Nabble - 3D Rendering

Re: TGSI simplification branch

2009-11-25T07:57:01Z

Keith Whitwell pisze:

> On Wed, 2009-11-25 at 06:28 -0800, michal wrote:
>
>> Keith Whitwell pisze:
>>
>>> I've pushed a feature branch with some tgsi simplifications in it. With
>>> these I've removed the biggest remaining oddities of that language, and
>>> it's getting to a place where I'm starting to be happy with it as a
>>> foundation for future work.
>>>
>>> Most of the surprising stuff like multiple negate flags, etc, is gone
>>> now, and the core tokens are quite a bit easier to understand than in
>>> previous iterations.
>>>
>>> I've still got my eye on reducing the verbosity of the names in the
>>> tgsi_parse.h "FullToken" world, and promoting the tgsi_any_token union
>>> into p_shader_tokens.h.
>>>
>>> It would be good if people can review the interface changes and provide
>>> feedback, and also test out their drivers on this branch. I've done
>>> minimal softpipe testing so far but will do more over the next few days.
>>>
>>>
>>>
>> All looks good to me, I'm happy somebody had the guts to cut off all the
>> cruft in one shot.
>>
>> I see some compile errors on windows build -- I will fix those along
>> with other minor bugs I have spotted.
>>
>> Now, looking at the interface, I'm thinking about removing some more tokens.
>>
>> 1) Remove tgsi_dimension and use tgsi_src_register directly with some
>> well-defined constraints.
>>
>> 2) Do the same to tgsi_instruction_predicate. Really, it's just an
>> optional src operand with some restrictions.
>>
>
> Interesting. I'd be keen to see a patch.
>
>
>

Attached. But the more I look at it the more lame it gets.

Another option would be to define tgsi_any_register that would have
File, Index, Indirect and Dimension fields. Then there would be more
specialised tgsi_*_register tokens, that would be binary compatible with
the first one. One could cast them using a union and avoid more mistakes
at compile time. That way we don't have to put the constraints in
comments, but be more strict and use the compiler to enforce them. I
will follow up with a patch.

diff --git a/src/gallium/include/pipe/p_shader_tokens.h b/src/gallium/include/pipe/p_shader_tokens.h
index 7d73d7d..7bea99a 100644
--- a/src/gallium/include/pipe/p_shader_tokens.h
+++ b/src/gallium/include/pipe/p_shader_tokens.h
@@ -290,7 +290,9 @@ union tgsi_immediate_data
* respectively. For a given operation code, those numbers are fixed and are
* present here only for convenience.
*
- * If Predicate is TRUE, tgsi_instruction_predicate token immediately follows.
+ * If Predicate is TRUE, tgsi_src_register token immediately follows. Only
+ * the File, Index, Negate and Swizzle* fields are valid. File must be set
+ * to TGSI_FILE_PREDICATE and Swizzle is either set to identity or replicate.
*
* Saturate controls how are final results in destination registers modified.
*/
@@ -350,21 +352,6 @@ struct tgsi_instruction_texture
unsigned Padding : 24;
};

-/*
- * For SM3, the following constraint applies.
- * - Swizzle is either set to identity or replicate.
- */
-struct tgsi_instruction_predicate
-{
- int Index : 16; /* SINT */
- unsigned SwizzleX : 2; /* TGSI_SWIZZLE_x */
- unsigned SwizzleY : 2; /* TGSI_SWIZZLE_x */
- unsigned SwizzleZ : 2; /* TGSI_SWIZZLE_x */
- unsigned SwizzleW : 2; /* TGSI_SWIZZLE_x */
- unsigned Negate : 1; /* BOOL */
- unsigned Padding : 7;
-};
-
/**
* File specifies the register array to access.
*
@@ -396,23 +383,12 @@ struct tgsi_src_register
};

/**
- * If tgsi_src_register::Modifier is TRUE, tgsi_src_register_modifier follows.
- *
- * Then, if tgsi_src_register::Indirect is TRUE, another tgsi_src_register
- * follows.
+ * If tgsi_src_register::Indirect is TRUE, tgsi_src_register follows.
*
- * Then, if tgsi_src_register::Dimension is TRUE, tgsi_dimension follows.
+ * If tgsi_src_register::Dimension is TRUE, tgsi_src_register follows.
+ * Only the Indirect, Dimension and Index fields are valid.
*/

-
-struct tgsi_dimension
-{
- unsigned Indirect : 1; /* BOOL */
- unsigned Dimension : 1; /* BOOL */
- unsigned Padding : 14;
- int Index : 16; /* SINT */
-};
-
struct tgsi_dst_register
{
unsigned File : 4; /* TGSI_FILE_ */

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: Unstable Vista Menus

2009-11-25T07:47:34Z

I had a problem at first because I couldn't get the canvas3d to show. Mine is called c and I had to manually set the size like so after I created it.

// New universe
c = createSplashUniverse();
jPanel1.add(c, java.awt.BorderLayout.CENTER);
c.setSize(jPanel1.getSize());

Maybe this is your problem? You can check first by getting the size of the canvas3d object - if it is 0x0 it will be invisible.
Dave
[Message sent by forum member 'dwsubc' ]

http://forums.java.net/jive/thread.jspa?messageID=373326

---------------------------------------------------------------------
To unsubscribe, e-mail: interest-unsubscribe@...
For additional commands, e-mail: interest-help@...

From forum: java.net - java3d interest

Re: Unstable Vista Menus

2009-11-25T07:07:49Z

hi, maybe you can help me.

i executed your example (nice work) and i tried some similar thing with java3d.
i have a borderlayout with a canvas3d and a jpanel.
the jpanel is invisble, but during runtime i want the jpanel to be visible.
but it remains invisible even set to true. how have you managed that
lightweight heavyweight problem?
thx in advance
[Message sent by forum member 'optimusprime1982' ]

http://forums.java.net/jive/thread.jspa?messageID=373319

---------------------------------------------------------------------
To unsubscribe, e-mail: interest-unsubscribe@...
For additional commands, e-mail: interest-help@...

From forum: java.net - java3d interest

Re: TGSI simplification branch

2009-11-25T06:30:32Z

On Wed, 2009-11-25 at 06:28 -0800, michal wrote:

> Keith Whitwell pisze:
> > I've pushed a feature branch with some tgsi simplifications in it. With
> > these I've removed the biggest remaining oddities of that language, and
> > it's getting to a place where I'm starting to be happy with it as a
> > foundation for future work.
> >
> > Most of the surprising stuff like multiple negate flags, etc, is gone
> > now, and the core tokens are quite a bit easier to understand than in
> > previous iterations.
> >
> > I've still got my eye on reducing the verbosity of the names in the
> > tgsi_parse.h "FullToken" world, and promoting the tgsi_any_token union
> > into p_shader_tokens.h.
> >
> > It would be good if people can review the interface changes and provide
> > feedback, and also test out their drivers on this branch. I've done
> > minimal softpipe testing so far but will do more over the next few days.
> >
> >
>
> All looks good to me, I'm happy somebody had the guts to cut off all the
> cruft in one shot.
>
> I see some compile errors on windows build -- I will fix those along
> with other minor bugs I have spotted.
>
> Now, looking at the interface, I'm thinking about removing some more tokens.
>
> 1) Remove tgsi_dimension and use tgsi_src_register directly with some
> well-defined constraints.
>
> 2) Do the same to tgsi_instruction_predicate. Really, it's just an
> optional src operand with some restrictions.

Interesting. I'd be keen to see a patch.

