Segfault when using simple shapes

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

Segfault when using simple shapes

sims-9
Hi
everybody,
 I'm trying to simulate some simple models with simple shapes but get
segfaults in the visualisation part. The model contains a simple
plane, cylinder and directional light. I've attached the file
"stap1.world" that I'm using. When I'm starting gazebo with
stap1.world, Gazebo crashes with a segfault after successful
initialization. I tried this with several basic shapes (box, cylinder,
cone, tube) but get the same behaviour. I've also tried it on
different machines with the same behaviour.
 The part in the world file which is causing the trouble is:
           unit_cylinder
           Gazebo/RustyBarrel
 If I remove this part from the world file, the gazebo does not
crash. But of course I don't see the cylinder.
 Bug 2870446 might be related. But the patched file is no longer part
of the trunk.
 I'm not sure if this is a bug in Gazebo, Ogre or in one of drivers.
Or if the world model is plainly wrong. Hopefully somebody can help me
with this one as it's really blocking for me and I'm pulling my hair
out.
 Here's some more info:
 - Gazebo trunk rev 8622
 - Ogre 1.6.4
 - gcc 4.4.3
 For more information I've included below (1) world file; (2)
terminal output Gazebo; (3) output of gdb showing where the segfault
occurs.
 Thanks,
 Barry
  This is the output I get from the terminal:
 $gazebo stap1.world
Gazebo multi-robot simulator, version 0.10.0

Part of the Player/Stage Project
[http://playerstage.sourceforge.net]. [1]
Copyright (C) 2003 Nate Koenig, Andrew Howard, and contributors.
Released under the GNU General Public License.

[/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:103]
  Gazebo Path[/usr/local/share/gazebo]
[/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:115]
  Ogre Path[/usr/lib/OGRE]
directory [/tmp/gazebo-barry-0] already xists (previous crash?)
but the owner gazebo server (pid=4430) is not running.
deleting the old information of the directory [/tmp/gazebo-barry-0]
Gazebo successfully initialized
Segmentation fault (core dumped)
 

Links:
------
[1] http://playerstage.sourceforge.net].

------------------------------------------------------------------------------

_______________________________________________
Playerstage-gazebo mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo

gdb.txt (12K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Segfault when using simple shapes

John Hsu-2
Hi Barry,
To help diagnose the problem, can you attach stap1.world?
also, what graphics card and driver are you using?
thanks,
John
On Sun, Apr 25, 2010 at 12:21 AM, <[hidden email]> wrote:

> Hi
> everybody,
>  I'm trying to simulate some simple models with simple shapes but get
> segfaults in the visualisation part. The model contains a simple
> plane, cylinder and directional light. I've attached the file
> "stap1.world" that I'm using. When I'm starting gazebo with
> stap1.world, Gazebo crashes with a segfault after successful
> initialization. I tried this with several basic shapes (box, cylinder,
> cone, tube) but get the same behaviour. I've also tried it on
> different machines with the same behaviour.
>  The part in the world file which is causing the trouble is:
>           unit_cylinder
>           Gazebo/RustyBarrel
>  If I remove this part from the world file, the gazebo does not
> crash. But of course I don't see the cylinder.
>  Bug 2870446 might be related. But the patched file is no longer part
> of the trunk.
>  I'm not sure if this is a bug in Gazebo, Ogre or in one of drivers.
> Or if the world model is plainly wrong. Hopefully somebody can help me
> with this one as it's really blocking for me and I'm pulling my hair
> out.
>  Here's some more info:
>  - Gazebo trunk rev 8622
>  - Ogre 1.6.4
>  - gcc 4.4.3
>  For more information I've included below (1) world file; (2)
> terminal output Gazebo; (3) output of gdb showing where the segfault
> occurs.
>  Thanks,
>  Barry
>  This is the output I get from the terminal:
>  $gazebo stap1.world
> Gazebo multi-robot simulator, version 0.10.0
>
> Part of the Player/Stage Project
> [http://playerstage.sourceforge.net]. [1]
> Copyright (C) 2003 Nate Koenig, Andrew Howard, and contributors.
> Released under the GNU General Public License.
>
> [/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:103]
>  Gazebo Path[/usr/local/share/gazebo]
> [/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:115]
>  Ogre Path[/usr/lib/OGRE]
> directory [/tmp/gazebo-barry-0] already xists (previous crash?)
> but the owner gazebo server (pid=4430) is not running.
> deleting the old information of the directory [/tmp/gazebo-barry-0]
> Gazebo successfully initialized
> Segmentation fault (core dumped)
>
>
> Links:
> ------
> [1] http://playerstage.sourceforge.net].
>
>
> ------------------------------------------------------------------------------
>
> _______________________________________________
> Playerstage-gazebo mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
>
>
------------------------------------------------------------------------------
_______________________________________________
Playerstage-gazebo mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
Reply | Threaded
Open this post in threaded view
|

