Michael Buesch wrote:
> On Sunday 10 September 2006 21:32, Larry Finger wrote:
>> Are the clean-room specs up to date for G PHY Revision 8? I'm 
> 
> Is there something that does not work?

This question relates to a BCM4310 PCI-E card with a Rev. 8 PHY that 
I'm debugging remotely. In our last test, it connects with the bus and 
appears to do output DMA transfers; however, we don't know if it is 
broadcasting. He doesn't have a second system to monitor the airwaves. 
It is definitely not receiving, so I'm checking all the bases.

>> particularly concerned about the 7 in the statement "If 2 <= PHY 
>> Revision <= 7, perform a B6PHY init and then continue with the GPHY 
>> Init" as found on  http://bcm-specs.sipsolutions.net/GPHY.
> 
> Well, it does not execute B6PHY init for PHYrev > 7.
> Do you think this is wrong? Why?
> Does changing this to <=8 make something work?

I don't know. That trial is currently in progress.

I also have additional questions in the section

  13. Otherwise

    a. If this is a 2050 radio with a Radio Revision of 8
          1. Write the saved value of TXCTL1 left shifted by 4, then
            bitwise OR'd with the saved value of TXCTL2

Where should this be written? Later in that section

    c. If the PHY Revision is 6 or greater

          1. MaskSet PHYRegister 0x36 with mask 0xF000 and set with
            FIXME left shifted by 12

What value is this FIXME? A little further down, there is

    e. Otherwise

         1. Write 0x807F to PHYRegister 0x2E

Here, the code has 0x3E, not 0x2E. Which is correct, the code or the 
specs?

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

Reply via email to