Keith

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: TGSI simplification branch

2009-11-25T06:28:33Z

Keith Whitwell pisze:

> I've pushed a feature branch with some tgsi simplifications in it. With
> these I've removed the biggest remaining oddities of that language, and
> it's getting to a place where I'm starting to be happy with it as a
> foundation for future work.
>
> Most of the surprising stuff like multiple negate flags, etc, is gone
> now, and the core tokens are quite a bit easier to understand than in
> previous iterations.
>
> I've still got my eye on reducing the verbosity of the names in the
> tgsi_parse.h "FullToken" world, and promoting the tgsi_any_token union
> into p_shader_tokens.h.
>
> It would be good if people can review the interface changes and provide
> feedback, and also test out their drivers on this branch. I've done
> minimal softpipe testing so far but will do more over the next few days.
>
>

All looks good to me, I'm happy somebody had the guts to cut off all the
cruft in one shot.

I see some compile errors on windows build -- I will fix those along
with other minor bugs I have spotted.

Now, looking at the interface, I'm thinking about removing some more tokens.

1) Remove tgsi_dimension and use tgsi_src_register directly with some
well-defined constraints.

2) Do the same to tgsi_instruction_predicate. Really, it's just an
optional src operand with some restrictions.

Thanks.

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

MESA_multithread_makecurrent (Was: Video Hackfest conclusions)

2009-11-25T06:24:23Z

On Tue, 2009-11-24 at 03:47 -0800, Benjamin Otte wrote:

> == locking in Mesa ==
> The GLX specification does not allow binding the same GLX context in
> multiple threads at the same time. This is a requirement for both
> Cairo's and GStreamer's threading model. An extension was proposed and
> initial code developed to support this requirement in the same way as
> Apple's GL does by default. Windows does not support this and would
> require potentially expensive fallback code.
> Links:
> * http://people.freedesktop.org/~anholt/MESA_multithread_makecurrent.spec
> Actions:
> * Get review for suggested extension and include it in future Mesa releases
> * Make Windows developers investigate the situation

Erik,

I think this extension will be quite useful.

My only concern with the current wording of
MESA_multithread_makecurrent is that the extension will affect all
applications, which will break any GL test or application that
checks/relies on glXMakeContextCurrent failing when binding the same
context to multiple threads. Perhaps such test/application doesn't
exist, but if such app is known or likely to exist then it might be
preferable to have the old behavior by default and require the
applications that want to use this to feature the new behavior.

Windows WGL spec is worded pretty similarly to GLX in this regard [1],
and the extension could easily be modified to cover WGL too if wasn't
the existence opengl32.dll. opengl32.dll implements all this logic and
comes bundled with Windows, so it is virtually impossible for IHVs to
change these semantics. The only way I see to make it work would be for
the app to call wglGetProcAddress for every GL function it uses, and add
a new "wglMakecurrentMESA" entrypoint that would do what we want...
quite ugly.

Jose

[1] http://msdn.microsoft.com/en-us/library/dd374387(VS.85).aspx

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: Unstable Vista Menus

2009-11-25T05:38:48Z

Finally found the fix - add these lines:

JPopupMenu.setDefaultLightWeightPopupEnabled( false );
ToolTipManager.sharedInstance().setLightWeightPopupEnabled(false);

before the NetBeans generated line:

initComponents();

Dave

ps I would still like to know if StarryOrbits works under Linux or Mac.
[Message sent by forum member 'dwsubc' ]

http://forums.java.net/jive/thread.jspa?messageID=373304

---------------------------------------------------------------------
To unsubscribe, e-mail: interest-unsubscribe@...
For additional commands, e-mail: interest-help@...

From forum: java.net - java3d interest

Re: Shape3D Picking Problems: Very bad accuracy in Java3D?

2009-11-25T03:13:12Z

Ok, I understand the problem now. getClosestIntersectionPoint I believe does only get the closest vertex point. So if your mesh is made up of a few large quads or triangles you might be getting the same closest vertex points depending on where your arrows are.

I assume you are using PickInfo.PICK_GEOMETRY and the flag PickInfo.CLOSEST_INTERSECTION_POINT. If you also use the flag PickInfo.CLOSEST_GEOM_INFO you can get access to the intersection info using getIntersectionInfos(). With the IntersectionInfo there is a method getIntersectionPoint(). I am not sure whether this point is any different, I have never used it, but try it out and see. Basically you want the intersection point of the ray through the polygon. If Java3D work this out for you, you can do it manually, you will need to use the intersection of a line through a plane where there are many examples of the equation online, e.g. http://local.wasp.uwa.edu.au/~pbourke/geometry/planeline/
To do this you need the normal of the intersected geometry, any point on the intersected face and the line geometry all of which you can gain access to.

That is indeed strange about having to repeat the picking using the old picking package, I have never had any problems using pickfast and the closest vertex has always been good enough for my needs. Hope this helps.
[Message sent by forum member 'chrismcband' ]

http://forums.java.net/jive/thread.jspa?messageID=373288

---------------------------------------------------------------------
To unsubscribe, e-mail: interest-unsubscribe@...
For additional commands, e-mail: interest-help@...

From forum: java.net - java3d interest

Re: [gst-devel] Video Hackfest conclusions

2009-11-25T02:56:52Z

On 11/24/2009 12:47 PM, Benjamin Otte wrote:

> = Timeline =
> The end goal of all of this is to get the code into users' hands
> quickly, but allow a smooth and non-disrupting upgrade. Desired dates
> have been put forth to achieve this:
> * End of January: release new versions of Cairo and Pixman that
> contain the new APIs reuired by GStreamer
> * End of March: X server and Mesa releases are due - make sure the
> required Mesa extension is part of this. Ideally XrenderPutImage would
> be included, too.
> * April/May, after next Fedora/Ubuntu releases: Merge Cairo support
> into gst-plugins-base and start porting elements to it. Encourage
> application developers (browser, Flash players) to make use of the new
> APIs
> * October: another Fedora/Ubuntu release that switches all users to
> the new APIs

thanks for the conclusion. one thing which would be very useful (for me
and may be others).
Fedora/Ubuntu is a very fast moving target, but i understood it as a
good target for gstreamer. although some people (like me) would like to
use a more robust os like rhel/centos. but currently it's not possible
to update gstreamer to newer version on rhel/centos since gstreamer
depend on glib/gtk which is too old in rhel for newer gstreamer.
as (probably) rhel-6 will be out in 2010 it means that version of
glib/gtk/cairo which will be in rhel-6 will remain for many years in the
rhel/centos-6.
so if i dare to ask that gtk/glib/cairo version in rhel-6 should have to
enough for these features. ie. rise the required gtk/glib/cairo version
in the current gstreamer to be enough for later version. otherwise these
features can't be used in rhel-6 for many years.
thanks in advance.
regards.

--
Levente "Si vis pacem para bellum!"

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: st_shader-varients merge tomorrow

2009-11-25T02:11:09Z

Hi Roland,

r300g doesn't use *semantic_index at all and due to this some shaders are broken on it. Merging the branch probably can't make it worse, so go for it. I already have fixes for this driver here and I am about to send them soon.

Marek

On Wed, Nov 25, 2009 at 4:47 AM, Roland Scheidegger <sroland@...> wrote:

I'm planning to merge st_shader-varients branch to master tomorrow.

This should not adversely affect drivers, unless they rely on generics
inputs/outputs semantic_index always starting at 0 without holes
(something that they shouldn't do but it would have worked previously).
Feedback for hw drivers welcome, I'll try i915 myself, but I can't test
the others, though some quick glance seemed to suggest they should be ok.