Re: Segfault when using simple shapes

sims-9
In reply to this post by sims-9

 Hi John,
 I've attached stap1.world (apparently something went wrong with
attaching it to the first mail).
 I'm running Gazebo on two setups.
 (1) The first is a VM Player setup (don't laugh).
 From /var/log/xorg.log:
 (--) PCI:*(0:0:15:0) 15ad:0405:15ad:0405 VMware Inc Abstract SVGA II
Adapter rev 0, Mem @ 0xd0000000/1
 34217728, 0xd8000000/8388608, I/O @ 0x000010d0/16, BIOS @
0x????????/32768
 (==) Using default built-in configuration (30 lines)
 (==) --- Start of built-in configuration ---
         Section "Device"
                 Identifier      "Builtin Default vmware Device 0"
                 Driver  "vmware"
         EndSection
         Section "Screen"
                 Identifier      "Builtin Default vmware Screen 0"
                 Device  "Builtin Default vmware Device 0"
         EndSection
 dmesg tells me that the segfault occured in "libdricore.so".
 (2) The second setup is also a very fast one:
 (--) PCI:*(0:0:2:0) 8086:2772:8086:464c Intel Corporation 82945G/GZ
Integrated Graphics Controller rev 2, Mem @ 0x90200000/524288,
0x80000000/268435456, 0x90280000/262144, I/O @ 0x000020e0/8, BIOS @
0x????????/131072
 (==) Using default built-in configuration (30 lines)
 (==) --- Start of built-in configuration ---
         Section "Device"
                 Identifier      "Builtin Default intel Device 0"
                 Driver  "intel"
         EndSection
         Section "Screen"
                 Identifier      "Builtin Default intel Screen 0"
                 Device  "Builtin Default intel Device 0"
         EndSection
         Section "Device"
                 Identifier      "Builtin Default vesa Device 0"
                 Driver  "vesa"
         EndSection
         Section "Screen"
                 Identifier      "Builtin Default vesa Screen 0"
                 Device  "Builtin Default vesa Device 0"
         EndSection
         Section "Device"
                 Identifier      "Builtin Default fbdev Device 0"
 I was accessing this second machine using vncserver. Then I had the
same problem. Also here dmesg tells me that "libdricore.so" causes the
problem.
 Inspired by your question, I run Gazebo while I was logged on
directly (from the console) this afternoon. Than it actually ran
successful! Apparently some problem with the (lack off) hardware
acceleration?
 Thanks,
 Barry
 On Sun 25/04/10 10:03 , John Hsu [hidden email] sent:
 Hi Barry,
 To help diagnose the problem, can you attach stap1.world?
 also, what graphics card and driver are you using?
 thanks,
 John
 On Sun, Apr 25, 2010 at 12:21 AM,   wrote:
 Hi
 everybody,
  I'm trying to simulate some simple models with simple shapes but
get
 segfaults in the visualisation part. The model contains a simple
 plane, cylinder and directional light. I've attached the file
 "stap1.world" that I'm using. When I'm starting gazebo with
 stap1.world, Gazebo crashes with a segfault after successful
 initialization. I tried this with several basic shapes (box,
cylinder,
 cone, tube) but get the same behaviour. I've also tried it on
 different machines with the same behaviour.
  The part in the world file which is causing the trouble is:
            unit_cylinder
            Gazebo/RustyBarrel
  If I remove this part from the world file, the gazebo does not
 crash. But of course I don't see the cylinder.
  Bug 2870446 might be related. But the patched file is no longer
part
 of the trunk.
  I'm not sure if this is a bug in Gazebo, Ogre or in one of drivers.
 Or if the world model is plainly wrong. Hopefully somebody can help
me
 with this one as it's really blocking for me and I'm pulling my hair
 out.
  Here's some more info:
  - Gazebo trunk rev 8622
  - Ogre 1.6.4
  - gcc 4.4.3
  For more information I've included below (1) world file; (2)
 terminal output Gazebo; (3) output of gdb showing where the segfault
 occurs.
  Thanks,
  Barry
   This is the output I get from the terminal:
  $gazebo stap1.world
 Gazebo multi-robot simulator, version 0.10.0
 Part of the Player/Stage Project
 [http://playerstage.sourceforge.net [2]]. [1]
 Copyright (C) 2003 Nate Koenig, Andrew Howard, and contributors.
 Released under the GNU General Public License.

[/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:103]
   Gazebo Path[/usr/local/share/gazebo]

[/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:115]
   Ogre Path[/usr/lib/OGRE]
 directory [/tmp/gazebo-barry-0] already xists (previous crash?)
 but the owner gazebo server (pid=4430) is not running.
 deleting the old information of the directory [/tmp/gazebo-barry-0]
 Gazebo successfully initialized
 Segmentation fault (core dumped)
 Links:
 ------
 [1] http://playerstage.sourceforge.net [3]].

