Re: [Flightgear-devel] RenderTexture bug
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Ampere K. Hardraade schrieb: However, there is some strange coloring issue: http://www.cs.yorku.ca/~cs233144/fgfs-screen-001.jpg You shouldn't take drugs and then fly!!! CU, Christian :) -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (MingW32) iD8DBQFDnqSplhWtxOxWNFcRAnsuAJ0RE85gDD2bwTX4gSnB/fyP0s9YlgCgh5mf aX6PZcM5x5jLliQ7ou7YUaA= =zu2G -END PGP SIGNATURE- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On December 13, 2005 01:28 am, Mathias Fröhlich wrote: Well I swapped the minor and major aguments ... Yep. That line that you have added works. That one, appart from the color issue? Other than the color issue, the simulation seems to run fine. Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On December 13, 2005 10:38 pm, Ampere K. Hardraade wrote: That one, appart from the color issue? Other than the color issue, the simulation seems to run fine. Ampere Actually, no. I've just noticed another problem. http://www.students.yorku.ca/~ampere/fgfs-screen-002.jpg Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On Donnerstag 08 Dezember 2005 02:08, Ampere K. Hardraade wrote:
https://bugs.freedesktop.org/show_bug.cgi?id=5142
So, the problem seems to be that pbuffer is not supported?
Can you test the attached patch?
Greetings
Mathias
--
Mathias Fröhlich, email: [EMAIL PROTECTED]
Index: simgear/screen/RenderTexture.cpp
===
RCS file: /var/cvs/SimGear-0.3/SimGear/simgear/screen/RenderTexture.cpp,v
retrieving revision 1.15
diff -u -r1.15 RenderTexture.cpp
--- simgear/screen/RenderTexture.cpp 27 Nov 2005 09:48:51 - 1.15
+++ simgear/screen/RenderTexture.cpp 12 Dec 2005 18:01:44 -
@@ -1820,6 +1820,9 @@
}
#elif defined( __APPLE__ )
#else
+Display* dpy = glXGetCurrentDisplay();
+int minor = 0, major = 0;
+glXQueryVersion(dpy, minor, major);
// First try the glX version 1.3 functions.
glXChooseFBConfigPtr = (glXChooseFBConfigProc)SGLookupFunction(glXChooseFBConfig);
@@ -1829,7 +1832,8 @@
glXDestroyPbufferPtr = (glXDestroyPbufferProc)SGLookupFunction(glXDestroyPbuffer);
glXQueryDrawablePtr = (glXQueryDrawableProc)SGLookupFunction(glXQueryDrawable);
-if (glXChooseFBConfigPtr
+if (((1 = major 3 = minor) || 2 = major)
+glXChooseFBConfigPtr
glXCreatePbufferPtr
glXGetVisualFromFBConfigPtr
glXCreateContextPtr
@@ -1838,6 +1842,12 @@
glXVersion1_3Present = true;
else
{
+int screen = DefaultScreen(dpy);
+const char* extString = glXQueryExtensionsString(dpy, screen);
+if (!SGSearchExtensionsString(extString, GLX_SGIX_fbconfig) ||
+!SGSearchExtensionsString(extString, GLX_SGIX_pbuffer))
+return false;
+
glXChooseFBConfigPtr = (glXChooseFBConfigProc)SGLookupFunction(glXChooseFBConfigSGIX);
glXCreateGLXPbufferPtr = (glXCreateGLXPbufferProc)SGLookupFunction(glXCreateGLXPbufferSGIX);
glXGetVisualFromFBConfigPtr = (glXGetVisualFromFBConfigProc)SGLookupFunction(glXGetVisualFromFBConfigSGIX);
___
Flightgear-devel mailing list
Flightgear-devel@flightgear.org
http://mail.flightgear.org/mailman/listinfo/flightgear-devel
2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On December 12, 2005 01:19 pm, Mathias Fröhlich wrote: On Donnerstag 08 Dezember 2005 02:08, Ampere K. Hardraade wrote: https://bugs.freedesktop.org/show_bug.cgi?id=5142 So, the problem seems to be that pbuffer is not supported? Can you test the attached patch? Greetings Mathias It didn't work. X still throws the GLXUnsupportPrivateRequest at line 496 in RenderTexture.cpp. Apparently, the return false; statement at line 1849 never got run. I will do some more investigation later on. GNU gdb 6.3.90_20051119-debian Copyright 2004 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type show copying to see the conditions. There is absolutely no warranty for GDB. Type show warranty for details. This GDB was configured as i486-linux-gnu...Using host libthread_db library /lib/tls/libthread_db.so.1. (gdb) break RenderTexture.cpp:496 Breakpoint 1 at 0x83ba20c: file RenderTexture.cpp, line 496. (gdb) break RenderTexture.cpp:521 Breakpoint 2 at 0x83ba2fb: file RenderTexture.cpp, line 521. (gdb) break RenderTexture.cpp:1849 Breakpoint 3 at 0x83b9654: file RenderTexture.cpp, line 1849. (gdb) run Starting program: /usr/local/FlightGear/bin/fgfs [Thread debugging using libthread_db enabled] [New Thread -1221092768 (LWP 10061)] Program received signal SIGFPE, Arithmetic exception. [Switching to Thread -1221092768 (LWP 10061)] 0xafd330f8 in ?? () from /usr/X11R6/lib/modules/dri//r200_dri.so (gdb) cont Continuing. Disabling HW TCL support opening file: /usr/local/FlightGear/share/FlightGear/Navaids/carrier_nav.dat /usr/local/FlightGear/share/FlightGear/Navaids/TACAN_freq.dat [New Thread -1492231248 (LWP 10066)] Breakpoint 1, RenderTexture::Initialize (this=0xae429d0, width=256, height=256, shareObjects=true, copyContext=false) at RenderTexture.cpp:496 496True); (gdb) next X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 129 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 30 Current serial number in output stream: 31 Program exited with code 01. (gdb) quit Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On December 12, 2005 07:09 pm, Ampere K. Hardraade wrote: On December 12, 2005 01:19 pm, Mathias Fröhlich wrote: On Donnerstag 08 Dezember 2005 02:08, Ampere K. Hardraade wrote: https://bugs.freedesktop.org/show_bug.cgi?id=5142 So, the problem seems to be that pbuffer is not supported? Can you test the attached patch? Greetings Mathias It didn't work. X still throws the GLXUnsupportPrivateRequest at line 496 in RenderTexture.cpp. Apparently, the return false; statement at line 1849 never got run. I will do some more investigation later on. After some investigation, I found that the new if...else statement never got run because RenderTexture was able to successfully try the glx version 1.3 functions. I moved the new if...else statement so that the it gets run before RenderTexture can try the glx version 1.3 functions. FlightGear now starts up fine. However, there is some strange coloring issue: http://www.cs.yorku.ca/~cs233144/fgfs-screen-001.jpg Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On Dienstag 13 Dezember 2005 03:31, Ampere K. Hardraade wrote:
After some investigation, I found that the new if...else statement never
got run because RenderTexture was able to successfully try the glx version
1.3 functions.
Well I swapped the minor and major aguments ...
That one, appart from the color issue?