Roland

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: Mesa 7.6.1 release candidate 1

2009-11-24T21:01:36Z

On Tue, Nov 24, 2009 at 3:55 PM, tom fogal <tfogal@...> wrote:

> Dan Nicholson <dbn.lists@...> writes:
>> On Tue, Nov 24, 2009 at 12:37 PM, tom fogal <tfogal@...> wrote:
>> > Dan Nicholson <dbn.lists@...> writes:
>> >> On Thu, Nov 19, 2009 at 8:28 AM, Brian Paul <brianp@...> wrote:
>> >> > tom fogal wrote:
>> >> >> Brian Paul <brianp@...> writes:
>> >> >>> Please test and report any problems ASAP. If there aren't any issues
>> >> >>> we'd like to release 7.6.1 on Friday or Saturday.
>> > [snip]
>> >> >> Secondly, the AIX build is failing in progs/Makefile.
>> >
>> > PROGRAM_DIRS =
>> > ^ (there's a space here [1] ;)
>> >
>> > in a config/ though. *shrug*. As I said, it doesn't matter much for
>> > me.
>>
>> Actually, I don't know why this is a problem. On my system with GNU
>> make (which you're using, too) having just spaces after the = gets
>> stripped.
>
> We're using the native make on AIX, not GNU make.
>
> Speaking of which, I guess AIX make *does* support $(strip ...).
>
>> $ cat > Makefile << "EOF"
>> DIRS = # bunch of spaces here
>> SUBDIRS = $(DIRS)
>> dirs:
>> @echo '"$(SUBDIRS)"'
>> @test -n "$(SUBDIRS)" && echo non-empty || echo empty
>> EOF
>> $ make dirs
>
> This test does in fact work fine.
>
> However, if you do:
> @for d in "$(SUBDIRS)" ; do \
> echo "$$d" \
> done
> after the @test -n line, AIX make chokes.
>
>> What is the actual error you're seeing?
>
> -bash-3.00$ cat Makefile
> DIRS =
> SUBDIRS = $(DIRS)
> dirs:
> @echo '"$(SUBDIRS)"'
> @test -n "$(SUBDIRS)" && echo non-empty || echo empty
> @for dir in $(SUBDIRS) ; do \
> echo "$$dir" \
> done
> -bash-3.00$ make
> ""
> empty
> /usr/bin/sh: 0403-057 Syntax error at line 1 : `;' is not expected.
> make: 1254-004 The error code from the last command is 2.
>
>
> Stop.
> -bash-3.00$
>
> If I quote the $(SUBDIRS) in the @for line, I get the error "`for' is
> not matched.".

Yeah, both errors are because of the AIX shell. It's not as resilient
to poorly constructed loops as bash, I guess.

>> 31f7e8efb25a77e3bdfb6e9850cf31e339060976 is definitely correct as the
>> quotes would cause all the directories to be a single argument. Can
>> you try the automake way, which I believe is used to workaround this
>> exact issue? It constructs the for loop using a shell variable instead
>> of a make variable.
>>
>> subdirs:
>> @list='$(SUBDIRS)'; for dir in $$list; do echo $$dir; done
>
> This works great with AIX (and of course GNU) make.
>
>> >> Tom, what arguments do you pass to configure?
>> >
>> > FWIW on AIX we do something like this:
>> >
>> > ./configure CC=xlc CXX=xlC CFLAGS=-qcpluscmt -qlanglvl=extc99
>> > -DUSE_MGL_NAMESPACE CXXFLAGS=-DUSE_MGL_NAMESPACE
>> > --prefix=/usr/common/homes/f/fogal1/sw
>> > --without-demos --with-driver=xxx --disable-gallium --with-max-width=16384
>> > --with-max-height=16384 --disable-glw --disable-glu --disable-egl
>> >
>> > Except it happens twice, the first time with xxx=xlib and the latter
>> > with xxx=osmesa. If you're *really* curious, see the function
>> > 'build_mesa' in:
>> >
>> > http://portal.nersc.gov/svn/visit/trunk/src/svn_bin/build_visit
>>
>> That's strange since --without-demos will substitute PROGRAM_DIRS="".
>> I think there's something else funky going on.
>
> The issue is SUBDIRS = $(PROGRAM_DIRS) in progs/Makefile, I think (but
> haven't verified). That seems to cause AIX make to think that SUBDIRS
> is a string with a single space, which != the empty string, and thus
> the test -n fails.
>
> I like your automake-esque solution. How's the attached patch? Note
> that I tested the approach but not Mesa on AIX yet (please don't
> apply).

Yeah, that's great. One of my _really_ rainy day projects was to
convert all the mesa recursion targets to use the automake style since
we know that's been thrown at tons of systems out in the wild.

Reviewed-by: Dan Nicholson <dbn.lists@...>

--
Dan

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: Can mesa library be built on aix 5.3 platform

2009-11-24T20:03:06Z

1. Please CC the list in replies.
2. You've again sent me a mail in a `gbk' charset, which was really
in iso-8859-1 or something similar, and all of that was needlessly
encoded via base64. This is the last reply you'll get from me
without fixing your mail client.

majun <majun.china@...> writes:
> Glad to receive your Emai. From your Email, if I can get these
> informations:
> 1. Before mesa_7_6, there is no mesa library which can support aix
> 5.3.

AFAICT, Mesa used to support AIX. I gather that few people use AIX,
so the support atrophied a bit until some recent testing revealed the
build was broken.

I haven't been here all that long though.

> 2. Only mesa_7_6_branch can support aix 5.3, but this version is
> latest, and I can not download, so I will wait.

You can download the current Mesa 7.6 tip from git. See:

http://www.mesa3d.org/repository.html

> In your Emai, I can not understand what isOSMesa is, Can you explain
> it for me?

I can but won't bother. Google is your friend.

I apparently don't have a good link on ML etiquette bookmarked. This
will do in the meantime:

http://www.freebsd.org/doc/en/articles/mailing-list-faq/etiquette.html

though it is of course specific to FreeBSD.

-tom

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

st_shader-varients merge tomorrow

2009-11-24T19:47:05Z

I'm planning to merge st_shader-varients branch to master tomorrow.

This should not adversely affect drivers, unless they rely on generics
inputs/outputs semantic_index always starting at 0 without holes
(something that they shouldn't do but it would have worked previously).
Feedback for hw drivers welcome, I'll try i915 myself, but I can't test
the others, though some quick glance seemed to suggest they should be ok.

Roland

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: FreeGLUT re-glutInit possible?

2009-11-24T17:37:23Z

John F. Fay wrote:
> Chris,
>
> There is a "glutExit" call that is supposed to undo all the
> "glutInit" stuff. Then you can call "glutInit" again and carry on merrily.
>
> - John

Great to know! It appears to be in FreeGLUT 2.6 so that will solve
a big piece of the problem when I get to upgrade from my current
2.4 version...

Thanks,
Chris

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Freeglut-developer mailing list
Freeglut-developer@...
https://lists.sourceforge.net/lists/listinfo/freeglut-developer

From forum: freeglut-developer

Re: FreeGLUT re-glutInit possible?

2009-11-24T17:36:09Z

Larry E. Ramey wrote:
> You can call the event loop yourself. It will execute 1 loop then return
> to you. Put this in your own thread that you stall and restart at will.
> DO NOT call glutInit twice as it is attempting to build an OpenGL context.
>
> Larry

That is basically what I am doing, pumping the loop with calls to
glutMainLoopEvent(). The problem is when I exit the wrong way,
defined as what I am doing now, I get in a situation where further
GLUT functions give the error that glutInit() needs to be called
first. Then when I call it, blooey! Still diagnosing this one
since it is likely that I am doing something wrong in the runtime
event processing and/or callbacks.

--Chris

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Freeglut-developer mailing list
Freeglut-developer@...
https://lists.sourceforge.net/lists/listinfo/freeglut-developer

From forum: freeglut-developer

Re: Mesa 7.6.1 release candidate 1

2009-11-24T16:34:56Z

tom fogal <tfogal@...> writes:

> Dan Nicholson <dbn.lists@...> writes:
> > On Tue, Nov 24, 2009 at 12:37 PM, tom fogal <tfogal@...> wrot=
> e:
> > > Dan Nicholson <dbn.lists@...> writes:
> > >> On Thu, Nov 19, 2009 at 8:28 AM, Brian Paul <brianp@...> wrote=
> :
> > >> > tom fogal wrote:
> > >> >> Brian Paul <brianp@...> writes:
> > >> >> Secondly, the AIX build is failing in progs/Makefile.
> > >
> > 31f7e8efb25a77e3bdfb6e9850cf31e339060976 is definitely correct as the
> > quotes would cause all the directories to be a single argument. Can
> > you try the automake way, which I believe is used to workaround this
> > exact issue? It constructs the for loop using a shell variable instead
> > of a make variable.
>
> > subdirs:
> > @list=3D'$(SUBDIRS)'; for dir in $$list; do echo $$dir; done
>
> This works great with AIX (and of course GNU) make.

