[Soekris] net5501 and GPIO

[Poul-Henning Kamp]

In message <20080129093524.G90367 at idiom.com>, Ralph Becker-Szendy writes:
>On Tue, 29 Jan 2008, Poul-Henning Kamp wrote:
>
>> This is the first I hear about that, can you elaborate which pins
>> you find unexpected places ?
>
>If I had the detailed pinout information with me, I would have posted
>it.  What I should have done is to post that I'll follow up tonight
>with the correct listing.
>
>If my vague memory is right, the pinout goes as follows:
>- Pins 1 and 2 are some sort of power.
>- From pins 3... on, it starts with GPIO16 and up.  The manual
>   had some different number assignment for those.

Ahh, hang on, somebody misunderstood something.

The PC87366 numbers it's gpio pins in octal: GPIO23 is bit 3 in the
second group of GPIO pins.

So what you call GPIO16 is what the chip calls GPIO20. Yours
is decimal, the chip uses octal.

>- If you start at pin 3, there are at least 6 or 7 useable
>   GPIO pins, followed by a ground pin right afterwards.

8 actually, GPIO20-27.

>- I found more IO pins, plus the serial pins, further down the
>   connector, in the expected places but again with the GPIO number
>   assignment not exactly matching the 4801 manual.

They are GPIO04/05 and GPIO12/13 in PC87366 numbering.

(Søren has confused the issue a bit by coming up with his own
NET4801 GPIO numbers that run from zero to eleven, disregard those).

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.