chromium-users — Mailing list for all users of the Chromuim rendering system

Re: [Chromium-users] Crash with multitile-sort

[Brian P. <bri...@tu...>]

Paul Melis wrote:
> Paul Melis wrote:
> 
> 
>>Hello,
>>While attempting to create a multi-tilesort configuration based on the 
>>multitilesort.conf file included with CR1.9 two things happen:
>>1. I get a warning about the state tracker being re-initialized
>>2. One of the crservers crashes (backtrace from the coredump shows it to 
>>be in crSPUCopyDispatchTable()).
>> 
>>
> 
> Just a bit more info, the crash is happening in spu_loader/spucopy.c, 
> crSPUCopyDispatchTable().
> In the first for-loop, the assignment temp = dst->copy_of->copyList sets 
> temp to 0x1, which is later used as temp->copy.
> 
> #0  0x5557f040 in crSPUCopyDispatchTable (dst=0x80d62c0, src=0x817d8c8) 
> at spucopy.c:630
> 630                             if (temp->copy == dst)
> (gdb) bt
> #0  0x5557f040 in crSPUCopyDispatchTable (dst=0x80d62c0, src=0x817d8c8) 
> at spucopy.c:630
> #1  0x08087fc5 in crStateDiffAPI (api=0x817d8c8) at state_flush.c:33
> #2  0x080515e0 in crServerInit (argc=1, argv=0xffffd494) at 
> server_main.c:240
> #3  0x08051655 in CRServerMain (argc=1, argv=0xffffd494) at 
> server_main.c:254
> #4  0x08051012 in main (argc=1, argv=0xffffd494) at main.c:22
> (gdb) p temp
> $1 = (struct _copy_list_node *) 0x1
> (gdb) p dst->copy_of
> $2 = (struct _spu_dispatch_table *) 0xffffa520
> (gdb)

The demo was crashing for me too, but in a different way. I've checked 
in fixes to tilesortspu_get.c and state_bufferobject.c that fixed 
things here.  Can you try the current CVS code?

-Brian

Re: [Chromium-users] Crash with multitile-sort

[Paul M. <p.e...@ru...>]

Paul Melis wrote:

>Hello,
>While attempting to create a multi-tilesort configuration based on the 
>multitilesort.conf file included with CR1.9 two things happen:
>1. I get a warning about the state tracker being re-initialized
>2. One of the crservers crashes (backtrace from the coredump shows it to 
>be in crSPUCopyDispatchTable()).
>  
>
Just a bit more info, the crash is happening in spu_loader/spucopy.c, 
crSPUCopyDispatchTable().
In the first for-loop, the assignment temp = dst->copy_of->copyList sets 
temp to 0x1, which is later used as temp->copy.

#0  0x5557f040 in crSPUCopyDispatchTable (dst=0x80d62c0, src=0x817d8c8) 
at spucopy.c:630
630                             if (temp->copy == dst)
(gdb) bt
#0  0x5557f040 in crSPUCopyDispatchTable (dst=0x80d62c0, src=0x817d8c8) 
at spucopy.c:630
#1  0x08087fc5 in crStateDiffAPI (api=0x817d8c8) at state_flush.c:33
#2  0x080515e0 in crServerInit (argc=1, argv=0xffffd494) at 
server_main.c:240
#3  0x08051655 in CRServerMain (argc=1, argv=0xffffd494) at 
server_main.c:254
#4  0x08051012 in main (argc=1, argv=0xffffd494) at main.c:22
(gdb) p temp
$1 = (struct _copy_list_node *) 0x1
(gdb) p dst->copy_of
$2 = (struct _spu_dispatch_table *) 0xffffa520
(gdb)


Paul

-- 
Paul Melis

VR Specialist,
Center for High-Performance Computing & Visualization,
University of Groningen,
The Netherlands

T: +31 50 363 9298
E: p.e...@ru...
W: http://www.rug.nl/rc/hpcv/index

[Chromium-users] Crash with multitile-sort

[Paul M. <p.e...@ru...>]

Hello,
While attempting to create a multi-tilesort configuration based on the 
multitilesort.conf file included with CR1.9 two things happen:
1. I get a warning about the state tracker being re-initialized
2. One of the crservers crashes (backtrace from the coredump shows it to 
be in crSPUCopyDispatchTable()).

