Bug 32924

Summary:	raylink driver having problems receiving packets from win98se
Product:	Base System	Reporter:	Adam Powers <adam.powers>
Component:	misc	Assignee:	Duncan Barclay <dmlb>
Status:	Closed FIXED
Severity:	Affects Only Me
Priority:	Normal
Version:	Unspecified
Hardware:	Any
OS:	Any

Description Adam Powers 2001-12-17 07:20:01 UTC

hardware is fully function when speaking win98se<->win98se.

ray_rx() drops all packets because of an incorrect 802.11 header version. upon modifying the driver to print out the contents of the packets it receives, it appears that all packets contain the same repeating byte: "0x55 0x55 0x55 ...". it seems that ray_tx() is working fine because ethereal running on the windows 98 machine shows arp requests from the fbsd box and shows windows sending arp replies. also note that previous to printing out packet contents, a debug message would occasionally say "incorrect header version: <some other number than 0x55>", but i haven't seen that since. 

my test code, kernel object and messages file are available on request. (adam.powers@intel.com)

How-To-Repeat: attempt to use webgear aviator 2.4 cards to communicate between windows 98se and freebsd

Comment 1 Sheldon Hearn freebsd_committer

2001-12-30 13:37:19 UTC

Responsible Changed
From-To: freebsd-bugs->ray

Duncan looks after the ray(4) driver.

Comment 2 Sheldon Hearn freebsd_committer

2001-12-30 13:47:04 UTC

Responsible Changed
From-To: ray->dmlb

Oops, I got Duncan's user id wrong.

Comment 3 Duncan Barclay freebsd_committer

2002-06-16 11:11:02 UTC

State Changed
From-To: open->closed

Contact submitter a couple of months ago to see if this was still a problem. 
Had no reply so am closing the PR. 

Problem is memory mapping of the RX buffer space. Adjustment of sysctl's 
machdep.pccard.mem_start and machdep.pccard.mem_end can solve the problem.