------------------------------------------------------------------------------
 _______________________________________________
 Playerstage-gazebo mailing list
 https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo [5]
 

Links:
------
[2] http://playerstage.sourceforge.net
[3] http://playerstage.sourceforge.net
[5] https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
------------------------------------------------------------------------------
_______________________________________________
Playerstage-gazebo mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
Reply | Threaded
Open this post in threaded view
|

Re: Segfault when using simple shapes

sims-9
In reply to this post by sims-9
Somehow the attachment of stap1.world is not coming through. Here's
the contents of the world model:
   5
     0.03
     0 0 -9.8
     10e-5
     0.8
     fltk
     800 600
     0 0
     0.4 0.4 0.4 1.0
       Gazebo/CloudySky
     0 0 0
     0 0 0
     true
         0 0 1
         2000 2000
         10 10
         100 100
         Gazebo/GrassFloor
     5 0 0.5
     0.0 0.0 0.0
     false
         0.5 1
         1.0
           unit_cylinder
           Gazebo/RustyBarrel
     true
       directional
       0 -0.8 -0.3
       0.9 0.9 0.9
       0.0 0.0 0.0
       100
       0.0 1.0 0.4
 On Sun 25/04/10 14:13 , [hidden email] sent:
  Hi John,
  I've attached stap1.world (apparently something went wrong with
 attaching it to the first mail).
  I'm running Gazebo on two setups.
  (1) The first is a VM Player setup (don't laugh).
  From /var/log/xorg.log:
  (--) PCI:*(0:0:15:0) 15ad:0405:15ad:0405 VMware Inc Abstract SVGA
II
 Adapter rev 0, Mem @ 0xd0000000/1
  34217728, 0xd8000000/8388608, I/O @ 0x000010d0/16, BIOS @
 0x????????/32768
  (==) Using default built-in configuration (30 lines)
  (==) --- Start of built-in configuration ---
          Section "Device"
                  Identifier      "Builtin Default vmware Device 0"
                  Driver  "vmware"
          EndSection
          Section "Screen"
                  Identifier      "Builtin Default vmware Screen 0"
                  Device  "Builtin Default vmware Device 0"
          EndSection
  dmesg tells me that the segfault occured in "libdricore.so".
  (2) The second setup is also a very fast one:
  (--) PCI:*(0:0:2:0) 8086:2772:8086:464c Intel Corporation 82945G/GZ
 Integrated Graphics Controller rev 2, Mem @ 0x90200000/524288,
 0x80000000/268435456, 0x90280000/262144, I/O @ 0x000020e0/8, BIOS @
 0x????????/131072
  (==) Using default built-in configuration (30 lines)
  (==) --- Start of built-in configuration ---
          Section "Device"
                  Identifier      "Builtin Default intel Device 0"
                  Driver  "intel"
          EndSection
          Section "Screen"
                  Identifier      "Builtin Default intel Screen 0"
                  Device  "Builtin Default intel Device 0"
          EndSection
          Section "Device"
                  Identifier      "Builtin Default vesa Device 0"
                  Driver  "vesa"
          EndSection
          Section "Screen"
                  Identifier      "Builtin Default vesa Screen 0"
                  Device  "Builtin Default vesa Device 0"
          EndSection
          Section "Device"
                  Identifier      "Builtin Default fbdev Device 0"
  I was accessing this second machine using vncserver. Then I had the
 same problem. Also here dmesg tells me that "libdricore.so" causes
the
 problem.
  Inspired by your question, I run Gazebo while I was logged on
 directly (from the console) this afternoon. Than it actually ran
 successful! Apparently some problem with the (lack off) hardware
 acceleration?
  Thanks,
  Barry
  On Sun 25/04/10 10:03 , John Hsu [hidden email] [1] sent:
  Hi Barry,
  To help diagnose the problem, can you attach stap1.world?
  also, what graphics card and driver are you using?
  thanks,
  John
  On Sun, Apr 25, 2010 at 12:21 AM,   wrote:
  Hi
  everybody,
   I'm trying to simulate some simple models with simple shapes but
 get
  segfaults in the visualisation part. The model contains a simple
  plane, cylinder and directional light. I've attached the file
  "stap1.world" that I'm using. When I'm starting gazebo with
  stap1.world, Gazebo crashes with a segfault after successful
  initialization. I tried this with several basic shapes (box,
 cylinder,
  cone, tube) but get the same behaviour. I've also tried it on
  different machines with the same behaviour.
   The part in the world file which is causing the trouble is:
             unit_cylinder
             Gazebo/RustyBarrel
   If I remove this part from the world file, the gazebo does not
  crash. But of course I don't see the cylinder.
   Bug 2870446 might be related. But the patched file is no longer
 part
  of the trunk.
   I'm not sure if this is a bug in Gazebo, Ogre or in one of
drivers.
  Or if the world model is plainly wrong. Hopefully somebody can help
 me
  with this one as it's really blocking for me and I'm pulling my
hair
  out.
   Here's some more info:
   - Gazebo trunk rev 8622
   - Ogre 1.6.4
   - gcc 4.4.3
   For more information I've included below (1) world file; (2)
  terminal output Gazebo; (3) output of gdb showing where the
segfault
  occurs.
   Thanks,
   Barry
    This is the output I get from the terminal:
   $gazebo stap1.world
  Gazebo multi-robot simulator, version 0.10.0
  Part of the Player/Stage Project
  [http://playerstage.sourceforge.net [2]]. [1]
  Copyright (C) 2003 Nate Koenig, Andrew Howard, and contributors.
  Released under the GNU General Public License.