This also happens with the included multitilesort.conf. I could not find 
any info on the warning in the archives of the mailing list. Has anybody 
had any luck with the multitilesort configuration?

Regards,
Paul

-- 
Paul Melis

VR Specialist,
Center for High-Performance Computing & Visualization,
University of Groningen,
The Netherlands

T: +31 50 363 9298
E: p.e...@ru...
W: http://www.rug.nl/rc/hpcv/index

Re: [Chromium-users] NV swap groups

[Paul M. <p.e...@ru...>]

Brian Paul wrote:

> Paul Melis wrote:
>
>> Hello Brian,
>>
>> I noticed that the render SPU has support for using NVidia's swap 
>> extensions. On a recent nvidia driver it seems that it is no longer 
>> supported to pass zero as the swap group and barrier number (which 
>> seems to be what chromium is hard-coded to do, ignoring options 
>> passed to the SPU from the configuration script). Has the swap 
>> support worked in the past? Just curious, before I spend time on 
>> updating it for the current driver.
>
>
> Haven't tested it in a while (or with a recent driver) but it used to 
> work.  Feel free to submit a patch...

Alright, will take a look in the near future

Paul



-- 
Paul Melis

VR Specialist,
Center for High-Performance Computing & Visualization,
University of Groningen,
The Netherlands

T: +31 50 363 9298
E: p.e...@ru...
W: http://www.rug.nl/rc/hpcv/index

Re: [Chromium-users] NV swap groups

[Brian P. <bri...@tu...>]

Paul Melis wrote:
> Hello Brian,
> 
> I noticed that the render SPU has support for using NVidia's swap 
> extensions. On a recent nvidia driver it seems that it is no longer 
> supported to pass zero as the swap group and barrier number (which seems 
> to be what chromium is hard-coded to do, ignoring options passed to the 
> SPU from the configuration script). Has the swap support worked in the 
> past? Just curious, before I spend time on updating it for the current 
> driver.

Haven't tested it in a while (or with a recent driver) but it used to 
work.  Feel free to submit a patch...

-Brian

[Chromium-users] NV swap groups

[Paul M. <p.e...@ru...>]

Hello Brian,

I noticed that the render SPU has support for using NVidia's swap 
extensions. On a recent nvidia driver it seems that it is no longer 
supported to pass zero as the swap group and barrier number (which seems 
to be what chromium is hard-coded to do, ignoring options passed to the 
SPU from the configuration script). Has the swap support worked in the 
past? Just curious, before I spend time on updating it for the current 
driver.

Paul

-- 
Paul Melis

VR Specialist,
Center for High-Performance Computing & Visualization,
University of Groningen,
The Netherlands

T: +31 50 363 9298
E: p.e...@ru...
W: http://www.rug.nl/rc/hpcv/index

[Chromium-users] Chromium 1.9 released

[Brian P. <bri...@tu...>]

Chromium 1.9 can be downloaded from SourceForge: 
http://sourceforge.net/project/showfiles.php?group_id=16529

Thanks to everyone that helped with release candidate testing, etc.

-Brian

Re: [Chromium-users] Inconsistent Line Endings for cr/include/cr_threads.h

[Brian P. <bri...@tu...>]

Heh, I guess I had it fixed here a while back but hadn't checked it 
in.  Done.

-Brian

Daniel E. Shipton wrote:
> Line 14 has the culprit ^M :)
> 
> [dshipton@regatta include]$ cp cr_threads.h cr_threads.h.old
> [dshipton@regatta include]$ dos2unix cr_threads.h
> dos2unix: converting file cr_threads.h to UNIX format ...
> [dshipton@regatta include]$ diff cr_threads.h cr_threads.h.old
> 14c14
> < #include "chromium.h"
> ---
>  > #include "chromium.h"
> 
> 
> On 8/31/06, * Brian Paul* <bri...@tu... 
> <mailto:bri...@tu...>> wrote:
> 
>     Daniel E. Shipton wrote:
>      > Might want to convert cr/include/cr_threads.h to Unix line
>     endings. It
>      > currently has both which svn complains about when using chromium
>     cvs as
>      > a vendor branch.
> 
>     Hmmm, I don't see any DOS-style carriage return chars.
> 
>     -Brian
> 
> 
>     -------------------------------------------------------------------------
>     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
>     <http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642>
>     _______________________________________________
>     Chromium-users mailing list
>     Chr...@li...
>     <mailto:Chr...@li...>
>     https://lists.sourceforge.net/lists/listinfo/chromium-users
> 
> 
> 
> 
> -- 
> Daniel E. Shipton
> Software Engineer, Infiscape Corp.