[snip]
> I like your automake-esque solution. How's the attached patch? Note
> that I tested the approach but not Mesa on AIX yet (please don't
> apply).
>
> From: Tom Fogal <tfogal@...>
> Date: Tue, 24 Nov 2009 16:46:31 -0700
> Subject: [PATCH] RFC: Simplify hackery added to fix AIX build.
[snip]

I just verified on AIX, and this patch works (in our configuration, at
least) fine.

-tom

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: Mesa 7.6.1 release candidate 1

2009-11-24T15:55:37Z

Dan Nicholson <dbn.lists@...> writes:

> On Tue, Nov 24, 2009 at 12:37 PM, tom fogal <tfogal@...> wrote:
> > Dan Nicholson <dbn.lists@...> writes:
> >> On Thu, Nov 19, 2009 at 8:28 AM, Brian Paul <brianp@...> wrote:
> >> > tom fogal wrote:
> >> >> Brian Paul <brianp@...> writes:
> >> >>> Please test and report any problems ASAP. If there aren't any issues
> >> >>> we'd like to release 7.6.1 on Friday or Saturday.
> > [snip]
> >> >> Secondly, the AIX build is failing in progs/Makefile.
> >
> > PROGRAM_DIRS =
> > ^ (there's a space here [1] ;)
> >
> > in a config/ though. *shrug*. As I said, it doesn't matter much for
> > me.
>
> Actually, I don't know why this is a problem. On my system with GNU
> make (which you're using, too) having just spaces after the = gets
> stripped.

We're using the native make on AIX, not GNU make.

Speaking of which, I guess AIX make *does* support $(strip ...).

> $ cat > Makefile << "EOF"
> DIRS = # bunch of spaces here
> SUBDIRS = $(DIRS)
> dirs:
> @echo '"$(SUBDIRS)"'
> @test -n "$(SUBDIRS)" && echo non-empty || echo empty
> EOF
> $ make dirs

This test does in fact work fine.

However, if you do:
@for d in "$(SUBDIRS)" ; do \
echo "$$d" \
done
after the @test -n line, AIX make chokes.

> What is the actual error you're seeing?

-bash-3.00$ cat Makefile
DIRS =
SUBDIRS = $(DIRS)
dirs:
@echo '"$(SUBDIRS)"'
@test -n "$(SUBDIRS)" && echo non-empty || echo empty
@for dir in $(SUBDIRS) ; do \
echo "$$dir" \
done
-bash-3.00$ make
""
empty
/usr/bin/sh: 0403-057 Syntax error at line 1 : `;' is not expected.
make: 1254-004 The error code from the last command is 2.

Stop.
-bash-3.00$

If I quote the $(SUBDIRS) in the @for line, I get the error "`for' is
not matched.".

> 31f7e8efb25a77e3bdfb6e9850cf31e339060976 is definitely correct as the
> quotes would cause all the directories to be a single argument. Can
> you try the automake way, which I believe is used to workaround this
> exact issue? It constructs the for loop using a shell variable instead
> of a make variable.
>
> subdirs:
> @list='$(SUBDIRS)'; for dir in $$list; do echo $$dir; done

This works great with AIX (and of course GNU) make.

> >> Tom, what arguments do you pass to configure?
> >
> > FWIW on AIX we do something like this:
> >
> > ./configure CC=xlc CXX=xlC CFLAGS=-qcpluscmt -qlanglvl=extc99
> > -DUSE_MGL_NAMESPACE CXXFLAGS=-DUSE_MGL_NAMESPACE
> > --prefix=/usr/common/homes/f/fogal1/sw
> > --without-demos --with-driver=xxx --disable-gallium --with-max-width=16384
> > --with-max-height=16384 --disable-glw --disable-glu --disable-egl
> >
> > Except it happens twice, the first time with xxx=xlib and the latter
> > with xxx=osmesa. If you're *really* curious, see the function
> > 'build_mesa' in:
> >
> > http://portal.nersc.gov/svn/visit/trunk/src/svn_bin/build_visit
>
> That's strange since --without-demos will substitute PROGRAM_DIRS="".
> I think there's something else funky going on.

The issue is SUBDIRS = $(PROGRAM_DIRS) in progs/Makefile, I think (but
haven't verified). That seems to cause AIX make to think that SUBDIRS
is a string with a single space, which != the empty string, and thus
the test -n fails.

I like your automake-esque solution. How's the attached patch? Note
that I tested the approach but not Mesa on AIX yet (please don't
apply).

-tom

From 8e97a68d0f85d72f81f4652dcfd0cb1665beec87 Mon Sep 17 00:00:00 2001
From: Tom Fogal <tfogal@...>
Date: Tue, 24 Nov 2009 16:46:31 -0700
Subject: [PATCH] RFC: Simplify hackery added to fix AIX build.

Borrow an idiom from the GNU build system which can handle `for'
loops over empty lists.
---
progs/Makefile | 26 +++++++++++---------------
1 files changed, 11 insertions(+), 15 deletions(-)

diff --git a/progs/Makefile b/progs/Makefile
index d5852fa..5bc444e 100644
--- a/progs/Makefile
+++ b/progs/Makefile
@@ -4,7 +4,7 @@ TOP = ..

include $(TOP)/configs/current

-SUBDIRS = "$(strip "$(PROGRAM_DIRS)")"
+SUBDIRS = $(PROGRAM_DIRS)

default: message subdirs
@@ -15,22 +15,18 @@ message:

subdirs:
- @if test -n "$(SUBDIRS)" ; then \
- for dir in $(SUBDIRS) ; do \
- if [ -d $$dir ] ; then \
- (cd $$dir && $(MAKE)) || exit 1 ; \
- fi \
- done \
- fi
+ @list='$(SUBDIRS)'; for dir in $$list ; do \
+ if [ -d $$dir ] ; then \
+ (cd $$dir && $(MAKE)) || exit 1 ; \
+ fi \
+ done

# Dummy install target
install:

clean:
- -@if test -n "$(SUBDIRS)" ; then \
- for dir in $(SUBDIRS) tests ; do \
- if [ -d $$dir ] ; then \
- (cd $$dir && $(MAKE) clean) ; \
- fi \
- done \
- fi
+ @list='$(SUBDIRS)'; for dir in $$list tests ; do \
+ if [ -d $$dir ] ; then \
+ (cd $$dir && $(MAKE) clean) ; \
+ fi \
+ done
--
1.6.3.3

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Stereo Rendering inverted on one eye

2009-11-24T15:33:16Z

Hi!

I hope, somebody of you knows this special problem.

If I use the regular Stereo Package from Java3D (with a nVidia/PNY Quadro FX card), I get, let's say, Eye Cancer - the reason, the right eye is rendered correctly like the regular view. But the left eye renders the opposite. This means, all objects are still in the correct place, but the objects are rendered the wrong way around: Objects in the back are in the front - and vice versa. I think I have to set a property somewhere - but where???

Thanks a lot for your suggestions!
Björn
[Message sent by forum member 'bjsommer' ]

http://forums.java.net/jive/thread.jspa?messageID=373236

---------------------------------------------------------------------
To unsubscribe, e-mail: interest-unsubscribe@...
For additional commands, e-mail: interest-help@...

From forum: java.net - java3d interest

Re: Shape3D Picking Problems: Very bad accuracy in Java3D?

2009-11-24T15:23:29Z

Okay, so thanks for the Info that the Picking package is obsolute - so I changed it to the new PickFast Package. But the problem remains the same. So I try to point it out precisely.