Greetings
Mathias
--
Mathias Fröhlich, email: [EMAIL PROTECTED]
Index: simgear/screen/RenderTexture.cpp
===
RCS file: /var/cvs/SimGear-0.3/SimGear/simgear/screen/RenderTexture.cpp,v
retrieving revision 1.15
diff -u -r1.15 RenderTexture.cpp
--- simgear/screen/RenderTexture.cpp 27 Nov 2005 09:48:51 - 1.15
+++ simgear/screen/RenderTexture.cpp 13 Dec 2005 06:24:54 -
@@ -1820,6 +1820,10 @@
}
#elif defined( __APPLE__ )
#else
+Display* dpy = glXGetCurrentDisplay();
+int minor = 0, major = 0;
+if (!glXQueryVersion(dpy, major, minor))
+return false;
// First try the glX version 1.3 functions.
glXChooseFBConfigPtr = (glXChooseFBConfigProc)SGLookupFunction(glXChooseFBConfig);
@@ -1829,7 +1833,8 @@
glXDestroyPbufferPtr = (glXDestroyPbufferProc)SGLookupFunction(glXDestroyPbuffer);
glXQueryDrawablePtr = (glXQueryDrawableProc)SGLookupFunction(glXQueryDrawable);
-if (glXChooseFBConfigPtr
+if (((1 = major 3 = minor) || 2 = major)
+glXChooseFBConfigPtr
glXCreatePbufferPtr
glXGetVisualFromFBConfigPtr
glXCreateContextPtr
@@ -1838,6 +1843,12 @@
glXVersion1_3Present = true;
else
{
+int screen = DefaultScreen(dpy);
+const char* extString = glXQueryExtensionsString(dpy, screen);
+if (!SGSearchExtensionsString(extString, GLX_SGIX_fbconfig) ||
+!SGSearchExtensionsString(extString, GLX_SGIX_pbuffer))
+return false;
+
glXChooseFBConfigPtr = (glXChooseFBConfigProc)SGLookupFunction(glXChooseFBConfigSGIX);
glXCreateGLXPbufferPtr = (glXCreateGLXPbufferProc)SGLookupFunction(glXCreateGLXPbufferSGIX);
glXGetVisualFromFBConfigPtr = (glXGetVisualFromFBConfigProc)SGLookupFunction(glXGetVisualFromFBConfigSGIX);
___
Flightgear-devel mailing list
Flightgear-devel@flightgear.org
http://mail.flightgear.org/mailman/listinfo/flightgear-devel
2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On Donnerstag 08 Dezember 2005 10:05, Erik Hofman wrote: Yes, and it also looks like the detection code is missing in RenderTexture.cpp ??! It looks like that ... I got the glxinfo's output from Ampere to check for exactly that but that slipped through my eyes. Erik, are you already working on that checks? Greetings Mathias -- Mathias Fröhlich, email: [EMAIL PROTECTED] ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Ampere K. Hardraade wrote: https://bugs.freedesktop.org/show_bug.cgi?id=5142 So, the problem seems to be that pbuffer is not supported? Yes, and it also looks like the detection code is missing in RenderTexture.cpp ??! Erik ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
https://bugs.freedesktop.org/show_bug.cgi?id=5142 So, the problem seems to be that pbuffer is not supported? Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Ampere K. Hardraade wrote: On November 30, 2005 08:22 pm, Curtis L. Olson wrote: Perhaps explain to them what our code is attempting to do, and then ask if they know of a GLX supported way to do it. I would do that if I can. However, I am not a programmer, and nothing in RenderTexture.cpp makes any sense to me. :( What exactly is our code attempting to do? Ampere The code creates a pbuffer the standard way, there is nothing special in rendertexture, the only things special is the choice of the api, using glx pre or post 1.3 - and this code won't work well for long since mesa lies on version numbers. The only standard test of pbuffers that I know is fgl_glxgears. But perhaps there is a test case somewhere in the mesa/xorg cvs, after all how do they test their code ? Harald. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On December 2, 2005 05:50 am, Harald JOHNSEN wrote: The code creates a pbuffer the standard way, there is nothing special in rendertexture, the only things special is the choice of the api, using glx pre or post 1.3 - and this code won't work well for long since mesa lies on version numbers. The only standard test of pbuffers that I know is fgl_glxgears. But perhaps there is a test case somewhere in the mesa/xorg cvs, after all how do they test their code ? Harald. Whether glXVersion1_3Present is true or false doesn't seem to make a difference. Last time I tried this, when glXVersion1_3Present is true, I get the GLXUnsupportPrivateRequest error at line 496. When glXVersion1_3Present is false, I get the same error at line 521. Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Mathias Fröhlich wrote : On Sonntag 27 November 2005 13:14, Frederic Bouvier wrote: Why not installing an X11 error handler to trap this instead of letting the default handler simply exiting ? Well, ist this possible? I was very excited about that idea, but found in the documentation that the error handler needs to call exit otherwise the calling function of the error handler will call exit past return of the error handler ... And what about a command line option, or a preferences.xml entry, that would disable the whole feature when the end user knows that there is a problem ? -Fred ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On November 30, 2005 08:22 pm, Curtis L. Olson wrote: Perhaps explain to them what our code is attempting to do, and then ask if they know of a GLX supported way to do it. I would do that if I can. However, I am not a programmer, and nothing in RenderTexture.cpp makes any sense to me. :( What exactly is our code attempting to do? Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Selon Ampere K. Hardraade:
On November 30, 2005 08:22 pm, Curtis L. Olson wrote:
Perhaps explain to them what our code is attempting to do, and then ask
if they know of a GLX supported way to do it.
I would do that if I can. However, I am not a programmer, and nothing in
RenderTexture.cpp makes any sense to me. :(
What exactly is our code attempting to do?
I won't look at the RenderTexture code. I'd rather do something like :
cvs -z4 -q diff -u bbcache.cxx (in directory
I:\FlightGear\cvs\SimGear\simgear\scene\sky\)
Index: bbcache.cxx
===
RCS file: /var/cvs/SimGear-0.3/SimGear/simgear/scene/sky/bbcache.cxx,v
retrieving revision 1.7
diff -u -r1.7 bbcache.cxx
--- bbcache.cxx 5 Sep 2005 08:17:37 - 1.7
+++ bbcache.cxx 2 Dec 2005 06:58:29 -
@@ -122,7 +122,7 @@
// rt-Reset(rgba tex2D ctt);
// rt-Reset(rgba tex2D);
- if( rt-Initialize(256, 256, true) ) {
+ if( !problemOnRenderTexture rt-Initialize(256, 256, true) ) {
SG_LOG(SG_ALL, SG_INFO, bbcache:Initialize sucessfull);
if (rt-BeginCapture())
{
And at other location where RenderTexture is used.
