Hi Larry,

 >I think this is the first time we have seen a device that uses b43legacy with
 >any radio kill hardware. I suspect that the b43legacy_is_hw_radio_enabled() is
 >not looking at the correct bit. The hardware will obey the value on PIN13, no
 >matter what the driver does. That is why changing ...enabled() did no good.
 >
 >Please build with the attached patch, and run 2 times, once with PIN13 
 >connected
 >and once with PIN13 disconnected. That should let us see where the state of 
 >that pin shows up.

I applied your patch.
dmesg if PIN13 is open (wire disconnected on minipci adapter):
b43legacy-phy4: Broadcom 4306 WLAN found
b43legacy-phy4 debug: Found PHY: Analog 1, Type 2, Revision 1
b43legacy-phy4 debug: Found Radio: Manuf 0x17F, Version 0x2050, Revision 2
b43legacy-phy4 debug: Radio initialized
b43legacy-phy4 debug: DebugFS (CONFIG_DEBUG_FS) not enabled in kernel config
phy4: Selected rate control algorithm 'minstrel'
udev: renamed network interface wlan0 to wlan2
Broadcom 43xx-legacy driver loaded [ Features: PLID, Firmware-ID: FW10 ]
b43legacy ssb0:0: firmware: requesting b43legacy/ucode4.fw
b43legacy ssb0:0: firmware: requesting b43legacy/pcm4.fw
b43legacy ssb0:0: firmware: requesting b43legacy/b0g0initvals2.fw
b43legacy-phy4: Loading firmware version 0x127, patch level 14 (2005-04-18 
02:36:27)
b43legacy-phy4 debug: Chip initialized
b43legacy-phy4 debug: 30-bit DMA initialized
Registered led device: b43legacy-phy4::tx
Registered led device: b43legacy-phy4::rx
Registered led device: b43legacy-phy4::radio
b43legacy: phy rev < 3: RADIO_ENAB_LO 0xFD71
b43legacy: RADIO_ENAB_HI 0x40
b43legacy-phy4 debug: Wireless interface started
b43legacy-phy4 debug: Adding Interface type 2

dmesg if PIN13 is connected (no modifications on minipci adapter):
b43legacy-phy2: Broadcom 4306 WLAN found
b43legacy-phy2 debug: Found PHY: Analog 1, Type 2, Revision 1
b43legacy-phy2 debug: Found Radio: Manuf 0x17F, Version 0x2050, Revision 2
b43legacy-phy2 debug: Radio initialized
b43legacy-phy2 debug: DebugFS (CONFIG_DEBUG_FS) not enabled in kernel config
phy2: Selected rate control algorithm 'minstrel'
Broadcom 43xx-legacy driver loaded [ Features: PLID, Firmware-ID: FW10 ]
udev: renamed network interface wlan0 to wlan1
b43legacy ssb0:0: firmware: requesting b43legacy/ucode4.fw
b43legacy ssb0:0: firmware: requesting b43legacy/pcm4.fw
b43legacy ssb0:0: firmware: requesting b43legacy/b0g0initvals2.fw
b43legacy-phy2: Loading firmware version 0x127, patch level 14 (2005-04-18 
02:36:27)
b43legacy-phy2 debug: Chip initialized
b43legacy-phy2 debug: 30-bit DMA initialized
Registered led device: b43legacy-phy2::tx
Registered led device: b43legacy-phy2::rx
Registered led device: b43legacy-phy2::radio
b43legacy: phy rev < 3: RADIO_ENAB_LO 0xDD71
b43legacy: RADIO_ENAB_HI 0x10001
b43legacy-phy2 debug: Wireless interface started
b43legacy-phy2 debug: Adding Interface type 2

wow, there are many different bits ...

 >Are these related to large packets? If you reduce the size to 1000, does it
 >still err?

I tested flood pings with packet size 1000 bytes.
If I do 10000 packets (it took 33 seconds), 8 packets got lost. I got 12 PHY 
transmission errors.

Yesterday I also tested downloading 200MByte via http-wget over the wireless 
link. I downgraded Linksys WRT54GL WLAN-Client to 1MBit/s bitrate. Nevertheless 
it downloaded with roughly 400kbyte/s ... ?!?
After 100 Mbyte the connection was gone.
The b43legacy hostapd-AP reported "controller restarted" after several PHY 
Transmission errors :(
here the exact dmesg:
b43legacy-phy3 ERROR: PHY transmission error
b43legacy-phy3 debug: Wireless interface stopped
b43legacy-phy3 debug: DMA-30 0x0260 (RX) max used slots: 10/64
b43legacy-phy3 debug: DMA-30 0x0200 (RX) max used slots: 11/64
b43legacy-phy3 debug: DMA-30 0x02A0 (TX) max used slots: 0/128
b43legacy-phy3 debug: DMA-30 0x0280 (TX) max used slots: 0/128
b43legacy-phy3 debug: DMA-30 0x0260 (TX) max used slots: 0/128
b43legacy-phy3 debug: DMA-30 0x0240 (TX) max used slots: 0/128
b43legacy-phy3 debug: DMA-30 0x0220 (TX) max used slots: 128/128
b43legacy-phy3 debug: DMA-30 0x0200 (TX) max used slots: 0/128
b43legacy-phy3 debug: Radio initialized
b43legacy-phy3 debug: Radio initialized
b43legacy-phy3: Loading firmware version 0x127, patch level 14 (2005-04-18 
02:36:27)
b43legacy-phy3 debug: Chip initialized
b43legacy-phy3 debug: 30-bit DMA initialized
Registered led device: b43legacy-phy3::tx
Registered led device: b43legacy-phy3::rx
Registered led device: b43legacy-phy3::radio
b43legacy-phy3 debug: Wireless interface started
__ratelimit: 270 callbacks suppressed
b43legacy-phy3: Controller restarted

I also got
b43legacy-phy2 warning: Unexpected value for chanstat (0x7C00)
b43legacy-phy2 warning: Unexpected value for chanstat (0x7C00)
before.

I'll try the driver version of b43legacy in kernel source 2.6.31.6. Perhaps it 
is not good to have bleeding edge :-)

Thanks,
Daniel

_______________________________________________
Bcm43xx-dev mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev

Reply via email to