So I use the VRML loader to load a 3D mesh into my project. It generates a Shape3D (I think it is not so important that it is originally a VRML object for this problem). Let's imagine, it is just a Sphere with many vertices.

Now want to distribute 100 small Shape3Ds onto this Sphere. Originally they are all placed close to each other on a rectangular area. The shapes are arrows, which are all pointing away from the area into the same direction. So they are like normal vectors to the area.

Now we want to place this area onto the sphere and then we want to move the arrows close to the sphere. While doing so, the distances between the arrows on the area should not change dramatically - so we can not just look for a vertex on the sphere and place it there, then the structure of the field of arrows would change drastically.

First we place the area with the arrows slidely outside the Sphere by using the TransformGroup of the area. Now it is floating a small distance across the Sphere.

Now the arrows should be moved downwards, until they meet the Sphere.

For this, we send a PickRay from the top of the arrow down into the direction of the sphere - by using the Normal and the position of each arrow.

Now we should get exactly the point on the sphere, where the arrow should be placed now.

But instead of this, we get a point in the neighbourhood of such a point. And although we are sending 100 different PickRays from 100 different positions into the direction of the sphere, we only get ca. 5 different PickingPoints by using getClosestIntersectionPoint().

P.S.: [i]The first while loop inside the for is redundant, you only need to do it once, doing it 10 times won't change the result unless the scene is animated.[/i]
Yes, but it is a workaround for the old Picking, which sometimes returns only hits when doing the Picking more often, up to 1M times. This is really strange, but works in other methods of my project. I have to check this with the new PickFast now.
[Message sent by forum member 'bjsommer' ]

http://forums.java.net/jive/thread.jspa?messageID=373235

---------------------------------------------------------------------
To unsubscribe, e-mail: interest-unsubscribe@...
For additional commands, e-mail: interest-help@...

From forum: java.net - java3d interest

Re: Mesa 7.6.1 release candidate 1

2009-11-24T15:22:34Z

On Tue, Nov 24, 2009 at 12:37 PM, tom fogal <tfogal@...> wrote:

> Dan Nicholson <dbn.lists@...> writes:
>> On Thu, Nov 19, 2009 at 8:28 AM, Brian Paul <brianp@...> wrote:
>> > tom fogal wrote:
>> >> Brian Paul <brianp@...> writes:
>> >>> Please test and report any problems ASAP. If there aren't any issues
>> >>> we'd like to release 7.6.1 on Friday or Saturday.
> [snip]
>> >> Secondly, the AIX build is failing in progs/Makefile. The
>> >> quoting introduced in 115edf24a9128b79dfa5f30482c990e2cb898357
>> >> and removed in 31f7e8efb25a77e3bdfb6e9850cf31e339060976 was
>> >> important. Otherwise, SUBDIRS will end up being " " or maybe " "
>> >> and test -n will (annoyingly) fail, causing the `for'-over-nothing
>> >> to execute.
>> >
>> > The $(strip ...) syntax is GNU-make specific, I think. If it
>> > causes trouble for anyone we'll revisit it.
>>
>> Sorry, I meant to comment on this earlier, but I was really busy.
>
> It happens, no worries.
>
>> I think the reason that the make variable ends up with a space
>> in it instead of being empty is the automatic demos checking in
>> configure. I think it would be better to fix it there so that you
>> either get a list of directories or an empty variable. I don't have
>> time to test, but I think this would fix it:
>
> I think you're right, but unfortunately that's not the only change that
> needs to be made. configs/default (IIRC) sets PROGRAM_DIRS as well,
> which I think is the issue Brian ran into that caused him to originally
> revert this part of the patch.
>
> I toyed with the idea of making:
> PROGRAM_DIRS = demos blah whatever
> into:
> PROGRAM_DIRS ="demos blah whatever"
>
> in configs/default, but I felt that was too fragile; anyone that
> then modified configs/default (or a config that included it, where
> they wanted to change this) would need to know about this weird
> stringification-with-no-spaces requirement, and no other variables need
> that behavior.
>
>> - PROGRAM_DIRS="$demos"
>> + PROGRAM_DIRS=`echo $demos | $SED 's/^ *//'`
>
> ... we might consider doing something like this instead of $(strip) in
> the makefile; this seems more portable (i.e. would work w/o GNU make).
>
> ... all that said, we could, of course, just go with something like
> this and ignore the non-autoconf case. I certainly don't care about
> the non-GBS for my work. It leaves the lurking case where one can't:
>
> PROGRAM_DIRS =
> ^ (there's a space here [1] ;)
>
> in a config/ though. *shrug*. As I said, it doesn't matter much for
> me.

Actually, I don't know why this is a problem. On my system with GNU
make (which you're using, too) having just spaces after the = gets
stripped.

$ cat > Makefile << "EOF"
DIRS = # bunch of spaces here
SUBDIRS = $(DIRS)
dirs:
@echo '"$(SUBDIRS)"'
@test -n "$(SUBDIRS)" && echo non-empty || echo empty
EOF
$ make dirs
""
empty

What is the actual error you're seeing?
31f7e8efb25a77e3bdfb6e9850cf31e339060976 is definitely correct as the
quotes would cause all the directories to be a single argument. Can
you try the automake way, which I believe is used to workaround this
exact issue? It constructs the for loop using a shell variable instead
of a make variable.

subdirs:
@list='$(SUBDIRS)'; for dir in $$list; do echo $$dir; done

>> Tom, what arguments do you pass to configure?
>
> FWIW on AIX we do something like this:
>
> ./configure CC=xlc CXX=xlC CFLAGS=-qcpluscmt -qlanglvl=extc99
> -DUSE_MGL_NAMESPACE CXXFLAGS=-DUSE_MGL_NAMESPACE
> --prefix=/usr/common/homes/f/fogal1/sw
> --without-demos --with-driver=xxx --disable-gallium --with-max-width=16384
> --with-max-height=16384 --disable-glw --disable-glu --disable-egl
>
> Except it happens twice, the first time with xxx=xlib and the latter
> with xxx=osmesa. If you're *really* curious, see the function
> 'build_mesa' in:
>
> http://portal.nersc.gov/svn/visit/trunk/src/svn_bin/build_visit

That's strange since --without-demos will substitute PROGRAM_DIRS="".
I think there's something else funky going on.

--
Dan

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: FreeGLUT re-glutInit possible?

2009-11-24T15:09:42Z

You can call the event loop yourself. It will execute 1 loop then return to you. Put this in your own thread that you stall and restart at will. DO NOT call glutInit twice as it is attempting to build an OpenGL context.

Larry

From: Chris Marshall <chm@...>
To: FreeGLUT developers list <freeglut-developer@...>
Sent: Mon, November 23, 2009 6:46:16 PM
Subject: [Freeglut-developer] FreeGLUT re-glutInit possible?

Is it possible to glutLeaveMainLoop() and stop GLUT
in a program and then restart GLUT again to do more
GUI stuff?

When I've tried this the second glutInit() call
always kills the program.

I don't expect this to work for standard GLUT, but
it seems a plausible extension for FreeGLUT since
it can exit the glutMainLoop() and return to
program control...

Cheers,
Chris

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Freeglut-developer mailing list
Freeglut-developer@...
https://lists.sourceforge.net/lists/listinfo/freeglut-developer

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Freeglut-developer mailing list
Freeglut-developer@...
https://lists.sourceforge.net/lists/listinfo/freeglut-developer

From forum: freeglut-developer

Re: Mesa 7.6.1 release candidate 1

2009-11-24T12:37:30Z

Dan Nicholson <dbn.lists@...> writes:
> On Thu, Nov 19, 2009 at 8:28 AM, Brian Paul <brianp@...> wrote:
> > tom fogal wrote:
> >> Brian Paul <brianp@...> writes:
> >>> Please test and report any problems ASAP. Â If there aren't any issues
> >>> we'd like to release 7.6.1 on Friday or Saturday.
[snip]

> >> Secondly, the AIX build is failing in progs/Makefile. Â The
> >> quoting introduced in 115edf24a9128b79dfa5f30482c990e2cb898357
> >> and removed in 31f7e8efb25a77e3bdfb6e9850cf31e339060976 was
> >> important. Â Otherwise, SUBDIRS will end up being " " or maybe " "
> >> and test -n will (annoyingly) fail, causing the `for'-over-nothing
> >> to execute.
> >
> > The $(strip ...) syntax is GNU-make specific, I think. Â If it
> > causes trouble for anyone we'll revisit it.
>
> Sorry, I meant to comment on this earlier, but I was really busy.