[/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:103]
    Gazebo Path[/usr/local/share/gazebo]

[/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:115]
    Ogre Path[/usr/lib/OGRE]
  directory [/tmp/gazebo-barry-0] already xists (previous crash?)
  but the owner gazebo server (pid=4430) is not running.
  deleting the old information of the directory [/tmp/gazebo-barry-0]
  Gazebo successfully initialized
  Segmentation fault (core dumped)
  Links:
  ------
  [1] http://playerstage.sourceforge.net [3]].

------------------------------------------------------------------------------
  _______________________________________________
  Playerstage-gazebo mailing list
  https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo [5]
 Links:
 ------
 [2] http://playerstage.sourceforge.net
 [3] http://playerstage.sourceforge.net
 [5] https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo

------------------------------------------------------------------------------
 _______________________________________________
 Playerstage-gazebo mailing list
 [hidden email] [2]
 https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
 

Links:
------
[1] mailto:[hidden email]
[2] mailto:[hidden email]
------------------------------------------------------------------------------
_______________________________________________
Playerstage-gazebo mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
Reply | Threaded
Open this post in threaded view
|

Re: Segfault when using simple shapes

John Hsu-2
Hi Bary,
Your posting of the world file is not coming through :( but from my
experience intel and older ati graphics cards and drivers don't have very
good support for opengl/glx.   I don't think people have had much success on
vm's either.  I tried to compile a list of compatible hardware
here<http://ros.org/wiki/simulator_gazebo/SystemRequirements>
.
Best,
John


On Sunday, April 25, 2010,  <[hidden email]> wrote:

> Somehow the attachment of stap1.world is not coming through. Here's
> the contents of the world model:
>    5
>      0.03
>      0 0 -9.8
>      10e-5
>      0.8
>      fltk
>      800 600
>      0 0
>      0.4 0.4 0.4 1.0
>        Gazebo/CloudySky
>      0 0 0
>      0 0 0
>      true
>          0 0 1
>          2000 2000
>          10 10
>          100 100
>          Gazebo/GrassFloor
>      5 0 0.5
>      0.0 0.0 0.0
>      false
>          0.5 1
>          1.0
>            unit_cylinder
>            Gazebo/RustyBarrel
>      true
>        directional
>        0 -0.8 -0.3
>        0.9 0.9 0.9
>        0.0 0.0 0.0
>        100
>        0.0 1.0 0.4
>  On Sun 25/04/10 14:13 , [hidden email] sent:
>   Hi John,
>   I've attached stap1.world (apparently something went wrong with
>  attaching it to the first mail).
>   I'm running Gazebo on two setups.
>   (1) The first is a VM Player setup (don't laugh).
>   From /var/log/xorg.log:
>   (--) PCI:*(0:0:15:0) 15ad:0405:15ad:0405 VMware Inc Abstract SVGA
> II
>  Adapter rev 0, Mem @ 0xd0000000/1
>   34217728, 0xd8000000/8388608, I/O @ 0x000010d0/16, BIOS @
>  0x????????/32768
>   (==) Using default built-in configuration (30 lines)
>   (==) --- Start of built-in configuration ---
>           Section "Device"
>                   Identifier      "Builtin Default vmware Device 0"
>                   Driver  "vmware"
>           EndSection
>           Section "Screen"
>                   Identifier      "Builtin Default vmware Screen 0"
>                   Device  "Builtin Default vmware Device 0"
>           EndSection
>   dmesg tells me that the segfault occured in "libdricore.so".
>   (2) The second setup is also a very fast one:
>   (--) PCI:*(0:0:2:0) 8086:2772:8086:464c Intel Corporation 82945G/GZ
>  Integrated Graphics Controller rev 2, Mem @ 0x90200000/524288,
>  0x80000000/268435456, 0x90280000/262144, I/O @ 0x000020e0/8, BIOS @
>  0x????????/131072
>   (==) Using default built-in configuration (30 lines)
>   (==) --- Start of built-in configuration ---
>           Section "Device"
>                   Identifier      "Builtin Default intel Device 0"
>                   Driver  "intel"
>           EndSection
>           Section "Screen"
>                   Identifier      "Builtin Default intel Screen 0"
>                   Device  "Builtin Default intel Device 0"
>           EndSection
>           Section "Device"
>                   Identifier      "Builtin Default vesa Device 0"
>                   Driver  "vesa"
>           EndSection
>           Section "Screen"
>                   Identifier      "Builtin Default vesa Screen 0"
>                   Device  "Builtin Default vesa Device 0"
>           EndSection
>           Section "Device"
>                   Identifier      "Builtin Default fbdev Device 0"
>   I was accessing this second machine using vncserver. Then I had the
>  same problem. Also here dmesg tells me that "libdricore.so" causes
> the
>  problem.
>   Inspired by your question, I run Gazebo while I was logged on
>  directly (from the console) this afternoon. Than it actually ran
>  successful! Apparently some problem with the (lack off) hardware
>  acceleration?
>   Thanks,
>   Barry
>   On Sun 25/04/10 10:03 , John Hsu [hidden email] [1] sent:
>   Hi Barry,
>   To help diagnose the problem, can you attach stap1.world?
>   also, what graphics card and driver are you using?
>   thanks,
>   John
>   On Sun, Apr 25, 2010 at 12:21 AM,   wrote:
>   Hi
>   everybody,
>    I'm trying to simulate some simple models with simple shapes but
>  get
>   segfaults in the visualisation part. The model contains a simple
>   plane, cylinder and directional light. I've attached the file
>   "stap1.world" that I'm using. When I'm starting gazebo with
>   stap1.world, Gazebo crashes with a segfault after successful
>   initialization. I tried this with several basic shapes (box,
>  cylinder,
>   cone, tube) but get the same behaviour. I've also tried it on
>   different machines with the same behaviour.
>    The part in the world file which is causing the trouble is:
>              unit_cylinder
>              Gazebo/RustyBarrel
>    If I remove this part from the world file, the gazebo does not
>   crash. But of course I don't see the cylinder.
>    Bug 2870446 might be related. But the patched file is no longer
>  part
>   of the trunk.
>    I'm not sure if this is a bug in Gazebo, Ogre or in one of
> drivers.
>   Or if the world model is plainly wrong. Hopefully somebody can help
>  me
>   with this one as it's really blocking for me and I'm pulling my
> hair
>   out.
>    Here's some more info:
>    - Gazebo trunk rev 8622
>    - Ogre 1.6.4
>    - gcc 4.4.3
>    For more information I've included below (1) world file; (2)
>   terminal output Gazebo; (3) output of gdb showing where the
> segfault
>   occurs.
>    Thanks,
>    Barry
>     This is the output I get from the terminal:
>    $gazebo stap1.world
>   Gazebo multi-robot simulator, version 0.10.0
>   Part of the Player/Stage Project
>   [http://playerstage.sourceforge.net [2]]. [1]
>   Copyright (C) 2003 Nate Koenig, Andrew Howard, and contributors.
>   Released under the GNU General Public License.
>
> [/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:103]
>     Gazebo Path[/usr/local/share/gazebo]
>
> [/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:115]
>     Ogre Path[/usr/lib/OGRE]
>   directory [/tmp/gazebo-barry-0] already xists (previous crash?)
>   but the owner gazebo server (pid=4430) is not running.
>   deleting the old information of the directory [/tmp/gazebo-barry-0]
>   Gazebo successfully initialized
>   Segmentation fault (core dumped)
>   Links:
>   ------
>   [1] http://playerstage.sourceforge.net [3]].
>
>
------------------------------------------------------------------------------

>   _______________________________________________
>   Playerstage-gazebo mailing list
>   https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo [5]
>  Links:
>  ------
>  [2] http://playerstage.sourceforge.net
>  [3] http://playerstage.sourceforge.net
>  [5] https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
>
>
------------------------------------------------------------------------------

>  _______________________________________________
>  Playerstage-gazebo mailing list
>  [hidden email] [2]
>  https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
>
>
> Links:
> ------
> [1] mailto:[hidden email]
> [2] mailto:[hidden email]
>
------------------------------------------------------------------------------
> _______________________________________________
> Playerstage-gazebo mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
>
------------------------------------------------------------------------------
_______________________________________________
Playerstage-gazebo mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
Reply | Threaded
Open this post in threaded view
|

Re: Segfault when using simple shapes

Scott Biddlestone-2
In reply to this post by sims-9
I had a similar problem with unit_cylinders and ATI drivers.

I posted a bug and a patch that fixed it for me at:
http://sourceforge.net/tracker/?func=detail&aid=2870446&group_id=42445&atid=433164

-Scott

On Sun, Apr 25, 2010 at 10:10 AM,  <[hidden email]> wrote:

> Somehow the attachment of stap1.world is not coming through. Here's
> the contents of the world model:
>   5
>     0.03
>     0 0 -9.8
>     10e-5
>     0.8
>     fltk
>     800 600
>     0 0
>     0.4 0.4 0.4 1.0
>       Gazebo/CloudySky
>     0 0 0
>     0 0 0
>     true
>         0 0 1
>         2000 2000
>         10 10
>         100 100
>         Gazebo/GrassFloor
>     5 0 0.5
>     0.0 0.0 0.0
>     false
>         0.5 1
>         1.0
>           unit_cylinder
>           Gazebo/RustyBarrel
>     true
>       directional
>       0 -0.8 -0.3
>       0.9 0.9 0.9
>       0.0 0.0 0.0
>       100
>       0.0 1.0 0.4
>  On Sun 25/04/10 14:13 , [hidden email] sent:
>  Hi John,
>  I've attached stap1.world (apparently something went wrong with
>  attaching it to the first mail).
>  I'm running Gazebo on two setups.
>  (1) The first is a VM Player setup (don't laugh).
>  From /var/log/xorg.log:
>  (--) PCI:*(0:0:15:0) 15ad:0405:15ad:0405 VMware Inc Abstract SVGA
> II
>  Adapter rev 0, Mem @ 0xd0000000/1
>  34217728, 0xd8000000/8388608, I/O @ 0x000010d0/16, BIOS @
>  0x????????/32768
>  (==) Using default built-in configuration (30 lines)
>  (==) --- Start of built-in configuration ---
>          Section "Device"
>                  Identifier      "Builtin Default vmware Device 0"
>                  Driver  "vmware"
>          EndSection
>          Section "Screen"
>                  Identifier      "Builtin Default vmware Screen 0"
>                  Device  "Builtin Default vmware Device 0"
>          EndSection
>  dmesg tells me that the segfault occured in "libdricore.so".
>  (2) The second setup is also a very fast one:
>  (--) PCI:*(0:0:2:0) 8086:2772:8086:464c Intel Corporation 82945G/GZ
>  Integrated Graphics Controller rev 2, Mem @ 0x90200000/524288,
>  0x80000000/268435456, 0x90280000/262144, I/O @ 0x000020e0/8, BIOS @
>  0x????????/131072
>  (==) Using default built-in configuration (30 lines)
>  (==) --- Start of built-in configuration ---
>          Section "Device"
>                  Identifier      "Builtin Default intel Device 0"
>                  Driver  "intel"
>          EndSection
>          Section "Screen"
>                  Identifier      "Builtin Default intel Screen 0"
>                  Device  "Builtin Default intel Device 0"
>          EndSection
>          Section "Device"
>                  Identifier      "Builtin Default vesa Device 0"
>                  Driver  "vesa"
>          EndSection
>          Section "Screen"
>                  Identifier      "Builtin Default vesa Screen 0"
>                  Device  "Builtin Default vesa Device 0"
>          EndSection
>          Section "Device"
>                  Identifier      "Builtin Default fbdev Device 0"
>  I was accessing this second machine using vncserver. Then I had the
>  same problem. Also here dmesg tells me that "libdricore.so" causes
> the
>  problem.
>  Inspired by your question, I run Gazebo while I was logged on
>  directly (from the console) this afternoon. Than it actually ran
>  successful! Apparently some problem with the (lack off) hardware
>  acceleration?
>  Thanks,
>  Barry
>  On Sun 25/04/10 10:03 , John Hsu [hidden email] [1] sent:
>  Hi Barry,
>  To help diagnose the problem, can you attach stap1.world?
>  also, what graphics card and driver are you using?
>  thanks,
>  John
>  On Sun, Apr 25, 2010 at 12:21 AM,   wrote:
>  Hi
>  everybody,
>   I'm trying to simulate some simple models with simple shapes but
>  get
>  segfaults in the visualisation part. The model contains a simple
>  plane, cylinder and directional light. I've attached the file
>  "stap1.world" that I'm using. When I'm starting gazebo with
>  stap1.world, Gazebo crashes with a segfault after successful
>  initialization. I tried this with several basic shapes (box,
>  cylinder,
>  cone, tube) but get the same behaviour. I've also tried it on
>  different machines with the same behaviour.
>   The part in the world file which is causing the trouble is:
>             unit_cylinder
>             Gazebo/RustyBarrel
>   If I remove this part from the world file, the gazebo does not
>  crash. But of course I don't see the cylinder.
>   Bug 2870446 might be related. But the patched file is no longer
>  part
>  of the trunk.
>   I'm not sure if this is a bug in Gazebo, Ogre or in one of
> drivers.
>  Or if the world model is plainly wrong. Hopefully somebody can help
>  me
>  with this one as it's really blocking for me and I'm pulling my
> hair
>  out.
>   Here's some more info:
>   - Gazebo trunk rev 8622
>   - Ogre 1.6.4
>   - gcc 4.4.3
>   For more information I've included below (1) world file; (2)
>  terminal output Gazebo; (3) output of gdb showing where the
> segfault
>  occurs.
>   Thanks,
>   Barry
>    This is the output I get from the terminal:
>   $gazebo stap1.world
>  Gazebo multi-robot simulator, version 0.10.0
>  Part of the Player/Stage Project
>  [http://playerstage.sourceforge.net [2]]. [1]
>  Copyright (C) 2003 Nate Koenig, Andrew Howard, and contributors.
>  Released under the GNU General Public License.
>
> [/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:103]
>    Gazebo Path[/usr/local/share/gazebo]
>
> [/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:115]
>    Ogre Path[/usr/lib/OGRE]
>  directory [/tmp/gazebo-barry-0] already xists (previous crash?)
>  but the owner gazebo server (pid=4430) is not running.
>  deleting the old information of the directory [/tmp/gazebo-barry-0]
>  Gazebo successfully initialized
>  Segmentation fault (core dumped)
>  Links:
>  ------
>  [1] http://playerstage.sourceforge.net [3]].
>
> ------------------------------------------------------------------------------
>  _______________________________________________
>  Playerstage-gazebo mailing list
>  https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo [5]
>  Links:
>  ------
>  [2] http://playerstage.sourceforge.net
>  [3] http://playerstage.sourceforge.net
>  [5] https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
>
> ------------------------------------------------------------------------------
>  _______________________________________________
>  Playerstage-gazebo mailing list
>  [hidden email] [2]
>  https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
>
>
> Links:
> ------
> [1] mailto:[hidden email]
> [2] mailto:[hidden email]
> ------------------------------------------------------------------------------
> _______________________________________________
> Playerstage-gazebo mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
>