Re: [Chromium-users] Inconsistent Line Endings for cr/include/cr_threads.h

[Daniel E. S. <dan...@gm...>]

Line 14 has the culprit ^M :)

[dshipton@regatta include]$ cp cr_threads.h cr_threads.h.old
[dshipton@regatta include]$ dos2unix cr_threads.h
dos2unix: converting file cr_threads.h to UNIX format ...
[dshipton@regatta include]$ diff cr_threads.h cr_threads.h.old
14c14
< #include "chromium.h"
---
> #include "chromium.h"


On 8/31/06, Brian Paul <bri...@tu...> wrote:
>
> Daniel E. Shipton wrote:
> > Might want to convert cr/include/cr_threads.h to Unix line endings. It
> > currently has both which svn complains about when using chromium cvs as
> > a vendor branch.
>
> Hmmm, I don't see any DOS-style carriage return chars.
>
> -Brian
>
>
> -------------------------------------------------------------------------
> 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
> _______________________________________________
> Chromium-users mailing list
> Chr...@li...
> https://lists.sourceforge.net/lists/listinfo/chromium-users
>



-- 
Daniel E. Shipton
Software Engineer, Infiscape Corp.

Re: [Chromium-users] Inconsistent Line Endings for cr/include/cr_threads.h

[Brian P. <bri...@tu...>]

Daniel E. Shipton wrote:
> Might want to convert cr/include/cr_threads.h to Unix line endings. It 
> currently has both which svn complains about when using chromium cvs as 
> a vendor branch.

Hmmm, I don't see any DOS-style carriage return chars.

-Brian

[Chromium-users] Inconsistent Line Endings for cr/include/cr_threads.h

[Daniel E. S. <dsh...@in...>]

Might want to convert cr/include/cr_threads.h to Unix line endings. It
currently has both which svn complains about when using chromium cvs as a
vendor branch.

-Dan

-- 
Daniel E. Shipton
Software Engineer, Infiscape Corp.

Re: [Chromium-users] Selecting what to tilesort?

[Brian P. <bri...@tu...>]

Paul Melis wrote:
> Hi Brian,
> 
> Brian Paul wrote:
> 
>> Paul Melis wrote:
>>
>>> Hello,
>>> I'm wondering if the following is possible with Chromium (I've 
>>> already got a bit of success, but no glory yet):
>>>
>>> I have an application, a molecular visualisation package, that I'd 
>>> like to run on a 3-screen cylindrical projection setup. I've got a 
>>> chromium configuration to work that does the non-planar tiling. But 
>>> since the application was meant as a normal desktop application, the 
>>> opengl window not only shows molecular structures and stuff in 3D, 
>>> but also contains GUI elements (buttons, text, a command input line, 
>>> etc) drawn in opengl. I don't think these elements, that are probably 
>>> rendered in ortho2d mode, can be made to display correctly on the 
>>> cylindrical screen, nor is it that important. I'd like to "cut out" 
>>> that part of the viewport where the molecules are visible and only 
>>> distribute the geometry visible in that part. Is there a way to set 
>>> up chromium to get this done?
>>
>>
>>
>> You might try writing a new SPU that examined the matrix operations 
>> (glOrtho, glFrustum(), etc) to determine whether the 2D or 3D stuff 
>> was being drawn, then no-op rendering commands as needed.
> 
> 
> I indeed took this route after your suggestion, writing a "filter3d" SPU 
> and it seems to work fine in identifying the right parts to keep and 
> what to throw away. But I'm having trouble with one aspect of this 
> approach.
> 
> As the GUI elements drawn provide functions that are still usable and 
> some even unavailable in the separate GUI app made in Tcl/Tk the 
> application is using, I'd like to keep the GUI elements on the master 
> node where the app is controlled. In effect letting the opengl stream 
> pass through unmodified.
> 
> So my basic idea is:
> 1. run the visualisation app on a master node, where it runs as usual in 
> the eyes of a user. the user can interact using mouse/keyboard and also 
> use the GUI elements drawn in opengl.
> 2. run 3 render nodes that each display a part of the 3D geometry, but 
> not any of the GUI elements
> 
> Chromium is working behind the scenes on the master node intercepting 
> the opengl stream and distributing it to the 3 nodes using the tilesort 
> SPU. The problem is where to put my filter3d SPU.
> If I try to attach it on the slave side, once for each of the 3 nodes, 
> it doesn't seem to have an effect. Could it be that opengl commands such 
> as Viewport and MatrixMode are influenced by the tile-sorting? Perhaps 
> an Ortho call is left out (as that is what the filter3d SPU basically 
> looks for)?