It happens, no worries.

> I think the reason that the make variable ends up with a space
> in it instead of being empty is the automatic demos checking in
> configure. I think it would be better to fix it there so that you
> either get a list of directories or an empty variable. I don't have
> time to test, but I think this would fix it:

I think you're right, but unfortunately that's not the only change that
needs to be made. configs/default (IIRC) sets PROGRAM_DIRS as well,
which I think is the issue Brian ran into that caused him to originally
revert this part of the patch.

I toyed with the idea of making:
PROGRAM_DIRS = demos blah whatever
into:
PROGRAM_DIRS ="demos blah whatever"

in configs/default, but I felt that was too fragile; anyone that
then modified configs/default (or a config that included it, where
they wanted to change this) would need to know about this weird
stringification-with-no-spaces requirement, and no other variables need
that behavior.

> - PROGRAM_DIRS="$demos"
> + PROGRAM_DIRS=`echo $demos | $SED 's/^ *//'`

... we might consider doing something like this instead of $(strip) in
the makefile; this seems more portable (i.e. would work w/o GNU make).

... all that said, we could, of course, just go with something like
this and ignore the non-autoconf case. I certainly don't care about
the non-GBS for my work. It leaves the lurking case where one can't:

PROGRAM_DIRS =
^ (there's a space here [1] ;)

in a config/ though. *shrug*. As I said, it doesn't matter much for
me.

> Tom, what arguments do you pass to configure?

FWIW on AIX we do something like this:

./configure CC=xlc CXX=xlC CFLAGS=-qcpluscmt -qlanglvl=extc99
-DUSE_MGL_NAMESPACE CXXFLAGS=-DUSE_MGL_NAMESPACE
--prefix=/usr/common/homes/f/fogal1/sw
--without-demos --with-driver=xxx --disable-gallium --with-max-width=16384
--with-max-height=16384 --disable-glw --disable-glu --disable-egl

Except it happens twice, the first time with xxx=xlib and the latter
with xxx=osmesa. If you're *really* curious, see the function
'build_mesa' in:

http://portal.nersc.gov/svn/visit/trunk/src/svn_bin/build_visit

-tom

[1] As a professor I had once liked to say, "It's hard to see nothing."

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: Mesa 7.6.1 release candidate 1

2009-11-24T12:14:21Z

On Thu, Nov 19, 2009 at 8:28 AM, Brian Paul <brianp@...> wrote:

> tom fogal wrote:
>> Brian Paul <brianp@...> writes:
>>> Please test and report any problems ASAP. If there aren't any issues
>>> we'd like to release 7.6.1 on Friday or Saturday.
>>
>> We've had the attached AIX patch locally for a while now. Sorry I
>> didn't send it earlier. I wrote a small C program to compare it to
>> Linux's bswap_32 and it seems to work, but I'll note it's untested in
>> Mesa; for build-dep reasons, we need an Xlib-based Mesa, but we never
>> call that code at runtime. Further, I inserted #error blah in the
>> define used here, and the Mesa build succeeded... so this is obviously
>> never even called in the configurations I am concerned with.
>>
>> Secondly, the AIX build is failing in progs/Makefile. The quoting
>> introduced in 115edf24a9128b79dfa5f30482c990e2cb898357 and removed in
>> 31f7e8efb25a77e3bdfb6e9850cf31e339060976 was important. Otherwise,
>> SUBDIRS will end up being " " or maybe " " and test -n will
>> (annoyingly) fail, causing the `for'-over-nothing to execute.
>>
>> The (second) attached patch fixes it, I believe in both cases. At
>> least, with Mesa-7.6.1-rc1 I get the error mentioned in 31f7.. from
>> `make linux', and now I don't (and it still works on AIX).
>
> OK, I've applied both patches.
>
> The $(strip ...) syntax is GNU-make specific, I think. If it causes
> trouble for anyone we'll revisit it.

Sorry, I meant to comment on this earlier, but I was really busy. I
think the reason that the make variable ends up with a space in it
instead of being empty is the automatic demos checking in configure. I
think it would be better to fix it there so that you either get a list
of directories or an empty variable. I don't have time to test, but I
think this would fix it:

diff --git a/configure.ac b/configure.ac
index cc588d5..5ffe93c 100644
--- a/configure.ac
+++ b/configure.ac
@@ -486,7 +486,7 @@ yes)
test -d "$srcdir/progs/$demo" || \
AC_MSG_ERROR([Program directory '$demo' doesn't exist])
done
- PROGRAM_DIRS="$demos"
+ PROGRAM_DIRS=`echo $demos | $SED 's/^ *//'`
;;
esac

Tom, what arguments do you pass to configure?

--
Dan

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: Can mesa library be built on aix 5.3 platform

2009-11-24T10:04:26Z

Please send your mails in plain text, preferably without any encoding.
Your base64 ends up raw in my replies, so I've just deleted it.

Anyway, I *just* fixed AIX 5.3 support last week. Use git master, git
mesa_7_6_branch, or just wait for the 7.6.1 release. I'd love to hear
how it works for you; we're only using OSMesa, so I think a generic
libGL doesn't see much testing on AIX.

Cheers,

-tom

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Unstable Vista Menus

2009-11-24T07:50:54Z

Hi,

I have almost finished my animated 3D star cluster simulation viewer and a semi-working prototype is available here.

http://www.manybody.org/wiki/index.php/StarryOrbits

This works fine under Windows XP but the menus and dialogs are very dodgy under Vista. The 3D scene sometimes shows up through the menus and dialogs. Can anyone point me towards a possible solution?

For info, StarryOrbits was produced in Netbeans using Swing. The viewer uses a SimpleUniverse object on a Canvas3D which is contained in a JPanel. I made all 3 menus heavyweight to see if that helped but it made no difference.

Many thanks
Dave

ps - I would love to know if StarryOrbits works under other operating systems as well.
[Message sent by forum member 'dwsubc' ]

http://forums.java.net/jive/thread.jspa?messageID=373128

---------------------------------------------------------------------
To unsubscribe, e-mail: interest-unsubscribe@...
For additional commands, e-mail: interest-help@...

From forum: java.net - java3d interest

Re: Cdk-user Digest, Vol 42, Issue 13

2009-11-24T06:45:22Z

Hi Aziz,

What I could see that on my computer with the given SMILES that it takes
about 1.5 minutes to run in Eclipse, so a lot shorter than 15 minutes.
The main bottleneck seems to be this bit:

/*processing for which bonds which are not in resonance*/
for(int number = 0; number < ac.getBondCount() ; number++){
IAtomContainer aa = setAntiFlags(container,ac, number,true);
if(aa != null){
IMoleculeSet ab = gR2.getStructures((IMolecule) aa);

For each bond, setAntiFlags is called and StructureResonanceGenerator's
method getStructures is called. Each call can take seconds, and that
adds up to more than a minute and a half.

