Am Samstag, 6. November 2004 01:39 schrieb Stephane Marchesin:
> Dieter Nützel wrote:
> >Am Donnerstag, 4. November 2004 23:09 schrieb Roland Scheidegger:
> >>(sorry for sending this first to the wrong list. replies should be in
> >>dri-devel, not mesa3d-dev.)
> >>
> >>Here's an updated version of the patch, which should work on more cards.
> >>Specifically, it works on my rv250, rv280 and 9100igp should probably
> >>work too. Not so sure about r200, only if you're lucky...
> >
> >I had to revert this line, to see something (r200):
> >
> >--- radeon_state.c.orig 2004-11-05 23:29:41.000000000 +0100
> >+++ radeon_state.c 2004-11-06 01:04:51.325423262 +0100
> >@@ -981,7 +981,7 @@
> > * rendering a quad into just those buffers. Thus, we have to
> > * make sure the 3D engine is configured correctly.
> > */
> >- else if ((dev_priv->microcode_version == UCODE_R200) &&
> >+ if ( (dev_priv->microcode_version==UCODE_R200) &&
> > (flags & (RADEON_DEPTH | RADEON_STENCIL))) {
> >
> > int tempPP_CNTL;
>
> You shouldn't revert this line.
> Basically, this is is the test that chooses between doing a fast zclear
> and a standard zclear.
> But you don't want to do both.
>
> That was a mistake in my first patch. The second one has this change.
I see, even the speed! ;-)
> >Now, I have 'only' the black lines (gears, gloss, etc.).
> >But isosurf it mostly fine?! --- Depth???
> >See isosurface1.
> >
> >But if the 3D window is vertically to high or if I move a small window to
> > the bottom of the screen, then there is additional pixel (texture)
> > corruption. See isosurface2
>
> Same problem as the last one, I think.
Could it be stencil (buffer) related in some way?
Have a look at ~/progs/tests/stencil_wrap.
/opt/Mesa> progs/tests/stencil_wrap
GL_RENDERER = Mesa DRI R200 20041007 AGP 4x x86/MMX+/3DNow!+/SSE TCL
GL_VERSION = 1.3 Mesa 6.3
All 5 squares should be the same color.
Stencil bits = 8, maximum stencil value = 0x000000ff
The _right_ two boxes are broken.
> >>Some things are still broken, notably apps which use stencil buffer
> >>(nwn, doomIII). I think there could be an issue when clearing stencil or
> >>z-buffer separately, or maybe something else (dinoshade at least seems
> >>to work).
> >
> >Here, too.
> >
> >But I saw something semilar (depth corruption?) with viewperf before
> > HyperZ.
Have anyone looked into this?
> >>Also, if another window is moved on top of the glxgears window so it
> >>covers part of the glxgears window then rendering is very broken -
> >>mostly just black, that seriously seems to break z buffer clearing.
> >
> >Nothing, here. (With my change).
>
> That's because only the first box is cleared. This can be easily fixed.
Stephane, any screws where I could play as a start?
-Dieter
-------------------------------------------------------
This SF.Net email is sponsored by:
Sybase ASE Linux Express Edition - download now for FREE
LinuxWorld Reader's Choice Award Winner for best database on Linux.
http://ads.osdn.com/?ad_idU88&alloc_id065&op=click
--
_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel
