Re: Re: Questions about hx4700-init.h

From: Eddi <h4100_at_dpeddi.com>
Date: Wed, 12 Jan 2005 21:54:21 +0100

Il giorno mer, 12-01-2005 alle 14:45 -0500, Todd Blumer ha scritto:

If I understand well, these init value may differ between each model.
We can focus at INIT_DIR, on my h4150 (I subscribed here to look at some
developement that may be reused on my h4150) I can have an asic gpio
connected to the blue led, on hx4700 this may be connected to another
gpio, so before decide the value to assign to these define, perhaps is
better to dump asic register with haret.

A question, the asic3 on hx4700 is the same as the one on h4100?
>From a noli mail:
>SL-A300 source is very
>important to H4000 devices because of the HTC Chip(HTC 30H88028-00
>2226560XB101 JAPAN 0320EAI F0103ZAF)

If yes:
locate xanadux cvs @ sourceforge and try to look at their asic3-base.c.
they have a version that should dump some asic registry on load, and can
be useful to see the uninitialized status.
Michaelo, you should have a old mail from shawn andersen with a diff
named platform.diff with some of this code and other (that was for
2.6.6_at_h4100)

>From an old discussion between me and sa about our h4100:
> now I will remove referencies to pcmcia.dll into regedit to see if
some
> wince flag are still initialized..
>
> After removing pcmcia.dll from registry, I get:
> - both bluetooth and wify doesn't works any more
> - wifi and bluetooth led doesn't come up any more...
> with pcmcia.dll
> HaRET(10)# pd 0x0C000000 30
> 0c000000 | ffff fc7f 0000 0000
> 0c000010 | 0000 0000 ffff 0000
> HaRET(11)# pd 0x0C000100 30
> 0c000100 | ffff dfbf 1c00 0000
> 0c000110 | 0000 0000 ffff 0000
> HaRET(12)# pd 0x0C000200 30
> 0c000200 | ffff ffff 4700 0000
> 0c000210 | 0000 0000 ffff 4000
> HaRET(13)# pd 0x0C000300 30
> 0c000300 | ff03 ef03 4e03 00fc
> 0c000310 | 0080 0000 ff83 0100
> HaRET(14)# pd 0x0C000400 30
>
> without pcmcia.dll
> HaRET(1)# pd 0x0C000000 30
> 0c000000 | ffff fc7f 0000 0000
> 0c000010 | 0000 0000 ffff 0000
> HaRET(2)# pd 0x0C000100 30
> 0c000100 | ffff dfbf 1c00 0000
> 0c000110 | 0000 0000 ffff 0000
> HaRET(3)# pd 0x0C000200 30
> 0c000200 | ffff ffff 4700 0000
> 0c000210 | 0000 0000 ffff 4000
> HaRET(4)# pd 0x0C000300 30
> 0c000300 | ff03 ef03 4e03 00fc
> 0c000310 | 0080 0000 ff83 0100
>
> But doesn't change any asic3 gpio in the a-d range.. and doesn't
change
> anything on pxa gpio too...
>
> any idea?

So I remember you... with haret if you have a asic by known the asic
address, always if it's the same asic3, you should get it with a MAP
from the bootloader, you may dump the register, but.....

I don't know how to track changes...
Haret should lack a function, to "pd" on gpio change...

I mean... pd command should be launched only when the asic3 gpio
connected to the pxa change its status.

My log may be wrong because i could have launched "pd" in this situation

0 "pd" 0->1,1->0 "pd" 0

>
> > Does this come from secret Asic3 docs that Jamey isn't allowed to share?

Jamey take some times to get h4100 specification from htc, I hope for
you that the real hx4700 producer have already sent this doc to CRL.

Eddi

PS: michael if something in my english may be difficult to understand, I
can tray to explain it to you in italian ;d
Received on Wed Jan 12 2005 - 15:55:33 EST

This archive was generated by hypermail 2.2.0 : Mon Jul 25 2005 - 17:21:03 EDT