I can't be of much more help, perhaps Miguel Rojas would have some
pointers how to possibly speed things up.

cheers,
Mark

> Dear Mark,
>
> > Could you send the source code that runs for 15 minutes to the list? It
> > should be relatively easy to debug it, to find out where things stall.
>
> Thank you for lookng at this issue. here is the jython code I am using
> to calculate partial charge for an atom.
>
>
> sp = smilesParser()
>
> smi =
> /"COC1=CC=C(C[C@H]2N(C)C(=O)[C@H](C)NC(=O)[C@@H](C)NC(=O)[C@H]3CC4=CC=C(O)C(OC5=CC=C(C=C5)C(O)[C@H](N(C)C(=O)[C@H](C)NC2=O)C(=O)N3C)=C4)C=C1"/
>
> mol = sp.parseSimles(smi)
>
> mol = reader.next()
>
> desc = PartialTChargePEOEDescriptor ()
>
> atoms = mol.atoms().iterator()
>
> firstAtom = atoms.next()
>
> pcharge = desc.calculate(firstAtom, mol).getValue().doubleValue()
>
>
>
>
>
> I also went through the code of of the one of the java object that is
> instanced by PartialTChargePEOEDescriptor and one of its methods that
> is called to calculate the partial charges, I found out some bizare code
> I paste here:
>
>
>
> public class GasteigerMarsiliPartialCharges implements IChargeCalculator {
>
> …
>
> public IAtomContainer
> assignGasteigerMarsiliSigmaPartialCharges(IAtomContainer ac, boolean
> setCharge) throws Exception {
>
> ….
>
> double[] q_old = new double[ac.getAtomCount()];
>
> for(int i = 0 ; i < q_old.length ; i++)
>
> q_old[*0*] = 20.0; ////// STRANGE
>
> ….
>
> the index should "i" not "0" if I understand this part of the code.
>
>
> thanks ,
>
>
> AYASRI
>

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Cdk-user mailing list
Cdk-user@...
https://lists.sourceforge.net/lists/listinfo/cdk-user

From forum: cdk-user

TGSI simplification branch

2009-11-24T06:32:29Z

I've pushed a feature branch with some tgsi simplifications in it. With
these I've removed the biggest remaining oddities of that language, and
it's getting to a place where I'm starting to be happy with it as a
foundation for future work.

Most of the surprising stuff like multiple negate flags, etc, is gone
now, and the core tokens are quite a bit easier to understand than in
previous iterations.

I've still got my eye on reducing the verbosity of the names in the
tgsi_parse.h "FullToken" world, and promoting the tgsi_any_token union
into p_shader_tokens.h.

It would be good if people can review the interface changes and provide
feedback, and also test out their drivers on this branch. I've done
minimal softpipe testing so far but will do more over the next few days.

Keith

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Video Hackfest conclusions

2009-11-24T03:47:38Z

Hi,

As you may be aware, we held a video hackfest last week in Barcelona.
Developers met to discuss how best to improve GPU support for video
applications. See http://gstreamer.freedesktop.org/wiki/VideoHackfest
for more details. In particular, you might be interested in the notes
some people took while the hackfest was going on. These can be found
at http://gstreamer.freedesktop.org/wiki/VideoHackfest/Notes

What follows is the results we arrived at after the hackfest. They are
taken from http://gstreamer.freedesktop.org/wiki/VideoHackfest/Conclusions
but certainly deserve wide circulation. We're interested in any
comments (or questions) you might have about them, so please don't
heistate to reply.

If you do reply, please make sure to reduce the recipient list appropriately.

Cheers,
Benjamin,

on behalf of all hackfest paticipants.

== YUV in pixman ==
It has been a goal for a while to make video a first-class citizen in
the image stack. A concrete API proposal on how to integrate video
into pixman was reached and reviewed.
Links:
* http://gstreamer.freedesktop.org/wiki/VideoHackfest/YCbCr_Formats
* http://gstreamer.freedesktop.org/wiki/VideoHackfest/pixman
Actions:
* GStreamer developers to provide a specification of important YUV formats
* Write implementation

== YUV in Cairo ==
To make it easy for applications to use these different formats, the
Cairo API needs to be extended to allow them. A proposal does exist
and has been reviewed, details need to be finalized.
Links:
* http://gstreamer.freedesktop.org/wiki/VideoHackfest/Cairo
Actions:
* Finalize proposal
* Write implementation

== locking in Cairo ==
GStreamer takes a lot of freedom in deciding which threads to schedule
elements in. Most hardware backends however require proper
serialization of commands. The current gst-plugins-cairo and
gst-plugins-gl code provide different, but ugly ways to achieve this.
Cairo's internal thread safety guarantees match these requirements
very well. But the different Cairo backends don't always keep these
guarantees. Interaction with these locking mechanisms from outside
applications is not possible yet either.
Links:
* http://gstreamer.freedesktop.org/wiki/VideoHackfest/Cairo
Actions:
* Improve Cairo backend implementations, in particular X11 and GL
* Expose Cairo's locking API to allow interaction with it
* Write testcases to squash bugs

== locking in Mesa ==
The GLX specification does not allow binding the same GLX context in
multiple threads at the same time. This is a requirement for both
Cairo's and GStreamer's threading model. An extension was proposed and
initial code developed to support this requirement in the same way as
Apple's GL does by default. Windows does not support this and would
require potentially expensive fallback code.
Links:
* http://people.freedesktop.org/~anholt/MESA_multithread_makecurrent.spec
Actions:
* Get review for suggested extension and include it in future Mesa releases
* Make Windows developers investigate the situation

== switch GStreamer to Cairo as default video transport model ==
The current approach to handling video in GStreamer is very outdated.
It does not allow hw-accelerated buffers nor does it provide a unified
API to modify video buffers, which leads to fragmentation and
duplication. It was agreed that using Cairo to provide an abstraction
as a drawing API and abstraction over different backends was a good
idea.
Links:
* http://gstreamer.freedesktop.org/wiki/VideoHackfest/Notes
* http://cgit.freedesktop.org/~company/gst-plugins-cairo
Actions:
* Rework gst-plugins-cairo to match improvements listed in previous points
* Get more review on API and to avoid regressions
* Merge into gst-plugins-base
* Switch plugins gradually to use new Cairo code

== gst-plugins-gl ==
The general consensus was that gst-plugins-gl is a hack. It was
necessary in the past to get things to work, but is not a good way
forward. However, the functionality or performance provided by the
current elements needs to continue working. Developers shared the
opinion that gst-plugins-cairo with cairo-gl surfaces is the best way
to achieve this.
Links:
* http://gstreamer.freedesktop.org/wiki/VideoHackfest/Notes
Actions:
* Get review of cairo-gl/gst-plugins-cairo code from gst-plugins-gl developers
* Port gst-plugins-gl elements to use Cairo early to ensure required
features are available

== XRenderPutImage ==
Currently there is no way to upload video data to the X server for
later use. The XV extension is not sufficient for anything more
complicated than a simple video player.
Links:
* http://gstreamer.freedesktop.org/wiki/VideoHackfest/Cairo
Actions:
* Figure out the best way to move YUV data into hw-accelerated Cairo
surfaces (GL vs X)
* Possibly draft an extension to XRender to handle this in the no-GL case

== JIT in Pixman ==
ORC was investigated as a potential JIT for pixman. A lot of talk has
happened, but no clear conclusions were reached. The idea to share a
JIT with Mesa's GLSL stack was also brought up.
Links:
* http://gstreamer.freedesktop.org/wiki/VideoHackfest/Notes
Actions:
* Continue investigating JIT options
* Prototype ORC integration into Pixman

== Video decoding Acceleration ==
Hardware video decoding is a feature that everybody is very excited
about and wants to integrate into GStreamer as soon as possible. But
developers feel that the current APIs available - the focus was on
VAAPI and VDPAU - to do hardware acceleration fail to integrate well.
No developer managed to create even demo-quality integration with
gst-plugins-cairo, even though quite a few tried. In particular, both
VDPAU and VAAPI...

... do not integrate with existing frameworks
Both libraries fail to use (or allow easy conversion to) existing
objects (GEM handles, GL textures, ...) but only provide access to an
encapsulation. This makes it hard for a flexible framework like
GStreamer to make use of its features without limiting itself.

