-post.

The transmit queue is full; you need to understand why.  Basic info 
includes the card info (dmes|grep ath) and ifconfig output.  It usually 
helps to explain how you make your problem to occur--if you can identify 
something you are doing.  Past that there are tools in src/tools/tools 
such as ath/athstats for dumping the driver statistics.

	Sam