------------------------------------------------------------------------------
_______________________________________________
Playerstage-gazebo mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
Reply | Threaded
Open this post in threaded view
|

Re: Segfault when using simple shapes

sims-9
In reply to this post by sims-9

 Hi John,
 Thanks for your advice. I'll try to get some decent hardware.
 Thanks,
 Barry
 On Mon 26/04/10 19:39 , John Hsu [hidden email] sent:
 Hi Bary,
 Your posting of the world file is not coming through :( but from my
experience intel and older ati graphics cards and drivers don't have
very good support for opengl/glx.   I don't think people have had much
success on vm's either.  I tried to compile a list of compatible
hardware here [1].
 Best,
 John
 On Sunday, April 25, 2010,   wrote:
 > Somehow the attachment of stap1.world is not coming through.
Here's
 > the contents of the world model:
 >    5
 >      0.03
 >      0 0 -9.8
 >      10e-5
 >      0.8
 >      fltk
 >      800 600
 >      0 0
 >      0.4 0.4 0.4 1.0
 >        Gazebo/CloudySky
 >      0 0 0
 >      0 0 0
 >      true
 >          0 0 1
 >          2000 2000
 >          10 10
 >          100 100
 >          Gazebo/GrassFloor
 >      5 0 0.5
 >      0.0 0.0 0.0
 >      false
 >          0.5 1
 >          1.0
 >            unit_cylinder
 >            Gazebo/RustyBarrel
 >      true
 >        directional
 >        0 -0.8 -0.3
 >        0.9 0.9 0.9
 >        0.0 0.0 0.0
 >        100
 >        0.0 1.0 0.4
 >  On Sun 25/04/10 14:13 ,  sent:
 >   Hi John,
 >   I've attached stap1.world (apparently something went wrong with
 >  attaching it to the first mail).
 >   I'm running Gazebo on two setups.
 >   (1) The first is a VM Player setup (don't laugh).
 >   From /var/log/xorg.log:
 >   (--) PCI:*(0:0:15:0) 15ad:0405:15ad:0405 VMware Inc Abstract
SVGA
 > II
 >  Adapter rev 0, Mem @ 0xd0000000/1
 >   34217728, 0xd8000000/8388608, I/O @ 0x000010d0/16, BIOS @
 >  0x????????/32768
 >   (==) Using default built-in configuration (30 lines)
 >   (==) --- Start of built-in configuration ---
 >           Section "Device"
 >                   Identifier      "Builtin Default vmware Device
0"
 >                   Driver  "vmware"
 >           EndSection
 >           Section "Screen"
 >                   Identifier      "Builtin Default vmware Screen
0"
 >                   Device  "Builtin Default vmware Device 0"
 >           EndSection
 >   dmesg tells me that the segfault occured in "libdricore.so".
 >   (2) The second setup is also a very fast one:
 >   (--) PCI:*(0:0:2:0) 8086:2772:8086:464c Intel Corporation
82945G/GZ
 >  Integrated Graphics Controller rev 2, Mem @ 0x90200000/524288,
 >  0x80000000/268435456, 0x90280000/262144, I/O @ 0x000020e0/8, BIOS
@
 >  0x????????/131072
 >   (==) Using default built-in configuration (30 lines)
 >   (==) --- Start of built-in configuration ---
 >           Section "Device"
 >                   Identifier      "Builtin Default intel Device 0"
 >                   Driver  "intel"
 >           EndSection
 >           Section "Screen"
 >                   Identifier      "Builtin Default intel Screen 0"
 >                   Device  "Builtin Default intel Device 0"
 >           EndSection
 >           Section "Device"
 >                   Identifier      "Builtin Default vesa Device 0"
 >                   Driver  "vesa"
 >           EndSection
 >           Section "Screen"
 >                   Identifier      "Builtin Default vesa Screen 0"
 >                   Device  "Builtin Default vesa Device 0"
 >           EndSection
 >           Section "Device"
 >                   Identifier      "Builtin Default fbdev Device 0"
 >   I was accessing this second machine using vncserver. Then I had
the
 >  same problem. Also here dmesg tells me that "libdricore.so"
causes
 > the
 >  problem.
 >   Inspired by your question, I run Gazebo while I was logged on
 >  directly (from the console) this afternoon. Than it actually ran
 >  successful! Apparently some problem with the (lack off) hardware
 >  acceleration?
 >   Thanks,
 >   Barry
 >   On Sun 25/04/10 10:03 , John Hsu  [1] sent:
 >   Hi Barry,