... do not have a clear scope
In particular VDPAU advertises itself as a video playback framework
using hardware acceleration. Unfortunately, they both limit themselves
to only the formats and functionalities supported by the hardware and
see fullscreen video playback as the single usecase. This needlessly
complicates integration with existing frameworks

... provide a very complicated API
For decoding video, an API like provided by libtheora, libmpeg2 or
even ffmpeg is very simple and allows easy integration with both the
demuxing and the postprocessing/display side. VAAPI and VDPAU require
complex setups and preprocessing of the data to work properly.

... overlap with GL functionality
It was felt that a lot of processing features provided by these
libraries are equally well available using shaders in GL or by
providing GL extensions. It was unclear why it is necessary to use a
separate way to do this.

Links:
* http://gstreamer.freedesktop.org/wiki/VideoHackfest/Notes
* http://gstreamer.freedesktop.org/wiki/VideoHackfest?action=AttachFile&do=view&target=VAinGST.pdf
Actions:
* Get in contact with developers of the APIs
* Encourage revision of public
* Work closer together in integrating the provided functionality into GStreamer

= Timeline =
The end goal of all of this is to get the code into users' hands
quickly, but allow a smooth and non-disrupting upgrade. Desired dates
have been put forth to achieve this:
* End of January: release new versions of Cairo and Pixman that
contain the new APIs reuired by GStreamer
* End of March: X server and Mesa releases are due - make sure the
required Mesa extension is part of this. Ideally XrenderPutImage would
be included, too.
* April/May, after next Fedora/Ubuntu releases: Merge Cairo support
into gst-plugins-base and start porting elements to it. Encourage
application developers (browser, Flash players) to make use of the new
APIs
* October: another Fedora/Ubuntu release that switches all users to
the new APIs

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: Mesa (master): tgsi: Fix POSITION and FACE fragment shader inputs.

2009-11-24T03:02:23Z

Keith Whitwell pisze:

> On Mon, 2009-11-23 at 17:28 -0800, Brian Paul wrote:
>
>> For OpenGL, the front-facing attribute is either 0 (back) or 1 (front) rather than +/-1.
>>
>> I think we'll need to do some additional work (insert a MAD instr?) in the Mesa->TGSI translation to account for this difference. I could dig into that someday...
>>
>
> I'm assuming DX or some other API uses +/-1?
>
> If we define tgsi to use +/-1, then the GL 0/1 version can be reached by
> just saturating. Getting from 0/1 to +/-1 looks like it would be a MAD
> as you say. Probably +/-1 is easy to calculate as the sign of the
> determinant, which would be an intermediate step to calculate GL's
> version.
>
> If it's OK, let's define TGSI's face reg as +/-1, and have Mesa insert
> the saturate if necessary.
>
>

OK, I have documented that as negative/positive since it gives us more
freedom in the future. If it is a problem, we can explicitly say it's
either -1 or +1 later.

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: Pure G3D libs

2009-11-24T02:47:25Z

On Sun, Nov 15, 2009 at 06:56:26PM +0000, Jakob Bornecrantz wrote:
> Hi I'm sorry for taking this long to respond.
> The basic idea is to create API struct that holds the functions for
> the different APIs and then just switch between those in EGL. There
> are also some changes to the in the way that framebuffers works. I
> have started to change the egl_xlib driver but haven't gotten that
> far.
> I have attached a patch series that adds the API and the work in
> progress conversion of egl_xlib. Please feel free to continue on the
> egl_xlib stuff.
I was busy with the opengl-es branch last week, and don't have the time
to think about it until now. Sorry for the delay.

I like the idea of st_framebuffer, and I am thinking about how EGLImage
can be supported. That is, a texture image or a renderbuffer can be
used to create an EGLImage which is shared with other APIs. And through
extensions such as GL_OES_EGL_image, an EGLImage (created from a
NativePixmap or other APIs' resources) can be used as a texture image or
renderbuffer.

It seems that an EGLImage can be mapped to an st_framebuffer without
flush_front. Then what's lacking is a way for client APIs to look up
the backing st_framebuffer of an EGLImage, as needed by GL_OES_EGL_image
and others. Maybe something like

struct st_co_api
{
struct st_framebuffer *(*lookup_framebuffer)(void *eglimage);
}

and a way to pass it to the client APIs?

--
Regards,
olv

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: Shape3D Picking Problems: Very bad accuracy in Java3D?

2009-11-24T02:17:47Z

Hmm, I'm not really sure what you are trying to do here? Can you explain exactly what you are trying to do, it sounds like you want the position of some objects loaded using the vrml loader. Why do you need picking for this? Can't you process the loaded scene in code to work out positions?. I don't know why you need to loop through the shapes, you are changing the coneray for each shape, is this because you need to orient the loaded shapes to face somewhere? The first while loop inside the for is redundant, you only need to do it once, doing it 10 times won't change the result unless the scene is animated.

Also you should use the picking tools from the pickfast package instead, the old picking package is obsolete. Please post some more details on what you are trying to achieve so we can help.
[Message sent by forum member 'chrismcband' ]

http://forums.java.net/jive/thread.jspa?messageID=373098

---------------------------------------------------------------------
To unsubscribe, e-mail: interest-unsubscribe@...
For additional commands, e-mail: interest-help@...

From forum: java.net - java3d interest

Re: PickTranslateBehavior: Change axis

2009-11-24T02:16:56Z

thx a lot
but there is some code missing..
[Message sent by forum member 'exfalso' ]

http://forums.java.net/jive/thread.jspa?messageID=373097

---------------------------------------------------------------------
To unsubscribe, e-mail: interest-unsubscribe@...
For additional commands, e-mail: interest-help@...

From forum: java.net - java3d interest

Can mesa library be built on aix 5.3 platform

2009-11-24T02:13:18Z

Hi everyone:

I have downloaded mesa 7.6 and 7.5.1, and both can be built on Linux platform, but when I build these two on aix platform, there are always errors, such as "<sys/endian.h> not found " 、CC not found etc. Can anybody tell me if mesa library can be built on aix 5.3 platform? Or can anybody give me some information about building mesa on aix platform?

Thanks a lot!!!

majun

看陆川杨幂新片《琴棋书画》，品网易3D国韵网游《天下贰》
------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: Mesa (master): tgsi: Fix POSITION and FACE fragment shader inputs.

2009-11-24T01:49:51Z

On Mon, 2009-11-23 at 17:28 -0800, Brian Paul wrote:
> For OpenGL, the front-facing attribute is either 0 (back) or 1 (front) rather than +/-1.
>
> I think we'll need to do some additional work (insert a MAD instr?) in the Mesa->TGSI translation to account for this difference. I could dig into that someday...

I'm assuming DX or some other API uses +/-1?

If we define tgsi to use +/-1, then the GL 0/1 version can be reached by
just saturating. Getting from 0/1 to +/-1 looks like it would be a MAD
as you say. Probably +/-1 is easy to calculate as the sign of the
determinant, which would be an intermediate step to calculate GL's
version.

If it's OK, let's define TGSI's face reg as +/-1, and have Mesa insert
the saturate if necessary.

Keith

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Mesa3d-dev mailing list
Mesa3d-dev@...
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

From forum: mesa3d-dev

Re: How to run a 2-node application ?

2009-11-23T23:54:03Z

On Mon, Nov 23, 2009 at 6:16 PM, <wangfeiz@...> wrote:

> I just setup a 2-node application.
> But the server machine stucked, I did not see any thing on the other.

There is no default launch mechanism like ssh on Windows. Search the
mailing list archives for my recommendations.

Maybe one of the Windows folks on the list can write up a How-To?

Cheers,

Stefan.

--
http://www.eyescale.ch
http://www.equalizergraphics.com
http://www.linkedin.com/in/eilemann

_______________________________________________
eq-dev mailing list
eq-dev@...
http://www.equalizergraphics.com/cgi-bin/mailman/listinfo/eq-dev
http://www.equalizergraphics.com

From forum: Equalizer - Parallel Rendering