So, I decided to give b43 in AP mode a go on my ASUS wl500gp with r14139
(kernel 2.6.25.20).  The results are both good and bad.

First the good news.  The problems I was seeing with b43/AP mode on my
Linksys WRT54GS are gone.  I don't see the CPU saturated with ksoftirqd.
Yay.

The bad news is that throughput still sucks rocks.

On a laptop as a client of the b43 wireless on the ASUS router, I ran
the following:

r...@brian-laptop:~# ssh pc cat /dev/sdd | dd of=~brian/tmp/sdd.img
bs=1k count=10000
r...@pc's password: 
10000+0 records in
10000+0 records out
10240000 bytes (10 MB) copied, 133.622 s, 76.6 kB/s

As you can see, the performance is not at all stellar.  If we change up
the block size/number of blocks we see the following results:

r...@brian-laptop:~# ssh pc cat /dev/sdd | dd of=~brian/tmp/sdd.img
bs=100k count=100
r...@pc's password: 
0+100 records in
0+100 records out
3129344 bytes (3.1 MB) copied, 50.2239 s, 62.3 kB/s

It would appear that with 100k blocks, it fails to even transfer
successfully.

During the first transfer, top reports:

Mem: 28496K used, 1432K free, 0K shrd, 4576K buff, 8676K cached
CPU:   0% usr   1% sys   0% nice  91% idle   0% io   0% irq   6% softirq
Load average: 0.08 0.08 0.08
  PID  PPID USER     STAT   VSZ %MEM %CPU COMMAND
 2359     1 root     S     3924  13%   2% hostapd -B /var/run/hostapd-wlan0.con
 2393  1973 root     R     2412   8%   1% top 
    3     2 root     SW<      0   0%   1% [ksoftirqd/0]

With everything below that at 0% CPU.  So as you can see clearly, this
is not a CPU bottleneck like it was on the Linksys unit.

iwconfig reports the following on the laptop:

wlan0     IEEE 802.11g  ESSID:"ilinx_exp"  Nickname:""
          Mode:Managed  Frequency:2.462 GHz  Access Point: XX:XX:XX:46:8F:7A   
          Bit Rate=54 Mb/s   Tx-Power=27 dBm   
          Retry min limit:7   RTS thr:off   Fragment thr=2346 B   
          Encryption key:XXXX-XXXX-XXXX-XXXX-XXXX-XXXX-XXXX-XXXX [2]
          Power Management:off
          Link Quality=50/100  Signal level=-75 dBm  Noise level=-95 dBm
          Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
          Tx excessive retries:0  Invalid misc:0   Missed beacon:0

FWIW, I do notice a bunch of these in the kernel ring buffer (i.e.
dmesg):

b43-phy0 ERROR: PHY transmission error
b43-phy0 ERROR: PHY transmission error
b43-phy0 ERROR: PHY transmission error
b43-phy0 ERROR: PHY transmission error
b43-phy0 ERROR: PHY transmission error
b43-phy0 ERROR: PHY transmission error

Any ideas where/what to investigate?

b.

Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
http://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel

Reply via email to