Martin Sperl [off-list ref] writes:

quoted

On 17.03.2016, at 17:57, Eric Anholt [off-list ref] wrote:


These don't exist on the hardware as far as I've been able to find.  "I
found it in a header file somewhere" is not sufficient justification to
expose it.

I'm working on getting a series of all of these reviewed and ready, so
I'm just dropping these PLLB hunks.

Fine with that - my test shows that these are not configured by the firmware.

As for headers: my experience is that these are the better resource
compared to all the public available documentation?

If you remember the ?frac? bit discussion where you said:

  Once again, trusting the docs turns out to be a bad idea.  You're right,
  the non-MASH clocks *do* have a bit 9 to enable fractional mode.  Sigh.

Also my understanding is that those headers are still used by the firmware
developers, so I guess that these are pretty stable and well maintained
(even if the ones available are by now a bit dated).

They even contain ifdefs for some earlier versions of the chip - 
see the dma-channels and their interrupts?
I.e: BCM2708A0 which does not have the DMA channels 9 to 15

So I would not discount all those pieces of information as totally
irrelevant.

I'm looking at the hardware source, and at the place where all the other
PLL dividers are, PLLB only has ARM.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 818 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20160317/ba579f79/attachment.sig>