 >   To help diagnose the problem, can you attach stap1.world?
 >   also, what graphics card and driver are you using?
 >   thanks,
 >   John
 >   On Sun, Apr 25, 2010 at 12:21 AM,   wrote:
 >   Hi
 >   everybody,
 >    I'm trying to simulate some simple models with simple shapes
but
 >  get
 >   segfaults in the visualisation part. The model contains a simple
 >   plane, cylinder and directional light. I've attached the file
 >   "stap1.world" that I'm using. When I'm starting gazebo with
 >   stap1.world, Gazebo crashes with a segfault after successful
 >   initialization. I tried this with several basic shapes (box,
 >  cylinder,
 >   cone, tube) but get the same behaviour. I've also tried it on
 >   different machines with the same behaviour.
 >    The part in the world file which is causing the trouble is:
 >              unit_cylinder
 >              Gazebo/RustyBarrel
 >    If I remove this part from the world file, the gazebo does not
 >   crash. But of course I don't see the cylinder.
 >    Bug 2870446 might be related. But the patched file is no longer
 >  part
 >   of the trunk.
 >    I'm not sure if this is a bug in Gazebo, Ogre or in one of
 > drivers.
 >   Or if the world model is plainly wrong. Hopefully somebody can
help
 >  me
 >   with this one as it's really blocking for me and I'm pulling my
 > hair
 >   out.
 >    Here's some more info:
 >    - Gazebo trunk rev 8622
 >    - Ogre 1.6.4
 >    - gcc 4.4.3
 >    For more information I've included below (1) world file; (2)
 >   terminal output Gazebo; (3) output of gdb showing where the
 > segfault
 >   occurs.
 >    Thanks,
 >    Barry
 >     This is the output I get from the terminal:
 >    $gazebo stap1.world
 >   Gazebo multi-robot simulator, version 0.10.0
 >   Part of the Player/Stage Project
 >   [http://playerstage.sourceforge.net [5] [2]]. [1]
 >   Copyright (C) 2003 Nate Koenig, Andrew Howard, and contributors.
 >   Released under the GNU General Public License.
 >
 >
[/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:103]
 >     Gazebo Path[/usr/local/share/gazebo]
 >
 >
[/home/barry/src/playerstage/code/gazebo-trunk/server/GazeboConfig.cc:115]
 >     Ogre Path[/usr/lib/OGRE]
 >   directory [/tmp/gazebo-barry-0] already xists (previous crash?)
 >   but the owner gazebo server (pid=4430) is not running.
 >   deleting the old information of the directory
[/tmp/gazebo-barry-0]
 >   Gazebo successfully initialized
 >   Segmentation fault (core dumped)
 >   Links:
 >   ------
 >   [1] http://playerstage.sourceforge.net [6] [3]].
 >
 >
------------------------------------------------------------------------------
 >   _______________________________________________
 >   Playerstage-gazebo mailing list
 >   https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
[7] [5]
 >  Links:
 >  ------
 >  [2] http://playerstage.sourceforge.net [8]
 >  [3] http://playerstage.sourceforge.net [9]
 >  [5]
https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo [10]
 >
 >
------------------------------------------------------------------------------
 >  _______________________________________________
 >  Playerstage-gazebo mailing list
 >   [2]
 >  https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
[12]
 >
 >
 > Links:
 > ------
 > [1] mailto:
 > [2] mailto:
 >
------------------------------------------------------------------------------
 > _______________________________________________
 > Playerstage-gazebo mailing list
 >
 > https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
[16]
 >
 

Links:
------
[1] http://ros.org/wiki/simulator_gazebo/SystemRequirements
[5] http://playerstage.sourceforge.net
[6] http://playerstage.sourceforge.net
[7] https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
[8] http://playerstage.sourceforge.net
[9] http://playerstage.sourceforge.net
[10] https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
[12] https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
[16] https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
------------------------------------------------------------------------------
_______________________________________________
Playerstage-gazebo mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo