The patch was based off the FIXME comment in the code [which suggested
replacing 11 with the boundary for the regulatory domain]. I'll take a
look at the vendors driver and see about modifying the patch to fit
more with its logic or just limmiting it to the FCC regulatory domain.

On 10/30/06, Uli Kunitz <[EMAIL PROTECTED]> wrote:
Johannes Berg wrote:
>> I'm not so sure about this. This patching might be US-specific and we
>> cannot simply apply the setting for top channel of another domain
>> instead of channel 11. One option would be to set the value only under
>> the US regulatory domain.
>
> ??
> What the patch does is replace the top channel which is hardcoded to 11
> by the top channel given by the current regulatory domain. How can that
> be wrong? Except that you may want to init the regulatory domain from
> the EEPROM but I'm not sure how the ieee80211 code works wrt. that.
>
> johannes

The problem is not so much that I don't trust the geo code, but whether
setting the register to that band-edge value for a higher channel is
the right thing to do. It looks like that this is a hack for FFC
compliance. Therefore I suggest to patch CR128 only
for the US regulatory domain.

Here is the code from the GPL vendor driver (zdhw.c):

    if (pObj->HWFeature & BIT_21)  //6321 for FCC regulation, enabled HWFeature 
6M band edge bit (for AL2230, AL2230S)
     {
         if (ChannelNo == 1 || ChannelNo == 11)  //MARK_003, band edge, these 
may depend on PCB layout
         {
             pObj->SetReg(reg, ZD_CR128, 0x12);
             pObj->SetReg(reg, ZD_CR129, 0x12);
             pObj->SetReg(reg, ZD_CR130, 0x10);
             pObj->SetReg(reg, ZD_CR47, 0x1E);
         }
         else //(ChannelNo 2 ~ 10, 12 ~ 14)
         {
             pObj->SetReg(reg, ZD_CR128, 0x14);
             pObj->SetReg(reg, ZD_CR129, 0x12);
             pObj->SetReg(reg, ZD_CR130, 0x10);
             pObj->SetReg(reg, ZD_CR47, 0x1E);
         }
     }

The patch from Holden would set ZD_CR128 to 0x12 for the highest channel,
which would not reflect the logic of the vendor driver.

Kind regards,

Uli

--
Uli Kunitz ([EMAIL PROTECTED])
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html



--
Cell: 613-276-1645
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to