Am 2004.10.16 00:59:07 +0200 schrieb(en) Ian Romanick:
Andreas Stenglein wrote:
Could someone else with an R100 test?
see attached patch ( some changes to make mesa compile with gcc2.95.3 omitted)
What were those changes? Were they to avoid the SSE2 code?
1) basically some brackets{} ,
Andreas Stenglein wrote:
Am 2004.10.16 00:59:07 +0200 schrieb(en) Ian Romanick:
Andreas Stenglein wrote:
Could someone else with an R100 test?
see attached patch ( some changes to make mesa compile with gcc2.95.3 omitted)
What were those changes? Were they to avoid the SSE2 code?
1) basically
On Wed, 13 Oct 2004 14:37:12 -0700, Ian Romanick [EMAIL PROTECTED] wrote:
Michel Dänzer wrote:
Anyway, I think we're on a tangent here, as the problem doesn't seem to
be PPC specific at all.
I dug out the Rage 128 that I have for the PC, and it works just fine.
glxgears, readpix, all of
Andreas Stenglein wrote:
Could someone else with an R100 test?
see attached patch ( some changes to make mesa compile with gcc2.95.3 omitted)
What were those changes? Were they to avoid the SSE2 code?
In any case, your patch looks good to me. In fact, it looks almost
identical to one that I'm
Am 2004.10.13 23:37:12 +0200 schrieb(en) Ian Romanick:
Michel Dänzer wrote:
Anyway, I think we're on a tangent here, as the problem doesn't seem to
be PPC specific at all.
I dug out the Rage 128 that I have for the PC, and it works just fine.
glxgears, readpix, all of it. :(
The PCI
Am 2004.10.13 23:37:12 +0200 schrieb(en) Ian Romanick:
Michel Dänzer wrote:
Anyway, I think we're on a tangent here, as the problem doesn't seem to
be PPC specific at all.
I dug out the Rage 128 that I have for the PC, and it works just fine.
glxgears, readpix, all of it. :(
The PCI
Am 2004.10.12 03:37:05 +0200 schrieb(en) Ian Romanick:
I was trying to test the latest version of my ReadPixels work to make
sure I didn't break anything on big-endian. However, it seems someone
beat me to it in the Rage128 driver. :) In a nutshell, I can get one
frame of gears, and then
On Wed, 13 Oct 2004, Benjamin Herrenschmidt wrote:
On Wed, 2004-10-13 at 00:41, Vladimir Dergachev wrote:
Just to check off the obvious, are you running a recent kernel with (I
assume framebuffer) ? It could be that the default might have changed to
configure the apertures to be bigendian.
On Wed, 2004-10-13 at 10:27 -0400, Vladimir Dergachev wrote:
On Wed, 13 Oct 2004, Benjamin Herrenschmidt wrote:
On Wed, 2004-10-13 at 00:41, Vladimir Dergachev wrote:
Just to check off the obvious, are you running a recent kernel with (I
assume framebuffer) ? It could be that the
Am 2004.10.12 18:38:18 +0200 schrieb(en) Ian Romanick:
Andreas Stenglein wrote:
It might be unrelated (and just some other silly mistake/problem):
Using my local version of radeon (r100) driver converted to t_vertex
I discovered a similar problem recently.
1) running glxgears I get the
Michel Dnzer wrote:
Anyway, I think we're on a tangent here, as the problem doesn't seem to
be PPC specific at all.
I dug out the Rage 128 that I have for the PC, and it works just fine.
glxgears, readpix, all of it. :(
The PCI ID on on the PC version is 1002:5046, and on the Mac it's
On Thu, 2004-10-14 at 00:27, Vladimir Dergachev wrote:
On Wed, 13 Oct 2004, Benjamin Herrenschmidt wrote:
On Wed, 2004-10-13 at 00:41, Vladimir Dergachev wrote:
Just to check off the obvious, are you running a recent kernel with (I
assume framebuffer) ? It could be that the default
Really ? Both the register and the framebuffer apertures ?
The reason I am asking as ati driver has some code to invert the
endianness for writing Xv data to the framebuffer.
No, only framebuffer, and we used, iirc, to switch the card's swapper
off when writing Xv data (maybe we just swap the data
On Mon, 11 Oct 2004, Ian Romanick wrote:
I was trying to test the latest version of my ReadPixels work to make sure I
didn't break anything on big-endian. However, it seems someone beat me to it
in the Rage128 driver. :) In a nutshell, I can get one frame of gears, and
then the 3D engine is
Michel Dnzer wrote:
On Mon, 2004-10-11 at 18:37 -0700, Ian Romanick wrote:
I was trying to test the latest version of my ReadPixels work to make
sure I didn't break anything on big-endian. However, it seems someone
beat me to it in the Rage128 driver. :) In a nutshell, I can get one
frame of
Andreas Stenglein wrote:
It might be unrelated (and just some other silly mistake/problem):
Using my local version of radeon (r100) driver converted to t_vertex
I discovered a similar problem recently.
1) running glxgears I get the hang with only mousepointer moving - reboot needed.
I get the
On Tue, 2004-10-12 at 08:36, Ian Romanick wrote:
It's a stock AGP G4. The card is the original Rage128. The kernel is
the debian 2.6.8-powerpc kernel (dittor for DRM), and X is yesterday's
X.org. The last time I did anything with that machine was about a month
ago with a 2.4.25 kernel
Am 2004.10.12 18:38:18 +0200 schrieb(en) Ian Romanick:
Andreas Stenglein wrote:
It might be unrelated (and just some other silly mistake/problem):
Using my local version of radeon (r100) driver converted to t_vertex
I discovered a similar problem recently.
1) running glxgears I get the
On Wed, 2004-10-13 at 00:41, Vladimir Dergachev wrote:
Just to check off the obvious, are you running a recent kernel with (I
assume framebuffer) ? It could be that the default might have changed to
configure the apertures to be bigendian.
Changed ? The apertures have always been BE on PPC
On Wed, 2004-10-13 at 09:07 +1000, Benjamin Herrenschmidt wrote:
On Wed, 2004-10-13 at 00:41, Vladimir Dergachev wrote:
Just to check off the obvious, are you running a recent kernel with (I
assume framebuffer) ? It could be that the default might have changed to
configure the apertures
I was trying to test the latest version of my ReadPixels work to make
sure I didn't break anything on big-endian. However, it seems someone
beat me to it in the Rage128 driver. :) In a nutshell, I can get one
frame of gears, and then the 3D engine is toast. After that frame is
drawn, gears
On Mon, 2004-10-11 at 18:37 -0700, Ian Romanick wrote:
I was trying to test the latest version of my ReadPixels work to make
sure I didn't break anything on big-endian. However, it seems someone
beat me to it in the Rage128 driver. :) In a nutshell, I can get one
frame of gears, and then
22 matches
Mail list logo