Yes, that may be the case.  The tilesort SPU and crservers do 
manipulate the viewport and projection.


> If I put filter3d before the tilesort SPU then the distributed scene is 
> correct, none of the GUI elements are left in and I only get the "pure 
> 3D" geometry.
> But then I don't have a non-filtered display on the master node anymore :-/
> 
> It feels like I need to split the opengl stream somewhere and send a 
> filtered one to the render nodes and an unfiltered one to the master. Is 
> this possible in an efficient way?

Other people have tried this, with mixed results, by adding another 
tilesort SPU to the mix which uses two identical tiles.


> In the docs it is mentioned that Chromium uses a graph-structure in 
> defining the connections between nodes, but I assume that with SPUs you 
> can only create a chain and not a split?

The only way to split today is with the tilesort SPU.

-Brian

Re: [Chromium-users] Selecting what to tilesort?

[Paul M. <p.e...@ru...>]

Hi Brian,

Brian Paul wrote:

> Paul Melis wrote:
>
>> Hello,
>> I'm wondering if the following is possible with Chromium (I've 
>> already got a bit of success, but no glory yet):
>>
>> I have an application, a molecular visualisation package, that I'd 
>> like to run on a 3-screen cylindrical projection setup. I've got a 
>> chromium configuration to work that does the non-planar tiling. But 
>> since the application was meant as a normal desktop application, the 
>> opengl window not only shows molecular structures and stuff in 3D, 
>> but also contains GUI elements (buttons, text, a command input line, 
>> etc) drawn in opengl. I don't think these elements, that are probably 
>> rendered in ortho2d mode, can be made to display correctly on the 
>> cylindrical screen, nor is it that important. I'd like to "cut out" 
>> that part of the viewport where the molecules are visible and only 
>> distribute the geometry visible in that part. Is there a way to set 
>> up chromium to get this done?
>
>
> You might try writing a new SPU that examined the matrix operations 
> (glOrtho, glFrustum(), etc) to determine whether the 2D or 3D stuff 
> was being drawn, then no-op rendering commands as needed.

I indeed took this route after your suggestion, writing a "filter3d" SPU 
and it seems to work fine in identifying the right parts to keep and 
what to throw away. But I'm having trouble with one aspect of this approach.

As the GUI elements drawn provide functions that are still usable and 
some even unavailable in the separate GUI app made in Tcl/Tk the 
application is using, I'd like to keep the GUI elements on the master 
node where the app is controlled. In effect letting the opengl stream 
pass through unmodified.

So my basic idea is:
1. run the visualisation app on a master node, where it runs as usual in 
the eyes of a user. the user can interact using mouse/keyboard and also 
use the GUI elements drawn in opengl.
2. run 3 render nodes that each display a part of the 3D geometry, but 
not any of the GUI elements

Chromium is working behind the scenes on the master node intercepting 
the opengl stream and distributing it to the 3 nodes using the tilesort 
SPU. The problem is where to put my filter3d SPU.
If I try to attach it on the slave side, once for each of the 3 nodes, 
it doesn't seem to have an effect. Could it be that opengl commands such 
as Viewport and MatrixMode are influenced by the tile-sorting? Perhaps 
an Ortho call is left out (as that is what the filter3d SPU basically 
looks for)?

