[ltp] [PATCH] Add support for FIR IrDA for T43/R52

[Henrique de Moraes Holschuh]

On Sat, 03 Jun 2006, Shem Multinymous wrote:
> There's a bunch of photos on ThinkWiki, maybe one of them happens to
> show the chip: http://thinkwiki.org/wiki/Category:T43

None. I think it is the NS/Winbond 8394T, it has as SuperIO on top of the
TPM functions, and there is no other SuperIO listed anywhere for the T43 and
R52.

The problem here is that the datasheet for the 8394T is available only under
NDA due to the TPM module inside it.  Again, IBM screwed us (T43/R52 onwers)
over: first with their cut-corners approach to using a *broken* SATA-PATA
bridge and disabling the ICH6M's native AHCI mode so that they need no
drivers for MS Windows, and now by using a TPM module with SuperIO built-in
which is a major crap to work with.

The *60 series get this right, they use a regular SuperIO chip, and no
SATA-PATA bridges that require HD firmware modifications due to bugs.  I
have no idea if they run the ICH6M/ICH7M on native mode, though.

> This may be even more dangerous, given:
> >Just a warning: do *not* remove nsc-ircc if other modules are using it, it

This is something I can try to fix (even if it is a bug caused by the big
fish upstream), as I don't need another FIR device to track down and fix the
issue.  I am on it, actually, but no promises.

I did notice the IR dongle autodetection is unreliable, and that's why we
always have to tell the driver which dongle to use.  It worked sometimes in
auto mode for me, but not always (and the first miss was *after* I sent the
email about it :-) ).

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh