headers
bugzilla-daemon | 1 May 2006 18:02

[Bug 6748] __gluInvertMatrixd:This function has error

Please do not reply to this email: if you want to comment on the bug, go to    

the URL shown below and enter yourcomments there.     

https://bugs.freedesktop.org/show_bug.cgi?id=6748          

brian.paul <at> tungstengraphics.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED

------- Additional Comments From brian.paul <at> tungstengraphics.com  2006-05-02 02:02 -------
OK, I've checked in the new function.  I did some tests and it looks good.  Thanks.          

     
--           
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email         

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
Permalink | Reply | 0 comments |
headers
bugzilla-daemon | 1 May 2006 23:03

[Bug 6706] Various bugs cause r300 to hang(hw lock being held)

Please do not reply to this email: if you want to comment on the bug, go to    

the URL shown below and enter yourcomments there.     

https://bugs.freedesktop.org/show_bug.cgi?id=6706          

aet <at> rasterburn.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jacekpoplawski <at> gmail.com

------- Additional Comments From aet <at> rasterburn.org  2006-05-02 07:03 -------
*** Bug 6804 has been marked as a duplicate of this bug. ***          

     
--           
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email         

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
Permalink | Reply | 0 comments |
headers
Alan Hourihane | 2 May 2006 12:46
Picon
Picon

Re: 3D performance loss with i915 driver

On Thu, 2006-04-27 at 13:55 +0200, Lukas Hejtmanek wrote:
> On Wed, Apr 26, 2006 at 08:06:51PM -0700, Eric Anholt wrote:
> > The open source driver isn't doing some important things that can be
> > done to improve memory bandwidth usage.  5x is more than I expected, but
> > it's not unbelievable.
> > 
> > If you had a clear performance regression between versions of the
> > open-source driver, we should definitely track that down, though.
> 
> Using CVS HEAD of xc (i.e. 6.9.0 Xorg) including bundled Mesa 6.4.1 I can reach
> 1110 FPS in glxgears and 27-30 FPS in ppracer (compared to 860 FPS glxgears and
> 11 FPS in ppracer using modular CVS HEAD of X.org and Mesa).
> 
> First, I thought that it's because of rotation support in i810 driver, but that
> version (6.9.0) also has rotation support so it is not the problem.

The problem here is batchbuffers.

Due to rotation being able to shuffle memory around it wasn't possible to
support batchbuffers in agp space as the 2D driver could rip up memory
allocation upon a rotation event. 

So the 3D driver falls back to the cmdbuffer path from system memory (rather
than AGP memory) which can be guaranteed.

To support batchbuffers again we'll need the new memory manager.

Alan.

-------------------------------------------------------
(Continue reading)

Permalink | Reply | 4 comments |
headers
Alan Hourihane | 2 May 2006 12:51
Picon
Picon

Re: 3D performance loss with i915 driver

On Tue, 2006-05-02 at 11:49 +0100, Alan Hourihane wrote:
> On Thu, 2006-04-27 at 13:55 +0200, Lukas Hejtmanek wrote:
> > On Wed, Apr 26, 2006 at 08:06:51PM -0700, Eric Anholt wrote:
> > > The open source driver isn't doing some important things that can be
> > > done to improve memory bandwidth usage.  5x is more than I expected, but
> > > it's not unbelievable.
> > > 
> > > If you had a clear performance regression between versions of the
> > > open-source driver, we should definitely track that down, though.
> > 
> > Using CVS HEAD of xc (i.e. 6.9.0 Xorg) including bundled Mesa 6.4.1 I can reach
> > 1110 FPS in glxgears and 27-30 FPS in ppracer (compared to 860 FPS glxgears and
> > 11 FPS in ppracer using modular CVS HEAD of X.org and Mesa).
> > 
> > First, I thought that it's because of rotation support in i810 driver, but that
> > version (6.9.0) also has rotation support so it is not the problem.
> 
> The problem here is batchbuffers.
> 
> Due to rotation being able to shuffle memory around it wasn't possible to
> support batchbuffers in agp space as the 2D driver could rip up memory
> allocation upon a rotation event. 
> 
> So the 3D driver falls back to the cmdbuffer path from system memory (rather
> than AGP memory) which can be guaranteed.
> 
> To support batchbuffers again we'll need the new memory manager.

to activate the batchbuffer path again for now though, you should be
able to do...
(Continue reading)

Permalink | Reply | 3 comments |
headers
Lukas Hejtmanek | 2 May 2006 14:23
Picon

Re: 3D performance loss with i915 driver

On Tue, May 02, 2006 at 11:51:05AM +0100, Alan Hourihane wrote:
> > The problem here is batchbuffers.
> > 
> > Due to rotation being able to shuffle memory around it wasn't possible to
> > support batchbuffers in agp space as the 2D driver could rip up memory
> > allocation upon a rotation event. 
> > 
> > So the 3D driver falls back to the cmdbuffer path from system memory (rather
> > than AGP memory) which can be guaranteed.
> > 
> > To support batchbuffers again we'll need the new memory manager.
> 
> to activate the batchbuffer path again for now though, you should be
> able to do...
> 
> INTEL_BATCH=1 glxgears

Using this increases glxgears ftp a little (940FPS vs 860FPS), but no change in
ppracer. 

Looks like there is still some issue with it.

--

-- 
Lukáš Hejtmánek

-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
(Continue reading)

Permalink | Reply | 2 comments |
headers
Alan Hourihane | 2 May 2006 14:37
Picon
Picon

Re: 3D performance loss with i915 driver

On Tue, 2006-05-02 at 14:23 +0200, Lukas Hejtmanek wrote:
> On Tue, May 02, 2006 at 11:51:05AM +0100, Alan Hourihane wrote:
> > > The problem here is batchbuffers.
> > > 
> > > Due to rotation being able to shuffle memory around it wasn't possible to
> > > support batchbuffers in agp space as the 2D driver could rip up memory
> > > allocation upon a rotation event. 
> > > 
> > > So the 3D driver falls back to the cmdbuffer path from system memory (rather
> > > than AGP memory) which can be guaranteed.
> > > 
> > > To support batchbuffers again we'll need the new memory manager.
> > 
> > to activate the batchbuffer path again for now though, you should be
> > able to do...
> > 
> > INTEL_BATCH=1 glxgears
> 
> Using this increases glxgears ftp a little (940FPS vs 860FPS), but no change in
> ppracer. 
> 
> Looks like there is still some issue with it.

You might want to check that your MTRR's are set correctly.

Alan.

-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
(Continue reading)

Permalink | Reply | 1 comment |
headers
Dave Airlie | 2 May 2006 14:47
Picon
Favicon

Re: 3D performance loss with i915 driver


> On Tue, 2006-05-02 at 14:23 +0200, Lukas Hejtmanek wrote:
>> On Tue, May 02, 2006 at 11:51:05AM +0100, Alan Hourihane wrote:
>>>> The problem here is batchbuffers.
>>>>
>>>> Due to rotation being able to shuffle memory around it wasn't possible to
>>>> support batchbuffers in agp space as the 2D driver could rip up memory
>>>> allocation upon a rotation event.
>>>>
>>>> So the 3D driver falls back to the cmdbuffer path from system memory (rather
>>>> than AGP memory) which can be guaranteed.
>>>>
>>>> To support batchbuffers again we'll need the new memory manager.
>>>
>>> to activate the batchbuffer path again for now though, you should be
>>> able to do...
>>>
>>> INTEL_BATCH=1 glxgears
>>
>> Using this increases glxgears ftp a little (940FPS vs 860FPS), but no change in
>> ppracer.
>>
>> Looks like there is still some issue with it.
>
> You might want to check that your MTRR's are set correctly.

To save Alan repeating himself, can we take this to bug 6814?

I'll post some more stuff to the bug... so far adding just the 
emit_invarient_state changes and I lose 100 FPS..
(Continue reading)

Permalink | Reply | 0 comments |
headers
bugzilla-daemon | 2 May 2006 16:14

[Bug 2372] better 'install' make targets

Please do not reply to this email: if you want to comment on the bug, go to    

the URL shown below and enter yourcomments there.     

https://bugs.freedesktop.org/show_bug.cgi?id=2372          

tilman <at> code-monkey.de changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
Attachment #5520 is|0                           |1
           obsolete|                            |

------- Additional Comments From tilman <at> code-monkey.de  2006-05-03 00:14 -------
Created an attachment (id=5548)
 --> (https://bugs.freedesktop.org/attachment.cgi?id=5548&action=view)
Patch

I don't know why I thought it was necessary to define PREFIX in the top level
Makefile. It's fixed in the new patch which works just fine.          

     
--           
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email         

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
(Continue reading)

Permalink | Reply | 0 comments |
headers
David Reveman | 3 May 2006 12:11
Picon
Favicon

Re: [compiz] Compilation failure as of 20060502

On Wed, 2006-05-03 at 02:44 +0200, Mark Rosenstand wrote:
> Hi,
> 
> The changes yesterday breaks compilation for me. I do have the latest
> glproto (1.4.7) and mesa cvs (20050603, i.e. a couple of hours ago) and
> xgl (20050603) have been recompiled after the updates.
> 
> Here's the initial error (with a whole lot similar):
> 
> texture.c: In function 'bindPixmapToTexture':
> texture.c:196: error: 'GLX_TEXTURE_FORMAT_EXT' undeclared (first use in
> this function)
> 
> What am I missing? :)

I think the problem is that mesa is missing some EXT_tfp enums in glx.h.
GLX_EXT_texture_from_pixmap is defined but enums are not. Try applying
the attached patch to mesa.

-David
Attachment (mesa-glx-h-tfp.diff): text/x-patch, 2336 bytes
Permalink | Reply | 1 comment |
headers
Mark Rosenstand | 3 May 2006 12:44

Re: [compiz] Compilation failure as of 20060502

On Wed, 2006-05-03 at 12:11 +0200, David Reveman wrote:
> On Wed, 2006-05-03 at 02:44 +0200, Mark Rosenstand wrote:
> > Hi,
> > 
> > The changes yesterday breaks compilation for me. I do have the latest
> > glproto (1.4.7) and mesa cvs (20050603, i.e. a couple of hours ago) and
> > xgl (20050603) have been recompiled after the updates.
> > 
> > Here's the initial error (with a whole lot similar):
> > 
> > texture.c: In function 'bindPixmapToTexture':
> > texture.c:196: error: 'GLX_TEXTURE_FORMAT_EXT' undeclared (first use in
> > this function)
> > 
> > What am I missing? :)
> 
> 
> I think the problem is that mesa is missing some EXT_tfp enums in glx.h.
> GLX_EXT_texture_from_pixmap is defined but enums are not. Try applying
> the attached patch to mesa.

That indeed fixed the problem. Thanks!

-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
Permalink | Reply | 0 comments |

Gmane