-Fred
___
Flightgear-devel mailing list
Flightgear-devel@flightgear.org
http://mail.flightgear.org/mailman/listinfo/flightgear-devel
2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On November 26, 2005 10:50 pm, Ampere K. Hardraade wrote: I finally managed to compile Xorg from source today and managed to get more information from gdb. I have also filed a bug report with Xorg: https://bugs.freedesktop.org/show_bug.cgi?id=5142 Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d Ah ha! A reply! https://bugs.freedesktop.org/show_bug.cgi?id=5142 The reply is as follow: X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 142 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 32 Current serial number in output stream: 33 Any clue what FG was trying to do when this was generated? All of the various FG mailing list threads referenced in this bug talk about RenderTexture. That makes me suspicious that FG is trying to do something with pbuffers or some other unsupported GLX functionality. Mesa warning: GL User Error: called without context: DeleteTextures That seems suspicious. What should I write in response? Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Ampere K. Hardraade wrote: On November 26, 2005 10:50 pm, Ampere K. Hardraade wrote: I finally managed to compile Xorg from source today and managed to get more information from gdb. I have also filed a bug report with Xorg: https://bugs.freedesktop.org/show_bug.cgi?id=5142 Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d Ah ha! A reply! https://bugs.freedesktop.org/show_bug.cgi?id=5142 What should I write in response? Perhaps explain to them what our code is attempting to do, and then ask if they know of a GLX supported way to do it. It sounds like the reply is suggesting that they have no support for pbuffers at all. If this is indeed the case, then a GLXUnsupportedPrivateRequest error might indeed be exactly what's going on. We are trying to do something RenderTexture that is not supported by GLX. The question is, can we do what we need with some approach that is supported by GLX? Curt. -- Curtis Olsonhttp://www.flightgear.org/~curt HumanFIRST Program http://www.humanfirst.umn.edu/ FlightGear Project http://www.flightgear.org Unique text:2f585eeea02e2c79d7b1d8c4963bae2d ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On Thu, 2005-11-24 at 10:06 +0100, Stefan Seifert wrote: I just wanted to note, that when it is clear, that it's a bug in ATI's drivers, someone should post a bugreport in the ATI driver Bugzilla: http://ati.cchtml.com/ This is actually a place where driver developers are watching and a good chance that such bugs get fixed. Before posting, you should read the instructions at: http://www.rage3d.com/board/showthread.php?t=33799828 which is btw. a thread in _the_ most useful forum for Linux using ATI card owners: http://www.rage3d.com/board/forumdisplay.php?f=88 I don't think it's ATI's binary driver fault. I have Ubuntu 5.10 with open-source radeon driver for ATI Radeon 9250 SE. After compiling 0.9.9 and SimGear 0.3.9 the following messages are printed after fgfs startup: [EMAIL PROTECTED] iso]$ /opt/fgfs/bin/fgfs opening file: /opt/fgfs/share/FlightGear/Navaids/carrier_nav.dat /opt/fgfs/share/FlightGear/Navaids/TACAN_freq.dat X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 143 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 40 Current serial number in output stream: 41 Timo ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
[EMAIL PROTECTED] iso]$ /opt/fgfs/bin/fgfs opening file: /opt/fgfs/share/FlightGear/Navaids/carrier_nav.dat /opt/fgfs/share/FlightGear/Navaids/TACAN_freq.dat X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 143 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 40 Current serial number in output stream: 41 Why not installing an X11 error handler to trap this instead of letting the default handler simply exiting ? -Fred ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On Sonntag 27 November 2005 13:14, Frederic Bouvier wrote: Why not installing an X11 error handler to trap this instead of letting the default handler simply exiting ? Well, ist this possible? I was very excited about that idea, but found in the documentation that the error handler needs to call exit otherwise the calling function of the error handler will call exit past return of the error handler ... Greetings Mathias -- Mathias Fröhlich, email: [EMAIL PROTECTED] ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Selon Mathias Fröhlich [EMAIL PROTECTED]: On Sonntag 27 November 2005 13:14, Frederic Bouvier wrote: Why not installing an X11 error handler to trap this instead of letting the default handler simply exiting ? Well, ist this possible? I was very excited about that idea, but found in the documentation that the error handler needs to call exit otherwise the calling function of the error handler will call exit past return of the error handler ... This is true only if the error is a fatal I/O that would trigger the handler installed by XSetIOErrorHandler ( http://tronche.com/gui/x/xlib/event-handling/protocol-errors/XSetIOErrorHandler.html ). This is not if the error triggers the handler installed by XSetErrorHandler ( http://tronche.com/gui/x/xlib/event-handling/protocol-errors/XSetErrorHandler.html ). -Fred ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On Sun, 2005-11-27 at 13:14 +0100, Frederic Bouvier wrote: [EMAIL PROTECTED] iso]$ /opt/fgfs/bin/fgfs opening file: /opt/fgfs/share/FlightGear/Navaids/carrier_nav.dat /opt/fgfs/share/FlightGear/Navaids/TACAN_freq.dat X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 143 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 40 Current serial number in output stream: 41 Why not installing an X11 error handler to trap this instead of letting the default handler simply exiting ? How do I install an X11 error handler? This is completely new concept for me. Quick googling didn't help much. Is there a command line option or should I modify FlightGear source or even X11 source? Timo ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Selon Timo Saarinen [EMAIL PROTECTED]: On Sun, 2005-11-27 at 13:14 +0100, Frederic Bouvier wrote: [EMAIL PROTECTED] iso]$ /opt/fgfs/bin/fgfs opening file: /opt/fgfs/share/FlightGear/Navaids/carrier_nav.dat /opt/fgfs/share/FlightGear/Navaids/TACAN_freq.dat X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 143 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 40 Current serial number in output stream: 41 Why not installing an X11 error handler to trap this instead of letting the default handler simply exiting ? How do I install an X11 error handler? This is completely new concept for me. Quick googling didn't help much. Is there a command line option or should I modify FlightGear source or even X11 source? A X11 error handler is a C/C++ function that is registered through the call of XSetErrorHandler or XSetIOErrorHandler functions. This is a developer task who needs to modify the existing SimGear code to do something that is not the default X11 behavior. -Fred ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
I finally managed to compile Xorg from source today and managed to get more information from gdb. I have also filed a bug report with Xorg: https://bugs.freedesktop.org/show_bug.cgi?id=5142 Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Mathias Fröhlich wrote: On Freitag 25 November 2005 04:28, Ampere K. Hardraade wrote: Ignore what I wrote in my last E-mail. I didn't realize I have errors when compiling. Whatever we end up here, I think that somebody with such hardware available should try to reproduce this with a small test program for two reasons: 1. Make sure that we do *not* have some kind of memory corruption within flightgear which is in the end responsible for that crash. 2. Have a simple testcase showing the problem. Just think if you would more /enjoy/ to work on a bug where you need to install plib/SimGear/Flightgear download the base package and then understand what is happening within that magitudes of code or if you would more likely take that problemreport with a simple 300 lines testcase where it is obvious to understand the problem... There is already a testcase besides the RenderTexture implementation. May be somebody can put together a testcase from that ... There is fgl_glxgears, it creates a pbuffer and render into it, simple code. Harald. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On November 25, 2005 01:35 am, Mathias Fröhlich wrote: I got the following errors when doing a make-install. What do I need to change? It looks like the RenderTexture stuff is not found at all. Did you get a compile error during RenderTexture compiles too? Greetings Mathias I think I have done something extremely stupid. :( I could use some instructions on how to use the TestRenderTexture.cpp file. Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Hi Mathias, I switched to the open source driver to see if I had better luck than with ATI's propietary one, and have encountered the same problem Ampere is having... I have tried to compile FlightGear using the TestRenderTexture.cpp in the cvs but am unable to compile it, here's a brief summary of the failed compilation: ../../src/Instrumentation/libInstrumentation.a(od_gauge.o): In function `~FGODGauge': /tmp/fg/source/src/Instrumentation/od_gauge.cxx:85: undefined reference to `RenderTexture::~RenderTexture()' ../../src/Instrumentation/libInstrumentation.a(od_gauge.o): In function `FGODGauge::endCapture(unsigned int)': /tmp/fg/source/src/Instrumentation/od_gauge.cxx:138: undefined reference to `RenderTexture::EndCapture()' ../../src/Instrumentation/libInstrumentation.a(od_gauge.o): In function `FGODGauge::allocRT()': /usr/local/SimGear-cvs/lib/libsgsky.a(bbcache.o): In function `SGBbCache::allocTextureMemory(int, int)': /tmp/SimGear/source/simgear/scene/sky/bbcache.cxx:84: undefined reference to `RenderTexture::BeginCapture()' /tmp/SimGear/source/simgear/scene/sky/bbcache.cxx:86: undefined reference to `RenderTexture::EndCapture()' I thought I'd ask you first instead of digging into that myself as I'm barely beginning to learn how the FG system works. thanks for trying to help us ATI owners to get over our misery :) cheers, -E Mathias Fröhlich wrote: On Donnerstag 24 November 2005 04:44, Ampere K. Hardraade wrote: X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 145 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 30 Current serial number in output stream: 31 Well, as far as I can see and remember: The client libraries send a request to the 'display system' and the 'display system' bails out with an 'unsupported request'. The error message is somehow misslieading, since the problem happens when communicating with dri/drm instead of the X server - the reason I called it 'display system' above. Since the same error happens regardless of using sgix pbuffers and glx-1.3 pbuffers and glxinfo announces sgix pbuffers (I got a glxinfo output in a private mail) it is a X/GLX/GL internal problem. Ampere maybe you can try to reproduce the problem using that TestRenderTexture.cpp. Greetings Mathias ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Mathias Fröhlich wrote: On Donnerstag 24 November 2005 04:44, Ampere K. Hardraade wrote: X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 145 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 30 Current serial number in output stream: 31 Well, as far as I can see and remember: The client libraries send a request to the 'display system' and the 'display system' bails out with an 'unsupported request'. The error message is somehow misslieading, since the problem happens when communicating with dri/drm instead of the X server - the reason I called it 'display system' above. I just wanted to note, that when it is clear, that it's a bug in ATI's drivers, someone should post a bugreport in the ATI driver Bugzilla: http://ati.cchtml.com/ This is actually a place where driver developers are watching and a good chance that such bugs get fixed. Before posting, you should read the instructions at: http://www.rage3d.com/board/showthread.php?t=33799828 which is btw. a thread in _the_ most useful forum for Linux using ATI card owners: http://www.rage3d.com/board/forumdisplay.php?f=88 Nine ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Ampere K. Hardraade wrote: First of all, thanks for looking into this. I have just tried your patch (I assume it was commited to the CVS). Unfortunately, it didn't work. I am still getting the GLXUnsupportPrivateRequest error at exactly the same spot. I'm afraid I can't help any further. I've read about any message I could find using google but no one seems to have a definitive answer. It seems to be an X-server related problem, so if if you feel brave you might want to install the latest version of X.org or XFree86. Erik ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Stefan Seifert wrote: Mathias Fröhlich wrote: On Donnerstag 24 November 2005 04:44, Ampere K. Hardraade wrote: X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 145 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 30 Current serial number in output stream: 31 Well, as far as I can see and remember: The client libraries send a request to the 'display system' and the 'display system' bails out with an 'unsupported request'. The error message is somehow misslieading, since the problem happens when communicating with dri/drm instead of the X server - the reason I called it 'display system' above. I just wanted to note, that when it is clear, that it's a bug in ATI's drivers, someone should post a bugreport in the ATI driver Bugzilla: http://ati.cchtml.com/ This is actually a place where driver developers are watching and a good chance that such bugs get fixed. Before posting, you should read the instructions at: http://www.rage3d.com/board/showthread.php?t=33799828 which is btw. a thread in _the_ most useful forum for Linux using ATI card owners: http://www.rage3d.com/board/forumdisplay.php?f=88 Nine ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d I'm not sure if it's the same one, but I have been following this one: http://ati.cchtml.com/show_bug.cgi?id=232 Doesn't seem to be any progress yet though. Josh ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On Thursday 24 November 2005 20:18, Josh Babcock wrote: Stefan Seifert wrote: Mathias Fröhlich wrote: On Donnerstag 24 November 2005 04:44, Ampere K. Hardraade wrote: X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 145 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 30 Current serial number in output stream: 31 Well, as far as I can see and remember: The client libraries send a request to the 'display system' and the 'display system' bails out with an 'unsupported request'. The error message is somehow misslieading, since the problem happens when communicating with dri/drm instead of the X server - the reason I called it 'display system' above. I just wanted to note, that when it is clear, that it's a bug in ATI's drivers, someone should post a bugreport in the ATI driver Bugzilla: http://ati.cchtml.com/ This is actually a place where driver developers are watching and a good chance that such bugs get fixed. Before posting, you should read the instructions at: http://www.rage3d.com/board/showthread.php?t=33799828 which is btw. a thread in _the_ most useful forum for Linux using ATI card owners: http://www.rage3d.com/board/forumdisplay.php?f=88 Nine ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d I'm not sure if it's the same one, but I have been following this one: http://ati.cchtml.com/show_bug.cgi?id=232 Doesn't seem to be any progress yet though. Josh Ampere says that they are not the same bug. Paul ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On Donnerstag 24 November 2005 10:06, Stefan Seifert wrote: I just wanted to note, that when it is clear, that it's a bug in ATI's drivers, someone should post a bugreport in the ATI driver Bugzilla: http://ati.cchtml.com/ No, it is not a problem of the ati binary driver, this is with the dri driver. Support from that guys is usually great. Much better than from the ati binary guys. But somebody needs to do ... The best chance to get support is to have a small testcase (don't call fg small! :) which could be submitted to bugs.freedesktop.org. Also the chance that this is already fixed with the next pending xorg release is high IMO. I dan no longer try, but if somebody can install the current release candidate and see if it is still there, we will know. Ok, I know about the upstream lag for X releases. This is a major problem, but it is also an offtopic for this list ... Greetings Mathias -- Mathias Fröhlich, email: [EMAIL PROTECTED] ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On November 24, 2005 01:50 am, Mathias Fröhlich wrote:
Well, as far as I can see and remember:
The client libraries send a request to the 'display system' and the
'display system' bails out with an 'unsupported request'.
The error message is somehow misslieading, since the problem happens when
communicating with dri/drm instead of the X server - the reason I called it
'display system' above.
Since the same error happens regardless of using sgix pbuffers and glx-1.3
pbuffers and glxinfo announces sgix pbuffers (I got a glxinfo output in a
private mail) it is a X/GLX/GL internal problem.
Ampere maybe you can try to reproduce the problem using that
TestRenderTexture.cpp.
Greetings
Mathias
Same problem:
Breakpoint 1, RenderTexture (this=0xad87d40, strMode=0x40 Address 0x40 out of
bounds) at RenderTexture.cpp:110
110 if (rt2-IsMipmapped())
(gdb) next
151 glTexParameteri(rt2-GetTextureTarget(),
(gdb)
154 }
(gdb)
166 if (rt2-IsTexture())
(gdb)
167 bShowDepthTexture = false;
(gdb)
168 }
(gdb)
169
(gdb)
172 }
(gdb)
178 void DestroyRenderTexture(RenderTexture *rt2)
(gdb)
180 delete rt2;
(gdb)
SGBbCache::init (this=0xad87d40, cacheCount=64) at bbcache.cxx:118
118 if( colorBits 8 )
(gdb)
121 rt-Reset(rgba ctt);
(gdb)
125 if( rt-Initialize(256, 256, true) ) {
(gdb)
X Error of failed request: GLXUnsupportedPrivateRequest
Major opcode of failed request: 145 (GLX)
Minor opcode of failed request: 16 (X_GLXVendorPrivate)
Serial number of failed request: 30
Current serial number in output stream: 31
Program exited with code 01.
(gdb)
The program is not being run.
(gdb)
The program is not being run.
(gdb)
The program is not being run.
(gdb) quit
Ampere
___
Flightgear-devel mailing list
Flightgear-devel@flightgear.org
http://mail.flightgear.org/mailman/listinfo/flightgear-devel
2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On November 24, 2005 03:01 pm, Mathias Fröhlich wrote: Also the chance that this is already fixed with the next pending xorg release is high IMO. I dan no longer try, but if somebody can install the current release candidate and see if it is still there, we will know. Ok, I know about the upstream lag for X releases. This is a major problem, but it is also an offtopic for this list ... Greetings Mathias I downloaded, compiled and installed the CVS version of Xorg some weeks ago. FlightGear crashed with the same message when launched. Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On November 24, 2005 01:50 am, Mathias Fröhlich wrote: Well, as far as I can see and remember: The client libraries send a request to the 'display system' and the 'display system' bails out with an 'unsupported request'. The error message is somehow misslieading, since the problem happens when communicating with dri/drm instead of the X server - the reason I called it 'display system' above. Since the same error happens regardless of using sgix pbuffers and glx-1.3 pbuffers and glxinfo announces sgix pbuffers (I got a glxinfo output in a private mail) it is a X/GLX/GL internal problem. Ampere maybe you can try to reproduce the problem using that TestRenderTexture.cpp. Greetings Mathias Ignore what I wrote in my last E-mail. I didn't realize I have errors when compiling. I got the following errors when doing a make-install. What do I need to change? ../../src/Instrumentation/libInstrumentation.a(od_gauge.o): In function `~FGODGauge':/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:85: undefined reference to `RenderTexture::~RenderTexture()' :/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:85: undefined reference to `RenderTexture::~RenderTexture()' :/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:85: undefined reference to `RenderTexture::~RenderTexture()' ../../src/Instrumentation/libInstrumentation.a(od_gauge.o): In function `FGODGauge::endCapture(unsigned int)':/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:138: undefined reference to `RenderTexture::EndCapture()' ../../src/Instrumentation/libInstrumentation.a(od_gauge.o): In function `FGODGauge::allocRT()':/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:45: undefined reference to `RenderTexture::RenderTexture(char const*)' :/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:47: undefined reference to `RenderTexture::Reset(char const*, ...)' :/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:51: undefined reference to `RenderTexture::Initialize(int, int, bool, bool)' :/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:53: undefined reference to `RenderTexture::BeginCapture()' :/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:77: undefined reference to `RenderTexture::EndCapture()' :/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:49: undefined reference to `RenderTexture::Reset(char const*, ...)' ../../src/Instrumentation/libInstrumentation.a(od_gauge.o): In function `FGODGauge::beginCapture(int)':/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:98: undefined reference to `RenderTexture::BeginCapture()' ../../src/Instrumentation/libInstrumentation.a(od_gauge.o): In function `FGODGauge::beginCapture()':/usr/local/src/flightgear/src/Instrumentation/od_gauge.cxx:110: undefined reference to `RenderTexture::BeginCapture()' /usr/local/lib/libsgsky.a(bbcache.o): In function `SGBbCache::allocTextureMemory(int, int)':/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:84: undefined reference to `RenderTexture::BeginCapture()' :/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:86: undefined reference to `RenderTexture::EndCapture()' /usr/local/lib/libsgsky.a(bbcache.o): In function `~SGBbCache':/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:106: undefined reference to `RenderTexture::~RenderTexture()' :/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:106: undefined reference to `RenderTexture::~RenderTexture()' /usr/local/lib/libsgsky.a(bbcache.o): In function `SGBbCache::init(int)':/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:115: undefined reference to `RenderTexture::RenderTexture(char const*)' :/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:119: undefined reference to `RenderTexture::Reset(char const*, ...)' :/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:121: undefined reference to `RenderTexture::Reset(char const*, ...)' :/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:125: undefined reference to `RenderTexture::Initialize(int, int, bool, bool)' :/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:127: undefined reference to `RenderTexture::BeginCapture()' :/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:150: undefined reference to `RenderTexture::EndCapture()' /usr/local/lib/libsgsky.a(bbcache.o): In function `SGBbCache::beginCapture()':/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:259: undefined reference to `RenderTexture::BeginCapture()' /usr/local/lib/libsgsky.a(bbcache.o): In function `SGBbCache::endCapture()':/usr/local/src/simgear/simgear/scene/sky/bbcache.cxx:304: undefined reference to `RenderTexture::EndCapture()' collect2: ld returned 1 exit status make[2]: *** [fgfs] Error 1 make[2]: Leaving directory `/usr/local/src/flightgear/src/Main' make[1]: *** [install-recursive] Error 1 make[1]: Leaving directory `/usr/local/src/flightgear/src' make: *** [install-recursive] Error 1 Ampere
Re: [Flightgear-devel] RenderTexture bug
On Freitag 25 November 2005 04:28, Ampere K. Hardraade wrote: Ignore what I wrote in my last E-mail. I didn't realize I have errors when compiling. Whatever we end up here, I think that somebody with such hardware available should try to reproduce this with a small test program for two reasons: 1. Make sure that we do *not* have some kind of memory corruption within flightgear which is in the end responsible for that crash. 2. Have a simple testcase showing the problem. Just think if you would more /enjoy/ to work on a bug where you need to install plib/SimGear/Flightgear download the base package and then understand what is happening within that magitudes of code or if you would more likely take that problemreport with a simple 300 lines testcase where it is obvious to understand the problem... There is already a testcase besides the RenderTexture implementation. May be somebody can put together a testcase from that ... I got the following errors when doing a make-install. What do I need to change? It looks like the RenderTexture stuff is not found at all. Did you get a compile error during RenderTexture compiles too? Greetings Mathias -- Mathias Fröhlich, email: [EMAIL PROTECTED] ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On November 22, 2005 01:06 pm, Erik Hofman wrote: Hi, I might have solved the nasty RenderTexture bug for ATI cards in CVS. Anyone cares to test it? Erik First of all, thanks for looking into this. I have just tried your patch (I assume it was commited to the CVS). Unfortunately, it didn't work. I am still getting the GLXUnsupportPrivateRequest error at exactly the same spot. [EMAIL PROTECTED]:~$ gdb /usr/local/FlightGear/bin/fgfs GNU gdb 6.3-debian Copyright 2004 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type show copying to see the conditions. There is absolutely no warranty for GDB. Type show warranty for details. This GDB was configured as i386-linux...Using host libthread_db library /lib/tls/libthread_db.so.1. (gdb) break RenderTexture.cpp:497 Breakpoint 1 at 0x838d864: file RenderTexture.cpp, line 497. (gdb) run Starting program: /usr/local/FlightGear/bin/fgfs [Thread debugging using libthread_db enabled] [New Thread -1220531616 (LWP 14145)] Program received signal SIGFPE, Arithmetic exception. [Switching to Thread -1220531616 (LWP 14145)] 0xafdb10a8 in ?? () from /usr/X11R6/lib/modules/dri//r200_dri.so (gdb) cont Continuing. Disabling HW TCL support opening file: /usr/local/FlightGear/share/FlightGear/Navaids/carrier_nav.dat /usr/local/FlightGear/share/FlightGear/Navaids/TACAN_freq.dat [New Thread -1491719248 (LWP 14148)] Breakpoint 1, RenderTexture::Initialize (this=0xae0b168, width=256, height=256, shareObjects=true, copyContext=false) at RenderTexture.cpp:497 497GL_TRUE); (gdb) next X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 145 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 30 Current serial number in output stream: 31 Program exited with code 01. (gdb) quit Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On Donnerstag 24 November 2005 04:44, Ampere K. Hardraade wrote: X Error of failed request: GLXUnsupportedPrivateRequest Major opcode of failed request: 145 (GLX) Minor opcode of failed request: 16 (X_GLXVendorPrivate) Serial number of failed request: 30 Current serial number in output stream: 31 Well, as far as I can see and remember: The client libraries send a request to the 'display system' and the 'display system' bails out with an 'unsupported request'. The error message is somehow misslieading, since the problem happens when communicating with dri/drm instead of the X server - the reason I called it 'display system' above. Since the same error happens regardless of using sgix pbuffers and glx-1.3 pbuffers and glxinfo announces sgix pbuffers (I got a glxinfo output in a private mail) it is a X/GLX/GL internal problem. Ampere maybe you can try to reproduce the problem using that TestRenderTexture.cpp. Greetings Mathias -- Mathias Fröhlich, email: [EMAIL PROTECTED] ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Quoting Erik Hofman : Hi, I might have solved the nasty RenderTexture bug for ATI cards in CVS. Anyone cares to test it? Erik, search the string '_vsnSG_LOG' in rendertexture.cpp -Fred ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Erik Hofman wrote: I might have solved the nasty RenderTexture bug for ATI cards in CVS. Anyone cares to test it? Yes, but I won't see 'my' PeeCee, the one at my customers location that's equipped with an r200 board, before thursday, Martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! -- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Frederic Bouvier wrote: search the string '_vsnSG_LOG' in rendertexture.cpp In case you missed the CVS log message; this is fixed now. Erik ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
On November 22, 2005 01:06 pm, Erik Hofman wrote: Hi, I might have solved the nasty RenderTexture bug for ATI cards in CVS. Anyone cares to test it? Erik I do. Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Hi Erik, I have tried tonight the cvs version and although I'm not suffering the RenderTexture problem (I use ATI's propietary binary driver - latest version 8.19.10), I have found the following error when starting up FG. RenderTexture Error: glXCreateGLXPbufferPtr() failed The program loads fine after that tho', however, I am still experiencing a segmentation fault when flying at night in the default San Francisco area. It only happens at night in this area, it won't happen when flying during the day in SF or when flying at night somewhere else (at least in the other scenery where I fly). When I take off (usually default runway 28R), I'd fly runway heading for some 15 to 30 seconds and then comes the segmentation fault. So far, running the program inside gdb gives little (at least to me), but I am pasting part of the output below, just hoping it can help anyone to track this issue: 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #0 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #1 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #2 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so . . . and goes on and on If you or anyone need any information or additional testing please let me know ... I will try to use the open source driver in the next few days to see how it goes for me. thanks, cheers -E Erik Hofman wrote: Hi, I might have solved the nasty RenderTexture bug for ATI cards in CVS. Anyone cares to test it? Erik ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Enrique Vaamonde wrote: Hi Erik, I have tried tonight the cvs version and although I'm not suffering the RenderTexture problem (I use ATI's propietary binary driver - latest version 8.19.10), I have found the following error when starting up FG. RenderTexture Error: glXCreateGLXPbufferPtr() failed The program loads fine after that tho', however, I am still experiencing a segmentation fault when flying at night in the default San Francisco area. It only happens at night in this area, it won't happen when flying during the day in SF or when flying at night somewhere else (at least in the other scenery where I fly). When I take off (usually default runway 28R), I'd fly runway heading for some 15 to 30 seconds and then comes the segmentation fault. So far, running the program inside gdb gives little (at least to me), but I am pasting part of the output below, just hoping it can help anyone to track this issue: 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #0 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #1 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #2 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so . . . and goes on and on If you or anyone need any information or additional testing please let me know ... I will try to use the open source driver in the next few days to see how it goes for me. The fact that the crash is inside the ati driver makes me just a little suspicious of a driver bug. What would be possibly useful is to see the lines further along in the backtrace so we can see where in the flightgear/simgear code the crash happen. Thanks, Curt. -- Curtis Olsonhttp://www.flightgear.org/~curt HumanFIRST Program http://www.humanfirst.umn.edu/ FlightGear Project http://www.flightgear.org Unique text:2f585eeea02e2c79d7b1d8c4963bae2d ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Hi Curt, as a matter of fact, the last line of the bt is this: #24939 0xafc88a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so (gdb) ...after that it's all empty...any clues? am I missing something? thank you :) -E Curtis L. Olson wrote: Enrique Vaamonde wrote: Hi Erik, I have tried tonight the cvs version and although I'm not suffering the RenderTexture problem (I use ATI's propietary binary driver - latest version 8.19.10), I have found the following error when starting up FG. RenderTexture Error: glXCreateGLXPbufferPtr() failed The program loads fine after that tho', however, I am still experiencing a segmentation fault when flying at night in the default San Francisco area. It only happens at night in this area, it won't happen when flying during the day in SF or when flying at night somewhere else (at least in the other scenery where I fly). When I take off (usually default runway 28R), I'd fly runway heading for some 15 to 30 seconds and then comes the segmentation fault. So far, running the program inside gdb gives little (at least to me), but I am pasting part of the output below, just hoping it can help anyone to track this issue: 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #0 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #1 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #2 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so . . . and goes on and on If you or anyone need any information or additional testing please let me know ... I will try to use the open source driver in the next few days to see how it goes for me. The fact that the crash is inside the ati driver makes me just a little suspicious of a driver bug. What would be possibly useful is to see the lines further along in the backtrace so we can see where in the flightgear/simgear code the crash happen. Thanks, Curt. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Enrique Vaamonde wrote: Hi Curt, as a matter of fact, the last line of the bt is this: #24939 0xafc88a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so (gdb) ...after that it's all empty...any clues? am I missing something? thank you :) There are situations where gdb barfs and says nothing useful. Not to pick on gdb ... I've never got the borland debugger to tell me anything useful after any crash, but that's a different story. Tracking down the specific spot of a crash can sometimes be easy (if gdb plays along) or can turn into an art form if the easy tricks fail ... Curt. -- Curtis Olsonhttp://www.flightgear.org/~curt HumanFIRST Program http://www.humanfirst.umn.edu/ FlightGear Project http://www.flightgear.org Unique text:2f585eeea02e2c79d7b1d8c4963bae2d ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Hi Curt, disregard the last message, I must have screwed up somewhere... anyway...here's the bt after the ATI driver errors: #87344 0x08418821 in ssgVtxTable::draw_geometry (this=0xecd8ba8) at ssgVtxTable.cxx:635 #87345 0x08417fe9 in ssgVtxTable::draw (this=0xecd8ba8) at ssgVtxTable.cxx:560 #87346 0x08405a1b in ssgSelector::cull (this=0xebd9bb0, f=0x8dd0020, m=0xbf8d62ec, test_needed=1) at ssgSelector.cxx:73 #87347 0x0840feea in ssgRangeSelector::cull (this=0xebd9a98, f=0x8dd0020, m=0xbf8d62ec, test_needed=1) at ssgRangeSelector.cxx:92 #87348 0x0840deb7 in ssgTransform::cull (this=0xebdc4d0, f=0x8dd0020, m=0x8dcffc8, test_needed=1) at ssgTransform.cxx:83 #87349 0x083f8230 in ssgBranch::cull (this=0xa719b18, f=0x8dd0020, m=0x8dcffc8, test_needed=1) at ssgBranch.cxx:276 #87350 0x083fa22b in ssgContext::cull (this=0x1bfc, r=0x0) at ssgContext.cxx:260 #87351 0x083f6102 in ssgCullAndDraw (r=0xa719b18) at ssg.cxx:303 #87352 0x08056ca8 in FGRenderer::update (refresh_camera_settings=true) at renderer.cxx:673 #87353 0x08054b7d in FGRenderer::update () at renderer.hxx:34 #87354 0x08081cc1 in GLUTdraw () at fg_os.cxx:118 #87355 0xb7f70884 in glutJoystickGetCenter () from /usr/lib/libglut.so.3 #87356 0xb7f74478 in fgEnumWindows () from /usr/lib/libglut.so.3 #87357 0xb7f70d6f in glutMainLoopEvent () from /usr/lib/libglut.so.3 #87358 0xb7f717fe in glutMainLoop () from /usr/lib/libglut.so.3 #87359 0x08051a82 in fgMainInit (argc=2, argv=0xbf8d6934) at main.cxx:1007 #87360 0x08051259 in main (argc=2, argv=0xbf8d6934) at bootstrap.cxx:193 I will try to understand this but like I mentioned before...I'm not really a graphics programmer. thanks for the help! -E Curtis L. Olson wrote: Enrique Vaamonde wrote: Hi Erik, I have tried tonight the cvs version and although I'm not suffering the RenderTexture problem (I use ATI's propietary binary driver - latest version 8.19.10), I have found the following error when starting up FG. RenderTexture Error: glXCreateGLXPbufferPtr() failed The program loads fine after that tho', however, I am still experiencing a segmentation fault when flying at night in the default San Francisco area. It only happens at night in this area, it won't happen when flying during the day in SF or when flying at night somewhere else (at least in the other scenery where I fly). When I take off (usually default runway 28R), I'd fly runway heading for some 15 to 30 seconds and then comes the segmentation fault. So far, running the program inside gdb gives little (at least to me), but I am pasting part of the output below, just hoping it can help anyone to track this issue: 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #0 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #1 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #2 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so . . . and goes on and on If you or anyone need any information or additional testing please let me know ... I will try to use the open source driver in the next few days to see how it goes for me. The fact that the crash is inside the ati driver makes me just a little suspicious of a driver bug. What would be possibly useful is to see the lines further along in the backtrace so we can see where in the flightgear/simgear code the crash happen. Thanks, Curt. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Hnmm, according to this you are crashing trying to draw the ambient ground light points. That should just be a collection of simple points with no strange opengl calls, unless you've turned on some of the fancy point lighting effects. If you have those on, you might want to turn them off. Otherwise, it looks like the driver is crashing just drawing simple untextured/colored points ... ( ? ) Curt. Enrique Vaamonde wrote: Hi Curt, disregard the last message, I must have screwed up somewhere... anyway...here's the bt after the ATI driver errors: #87344 0x08418821 in ssgVtxTable::draw_geometry (this=0xecd8ba8) at ssgVtxTable.cxx:635 #87345 0x08417fe9 in ssgVtxTable::draw (this=0xecd8ba8) at ssgVtxTable.cxx:560 #87346 0x08405a1b in ssgSelector::cull (this=0xebd9bb0, f=0x8dd0020, m=0xbf8d62ec, test_needed=1) at ssgSelector.cxx:73 #87347 0x0840feea in ssgRangeSelector::cull (this=0xebd9a98, f=0x8dd0020, m=0xbf8d62ec, test_needed=1) at ssgRangeSelector.cxx:92 #87348 0x0840deb7 in ssgTransform::cull (this=0xebdc4d0, f=0x8dd0020, m=0x8dcffc8, test_needed=1) at ssgTransform.cxx:83 #87349 0x083f8230 in ssgBranch::cull (this=0xa719b18, f=0x8dd0020, m=0x8dcffc8, test_needed=1) at ssgBranch.cxx:276 #87350 0x083fa22b in ssgContext::cull (this=0x1bfc, r=0x0) at ssgContext.cxx:260 #87351 0x083f6102 in ssgCullAndDraw (r=0xa719b18) at ssg.cxx:303 #87352 0x08056ca8 in FGRenderer::update (refresh_camera_settings=true) at renderer.cxx:673 #87353 0x08054b7d in FGRenderer::update () at renderer.hxx:34 #87354 0x08081cc1 in GLUTdraw () at fg_os.cxx:118 #87355 0xb7f70884 in glutJoystickGetCenter () from /usr/lib/libglut.so.3 #87356 0xb7f74478 in fgEnumWindows () from /usr/lib/libglut.so.3 #87357 0xb7f70d6f in glutMainLoopEvent () from /usr/lib/libglut.so.3 #87358 0xb7f717fe in glutMainLoop () from /usr/lib/libglut.so.3 #87359 0x08051a82 in fgMainInit (argc=2, argv=0xbf8d6934) at main.cxx:1007 #87360 0x08051259 in main (argc=2, argv=0xbf8d6934) at bootstrap.cxx:193 I will try to understand this but like I mentioned before...I'm not really a graphics programmer. thanks for the help! -E Curtis L. Olson wrote: Enrique Vaamonde wrote: Hi Erik, I have tried tonight the cvs version and although I'm not suffering the RenderTexture problem (I use ATI's propietary binary driver - latest version 8.19.10), I have found the following error when starting up FG. RenderTexture Error: glXCreateGLXPbufferPtr() failed The program loads fine after that tho', however, I am still experiencing a segmentation fault when flying at night in the default San Francisco area. It only happens at night in this area, it won't happen when flying during the day in SF or when flying at night somewhere else (at least in the other scenery where I fly). When I take off (usually default runway 28R), I'd fly runway heading for some 15 to 30 seconds and then comes the segmentation fault. So far, running the program inside gdb gives little (at least to me), but I am pasting part of the output below, just hoping it can help anyone to track this issue: 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #0 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #1 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #2 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so . . . and goes on and on If you or anyone need any information or additional testing please let me know ... I will try to use the open source driver in the next few days to see how it goes for me. The fact that the crash is inside the ati driver makes me just a little suspicious of a driver bug. What would be possibly useful is to see the lines further along in the backtrace so we can see where in the flightgear/simgear code the crash happen. Thanks, Curt. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d -- Curtis Olsonhttp://www.flightgear.org/~curt HumanFIRST Program http://www.humanfirst.umn.edu/ FlightGear Project http://www.flightgear.org Unique text:2f585eeea02e2c79d7b1d8c4963bae2d ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Acctually I shouldn't be using any advanced or fancy point lightning effects unless they come turned on in the default preferences in the cvs. Funny thing is: why is the driver crashing only in the bay area scenery? I don't seem to have any night flying problems with the other area I use (w070n10). Could it be an object? I will try removing 3D objects around the bay area scenery and will also try going back to freeglut 2.2 and let you know. thanks for your help, -E Curtis L. Olson wrote: Hnmm, according to this you are crashing trying to draw the ambient ground light points. That should just be a collection of simple points with no strange opengl calls, unless you've turned on some of the fancy point lighting effects. If you have those on, you might want to turn them off. Otherwise, it looks like the driver is crashing just drawing simple untextured/colored points ... ( ? ) Curt. Enrique Vaamonde wrote: Hi Curt, disregard the last message, I must have screwed up somewhere... anyway...here's the bt after the ATI driver errors: #87344 0x08418821 in ssgVtxTable::draw_geometry (this=0xecd8ba8) at ssgVtxTable.cxx:635 #87345 0x08417fe9 in ssgVtxTable::draw (this=0xecd8ba8) at ssgVtxTable.cxx:560 #87346 0x08405a1b in ssgSelector::cull (this=0xebd9bb0, f=0x8dd0020, m=0xbf8d62ec, test_needed=1) at ssgSelector.cxx:73 #87347 0x0840feea in ssgRangeSelector::cull (this=0xebd9a98, f=0x8dd0020, m=0xbf8d62ec, test_needed=1) at ssgRangeSelector.cxx:92 #87348 0x0840deb7 in ssgTransform::cull (this=0xebdc4d0, f=0x8dd0020, m=0x8dcffc8, test_needed=1) at ssgTransform.cxx:83 #87349 0x083f8230 in ssgBranch::cull (this=0xa719b18, f=0x8dd0020, m=0x8dcffc8, test_needed=1) at ssgBranch.cxx:276 #87350 0x083fa22b in ssgContext::cull (this=0x1bfc, r=0x0) at ssgContext.cxx:260 #87351 0x083f6102 in ssgCullAndDraw (r=0xa719b18) at ssg.cxx:303 #87352 0x08056ca8 in FGRenderer::update (refresh_camera_settings=true) at renderer.cxx:673 #87353 0x08054b7d in FGRenderer::update () at renderer.hxx:34 #87354 0x08081cc1 in GLUTdraw () at fg_os.cxx:118 #87355 0xb7f70884 in glutJoystickGetCenter () from /usr/lib/libglut.so.3 #87356 0xb7f74478 in fgEnumWindows () from /usr/lib/libglut.so.3 #87357 0xb7f70d6f in glutMainLoopEvent () from /usr/lib/libglut.so.3 #87358 0xb7f717fe in glutMainLoop () from /usr/lib/libglut.so.3 #87359 0x08051a82 in fgMainInit (argc=2, argv=0xbf8d6934) at main.cxx:1007 #87360 0x08051259 in main (argc=2, argv=0xbf8d6934) at bootstrap.cxx:193 I will try to understand this but like I mentioned before...I'm not really a graphics programmer. thanks for the help! -E Curtis L. Olson wrote: Enrique Vaamonde wrote: Hi Erik, I have tried tonight the cvs version and although I'm not suffering the RenderTexture problem (I use ATI's propietary binary driver - latest version 8.19.10), I have found the following error when starting up FG. RenderTexture Error: glXCreateGLXPbufferPtr() failed The program loads fine after that tho', however, I am still experiencing a segmentation fault when flying at night in the default San Francisco area. It only happens at night in this area, it won't happen when flying during the day in SF or when flying at night somewhere else (at least in the other scenery where I fly). When I take off (usually default runway 28R), I'd fly runway heading for some 15 to 30 seconds and then comes the segmentation fault. So far, running the program inside gdb gives little (at least to me), but I am pasting part of the output below, just hoping it can help anyone to track this issue: 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #0 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #1 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #2 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so . . . and goes on and on If you or anyone need any information or additional testing please let me know ... I will try to use the open source driver in the next few days to see how it goes for me. The fact that the crash is inside the ati driver makes me just a little suspicious of a driver bug. What would be possibly useful is to see the lines further along in the backtrace so we can see where in the flightgear/simgear code the crash happen. Thanks, Curt. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] RenderTexture bug
Well, neither disabling 3D objects nor going back to freeglut 2.2 (and recompiling simgear/fg with it, just in case) fixed the problem... if any of you has a suggestion, please do speak up :) By the way...additional problems: no 3d clouds at all and the sun is displayed as a black circle. I will try to use the open source drivers and see if anything improves, which I doubt. thanks! cheers, -E Enrique Vaamonde wrote: Acctually I shouldn't be using any advanced or fancy point lightning effects unless they come turned on in the default preferences in the cvs. Funny thing is: why is the driver crashing only in the bay area scenery? I don't seem to have any night flying problems with the other area I use (w070n10). Could it be an object? I will try removing 3D objects around the bay area scenery and will also try going back to freeglut 2.2 and let you know. thanks for your help, -E Curtis L. Olson wrote: Hnmm, according to this you are crashing trying to draw the ambient ground light points. That should just be a collection of simple points with no strange opengl calls, unless you've turned on some of the fancy point lighting effects. If you have those on, you might want to turn them off. Otherwise, it looks like the driver is crashing just drawing simple untextured/colored points ... ( ? ) Curt. Enrique Vaamonde wrote: Hi Curt, disregard the last message, I must have screwed up somewhere... anyway...here's the bt after the ATI driver errors: #87344 0x08418821 in ssgVtxTable::draw_geometry (this=0xecd8ba8) at ssgVtxTable.cxx:635 #87345 0x08417fe9 in ssgVtxTable::draw (this=0xecd8ba8) at ssgVtxTable.cxx:560 #87346 0x08405a1b in ssgSelector::cull (this=0xebd9bb0, f=0x8dd0020, m=0xbf8d62ec, test_needed=1) at ssgSelector.cxx:73 #87347 0x0840feea in ssgRangeSelector::cull (this=0xebd9a98, f=0x8dd0020, m=0xbf8d62ec, test_needed=1) at ssgRangeSelector.cxx:92 #87348 0x0840deb7 in ssgTransform::cull (this=0xebdc4d0, f=0x8dd0020, m=0x8dcffc8, test_needed=1) at ssgTransform.cxx:83 #87349 0x083f8230 in ssgBranch::cull (this=0xa719b18, f=0x8dd0020, m=0x8dcffc8, test_needed=1) at ssgBranch.cxx:276 #87350 0x083fa22b in ssgContext::cull (this=0x1bfc, r=0x0) at ssgContext.cxx:260 #87351 0x083f6102 in ssgCullAndDraw (r=0xa719b18) at ssg.cxx:303 #87352 0x08056ca8 in FGRenderer::update (refresh_camera_settings=true) at renderer.cxx:673 #87353 0x08054b7d in FGRenderer::update () at renderer.hxx:34 #87354 0x08081cc1 in GLUTdraw () at fg_os.cxx:118 #87355 0xb7f70884 in glutJoystickGetCenter () from /usr/lib/libglut.so.3 #87356 0xb7f74478 in fgEnumWindows () from /usr/lib/libglut.so.3 #87357 0xb7f70d6f in glutMainLoopEvent () from /usr/lib/libglut.so.3 #87358 0xb7f717fe in glutMainLoop () from /usr/lib/libglut.so.3 #87359 0x08051a82 in fgMainInit (argc=2, argv=0xbf8d6934) at main.cxx:1007 #87360 0x08051259 in main (argc=2, argv=0xbf8d6934) at bootstrap.cxx:193 I will try to understand this but like I mentioned before...I'm not really a graphics programmer. thanks for the help! -E Curtis L. Olson wrote: Enrique Vaamonde wrote: Hi Erik, I have tried tonight the cvs version and although I'm not suffering the RenderTexture problem (I use ATI's propietary binary driver - latest version 8.19.10), I have found the following error when starting up FG. RenderTexture Error: glXCreateGLXPbufferPtr() failed The program loads fine after that tho', however, I am still experiencing a segmentation fault when flying at night in the default San Francisco area. It only happens at night in this area, it won't happen when flying during the day in SF or when flying at night somewhere else (at least in the other scenery where I fly). When I take off (usually default runway 28R), I'd fly runway heading for some 15 to 30 seconds and then comes the segmentation fault. So far, running the program inside gdb gives little (at least to me), but I am pasting part of the output below, just hoping it can help anyone to track this issue: 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #0 0xafc361b3 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #1 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so #2 0xafc35a19 in __glim_R200TCLDrawArrays () from /usr/X11R6/lib/modules/dri/atiogl_a_dri.so . . . and goes on and on If you or anyone need any information or additional testing please let me know ... I will try to use the open source driver in the next few days to see how it goes for me. The fact that the crash is inside the ati driver makes me just a little suspicious of a driver bug. What would be possibly useful is to see the lines further along in the backtrace so we can see where in the flightgear/simgear code the crash happen. Thanks, Curt. ___ Flightgear-devel mailing list
Re: [Flightgear-devel] RenderTexture bug
Enrique Vaamonde wrote: I will try to use the open source drivers and see if anything improves, which I doubt. The OpenSource drivers should do - at least they did for me for a long time, Martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! -- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
