in SMI_Save(), SR6C and SR6D (and SR68) are saved only if pSmi->Dualhead,
why ? seems it should always be done ?
in SMI_WriteMode, those 2 are written twice, once everytime + once if
pSmi->Dualhead
What's your take about CCR68 (SR68), also written only in dualhead
mode... seems if SR6C/SR6D are to be active, SR68 has to be configured
too.
bruno
On Fri, 9 Mar 2007, Alex Deucher wrote:
On 3/9/07, bruno schwander <[EMAIL PROTECTED]> wrote:
On Fri, 9 Mar 2007, Alex Deucher wrote:
> The latest version of the siliconmotion driver in xorg git head should
> have the lockup fix you need. The problem is the engine doesn't need
> to be synced until it has been started. I've also added dualhead
> support. I haven't yet pushed the pll fix, that's still in my local
> tree at home along with some other stuff I'm working on. I can send
> you a patch later if you want.
that would be awesome, yes ! so far what I added (enabling in CCR68
vclock scaled from CCR6C and CCR6D) crashes my pc... I used
SMI_CommonCalcClock() but I get wrong values back for the register values.
I am certainly not calling it right.
I just put my local tree up here:
http://gitweb.freedesktop.org/?p=users/agd5f/xf86-video-siliconmotion.git;a=summary
The relevant commit is a328d4378c28b788acf320bee9fbdfd129e0923d
I haven't tested any of this new code yet, so YMMV. I haven't had
time this week and I need to rebuild my laptop with the smi chip, but
I hope to do that this weekend, at which point I should be able to
sort out any bugs. Let me know if you haven any questions.
Alex
bruno
_______________________________________________
Devel mailing list
Devel@XFree86.Org
http://XFree86.Org/mailman/listinfo/devel
_______________________________________________
Devel mailing list
Devel@XFree86.Org
http://XFree86.Org/mailman/listinfo/devel
