cnfRcvCrcError problems

[Matthew Ressler]

Hello,

Has anyone ever successfully enabled the cnfRcvCrcError RID (=0xFC30) on 
PRISM 2/2.5/3 chips and verified that frames with CRC errors are passed 
to the driver (i.e., by checking the status bits of the Rx frame). I've 
tried on a number of different firmware versions (1.5.4, 1.7.4, 1.8.0) 
but with no success. The documentation in RM025 (2.2) is a somewhat 
inconsistent -- it claims to write to bit 1 of the data portion RID, set 
bits 2-15 to zero, but mentions nothing about bit 0. Also, it claims the 
WORD OFFSET for this data as 3, but the dataLen is 2, which also seems 
inconsistent. Regardless, I tried a number of values in the data portion 
of the RID, but with no success in any case.

The only way I have been able to get frames with CRC errors to be passed 
to the driver is to enable Promiscuous mode. But promiscuous mode seems 
to have a firmware bug where received frames (with valid CRC) intended 
for the host are not ack'd, thereby severely limiting throughput (by at 
least a factor of 10).

Thanks in advance,

Matt

-- 
Matthew B. Ressler
Software Engineer
SRI International, SLO Office
805-542-9330 x101
matthew.ressler at sri.com

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 5217 bytes
Desc: S/MIME Cryptographic Signature
Url : http://lists.shmoo.com/pipermail/hostap/attachments/20040319/130a0fb5/attachment.bin