If I put filter3d before the tilesort SPU then the distributed scene is 
correct, none of the GUI elements are left in and I only get the "pure 
3D" geometry.
But then I don't have a non-filtered display on the master node anymore :-/

It feels like I need to split the opengl stream somewhere and send a 
filtered one to the render nodes and an unfiltered one to the master. Is 
this possible in an efficient way?
In the docs it is mentioned that Chromium uses a graph-structure in 
defining the connections between nodes, but I assume that with SPUs you 
can only create a chain and not a split?

Regards,
Paul


-- 
Paul Melis

VR Specialist,
Center for High-Performance Computing & Visualization,
University of Groningen,
The Netherlands

T: +31 50 363 9298
E: p.e...@ru...
W: http://www.rug.nl/rc/hpcv/index

Re: [Chromium-users] Selecting what to tilesort?

[Brian P. <bri...@tu...>]

Paul Melis wrote:
> Hello,
> I'm wondering if the following is possible with Chromium (I've already 
> got a bit of success, but no glory yet):
> 
> I have an application, a molecular visualisation package, that I'd like 
> to run on a 3-screen cylindrical projection setup. I've got a chromium 
> configuration to work that does the non-planar tiling. But since the 
> application was meant as a normal desktop application, the opengl window 
> not only shows molecular structures and stuff in 3D, but also contains 
> GUI elements (buttons, text, a command input line, etc) drawn in opengl. 
> I don't think these elements, that are probably rendered in ortho2d 
> mode, can be made to display correctly on the cylindrical screen, nor is 
> it that important. I'd like to "cut out" that part of the viewport where 
> the molecules are visible and only distribute the geometry visible in 
> that part. Is there a way to set up chromium to get this done?

You might try writing a new SPU that examined the matrix operations 
(glOrtho, glFrustum(), etc) to determine whether the 2D or 3D stuff 
was being drawn, then no-op rendering commands as needed.

-Brian

[Chromium-users] Selecting what to tilesort?

[Paul M. <p.e...@ru...>]

Hello,
I'm wondering if the following is possible with Chromium (I've already 
got a bit of success, but no glory yet):

I have an application, a molecular visualisation package, that I'd like 
to run on a 3-screen cylindrical projection setup. I've got a chromium 
configuration to work that does the non-planar tiling. But since the 
application was meant as a normal desktop application, the opengl window 
not only shows molecular structures and stuff in 3D, but also contains 
GUI elements (buttons, text, a command input line, etc) drawn in opengl. 
I don't think these elements, that are probably rendered in ortho2d 
mode, can be made to display correctly on the cylindrical screen, nor is 
it that important. I'd like to "cut out" that part of the viewport where 
the molecules are visible and only distribute the geometry visible in 
that part. Is there a way to set up chromium to get this done?

Thanks in advance,
Paul

-- 
Paul Melis

VR Specialist,
Center for High-Performance Computing & Visualization,
University of Groningen,
The Netherlands

T: +31 50 363 9298
E: p.e...@ru...
W: http://www.rug.nl/rc/hpcv/index

Re: [Chromium-users] Windows app_faker

[Alan H. <al...@tu...>]

On Mon, 2006-08-28 at 14:58 -0500, Daniel E. Shipton wrote:
> The Windows app_faker seems to be severely limited in what it can do
> (compared to other platforms).  Applications that have their own data
> to load do not work very well because of the temporary infrastructure
> that Chromium sets up to run it.  How do most people deal with this
> issue?  I would like to see the Chromium not specify the application
> to run in the mothership config file and instead let the user run the
> app_faker /with/the/path/to/what/they/want/to/run .  I don't see why
> this would be a problem(on All platforms). I find it quite unnatural
> to specify the file path to a Windows executable from a config file.
> Thoughts? Comments? 

Dan,

You don't need to use the app_faker at all. As long as the Chromium
opengl32.dll is chosen for the application startup, you can rule out
app_faker just as you can on other platforms.

You are probably best to read the autostart document in the Chromium
documentation to get you going. Which explains the Windows route to
application starting.

Alan.

Re: [Chromium-users] Windows app_faker

[Brian P. <bri...@tu...>]

Daniel E. Shipton wrote:
> The Windows app_faker seems to be severely limited in what it can do 
> (compared to other platforms).  Applications that have their own data to 
> load do not work very well because of the temporary infrastructure that 
> Chromium sets up to run it.  How do most people deal with this issue?  I 
> would like to see the Chromium not specify the application to run in the 
> mothership config file and instead let the user run the app_faker 
> /with/the/path/to/what/they/want/to/run .  I don't see why this would be 
> a problem(on All platforms). I find it quite unnatural to specify the 
> file path to a Windows executable from a config file.  Thoughts? Comments?

Not being a Windows user, I'm not sure of the story behind this.  But 
what you say sounds reasonable.  Maybe someone more familiar with this 
can comment.

I'm hoping to _finally_ release Cr 1.9 this week and, in the interest 
of stability, don't want to make any big changes right now.

-Brian

[Chromium-users] Windows app_faker

[Daniel E. S. <dan...@gm...>]

The Windows app_faker seems to be severely limited in what it can do
(compared to other platforms).  Applications that have their own data to
load do not work very well because of the temporary infrastructure that
Chromium sets up to run it.  How do most people deal with this issue?  I
would like to see the Chromium not specify the application to run in the
mothership config file and instead let the user run the app_faker
/with/the/path/to/what/they/want/to/run .  I don't see why this would be a
problem(on All platforms). I find it quite unnatural to specify the file
path to a Windows executable from a config file.  Thoughts? Comments?
-Dan

Re: [Chromium-users] Question on 'display_string" for tilesort spu

[Qinghuai G. <Qin...@Su...>]

Brian Paul wrote On 08/25/06 10:23,:

>Qinghuai Gao wrote:
>  
>
>>The document on the parameter "display_string" for tilesort spu is:
>>
>>Specifies the DMX display string. This may be used to override the
>>setting of the DISPLAY environment variable.
>>
>>Does this mean I can use this parameter to assign a graphics device for
>>each render spu when I use chromium tilesort spu without DMX on a single
>>computer with more than one graphics device?  If not, any way I can
>>assign a separate graphic device for each render spu when use
>>tilesortspu without DMX?
>>    
>>
>
>By "graphics device" do you mean "X server display"?
>
>To set the render SPU's display set the render SPU's display_string, 
>not the tilesort SPU's display_string.
>  
>
Oops.  Yes., it should be the display_string of render spu.  Seems I
confused myself.  Now I am fully wake up :-) Thank,

>-Brian
>  
>

Re: [Chromium-users] Question on 'display_string" for tilesort spu

[Brian P. <bri...@tu...>]

Qinghuai Gao wrote:
> The document on the parameter "display_string" for tilesort spu is:
> 
> Specifies the DMX display string. This may be used to override the
> setting of the DISPLAY environment variable.
> 
> Does this mean I can use this parameter to assign a graphics device for
> each render spu when I use chromium tilesort spu without DMX on a single
> computer with more than one graphics device?  If not, any way I can
> assign a separate graphic device for each render spu when use
> tilesortspu without DMX?

By "graphics device" do you mean "X server display"?

To set the render SPU's display set the render SPU's display_string, 
not the tilesort SPU's display_string.

-Brian

[Chromium-users] Question on 'display_string" for tilesort spu

[Qinghuai G. <Qin...@Su...>]

The document on the parameter "display_string" for tilesort spu is:

Specifies the DMX display string. This may be used to override the
setting of the DISPLAY environment variable.

Does this mean I can use this parameter to assign a graphics device for
each render spu when I use chromium tilesort spu without DMX on a single
computer with more than one graphics device?  If not, any way I can
assign a separate graphic device for each render spu when use
tilesortspu without DMX?

Thanks,

Qinghuai

Re: [Chromium-users] 64bit debug compiler flag issue

[Brian P. <bri...@tu...>]

Daniel E. Shipton wrote:
> When using -Werror for 64bit builds(default for Linux) you come across a 
> compile failure in printspu_init with gcc 4.1.1:
> 
> Compiling printspu_init.c
> cc1: warnings being treated as errors
> printspu_init.c: In function 'printspu_signal_handler':
> printspu_init.c:47: warning: field precision should have type 'int', but 
> argument 3 has type 'long int'
> gmake[3]: *** [../../built/printspu/Linux/printspu_init.o] Error 1
> 
> on this lines 47 and 84:
>        fprintf(print_spu.fp, "%.*s", endOfVariable - nextVariable + 1, 
> nextVariable);
> 
> This warning (turned error) can be avoided by casting the result of the 
> 3rd arg to an int. fprintf expects you to give it an int if specifying 
> how much to print.  endofVariable and nextVariable are both char* which 
> makes the whole expression return a long int.
>        fprintf(print_spu.fp, "%.*s", (int)(endOfVariable - nextVariable 
> + 1), nextVariable);

I've added the casts in CVS.  Thanks!

-Brian

[Chromium-users] 64bit debug compiler flag issue

[Daniel E. S. <dan...@gm...>]

When using -Werror for 64bit builds(default for Linux) you come across a
compile failure in printspu_init with gcc 4.1.1:

Compiling printspu_init.c
cc1: warnings being treated as errors
printspu_init.c: In function 'printspu_signal_handler':
printspu_init.c:47: warning: field precision should have type 'int', but
argument 3 has type 'long int'
gmake[3]: *** [../../built/printspu/Linux/printspu_init.o] Error 1

on this lines 47 and 84:
       fprintf(print_spu.fp, "%.*s", endOfVariable - nextVariable + 1,
nextVariable);

This warning (turned error) can be avoided by casting the result of the 3rd
arg to an int. fprintf expects you to give it an int if specifying how much
to print.  endofVariable and nextVariable are both char* which makes the
whole expression return a long int.
       fprintf(print_spu.fp, "%.*s", (int)(endOfVariable - nextVariable +
1), nextVariable);

By changing config/Linux.mk and changing -Werror to -Wall will also let this
build.

When it rains, it pours ;)
-Dan

-- 
Daniel E. Shipton
Software Engineer, Infiscape Corp.

Re: [Chromium-users] AMD64 run patch

[Brian P. <bri...@tu...>]

Daniel E. Shipton wrote:
> Attached is a patch I needed to *run* chromium cvs on an AMD64 box 
> running FC5.

Thanks.  I just moved the entry to the later part of the file where 
extensions are handled.

-Brian

Re: [Chromium-users] Window creation questions

[Daniel E. S. <dan...@gm...>]

On 8/24/06, Brian Paul <bri...@tu...> wrote:
>
> Daniel E. Shipton wrote:
> > [snip]
> >
> >      >
> >      > I have seen the render_to_app_window, but this is for the desktop
> >      > application, not the server(render nodes).
> >      > I want the render nodes that are rendering the crapfaker graphics
> >     to be
> >      > essentially injecting the OpenGL stream into this passive
> window.  I
> >      > ulitimately am going to be changing the OpenGL stream to cut out
> the
> >      > view matrices and such but this is the first step before I get to
> >     that.
> >
> >     Have you looked at the reassemble.conf file?  It does tilesort then
> >     puts the results back into the original application window.
> >
> >
> > Brian,
> >
> > I am trying to take the graphics stream from a desktop app and put the
> > stream back into a passive window (chromium would not control this
> > window at all) with the projection matrices changed to my own custom
> > view or views(think 6 sided CAVE) so I can move around within the
> > graphics stream.  Hopefully this gives you a better idea of what I am
> > trying to accomplish.
>
> OK, I didn't know what you meant by a passive window.
>
> The only thing I can think of would be for the app that controls your
> passive window to somehow pass it's ID to the Cr mothership as the
> 'crut' window.  Then, enable the render SPU's 'render_to_crut_window'
> option.


For my prototype that is exactly where I started.  But it didn't work
because it would always complain about the visual s not matching up.
Eventually I just made the renderspu not call SystemMakeCurrent.  That ended
up working but I this solution won't work when there are multiple windows
opening on a display.

The other thing I am trying to accomplish is removing the mothership from
the equation as much as possible.  So passing things to the mothership will
not be very viable. I just want to multiplex the stream out to render nodes.
So the mothership has little configuration to know about.

-Dan