GPU laser crashes with more than 16k samples

[osrf-migration]

Original report (archived issue) by Louise Poubel (Bitbucket: chapulina, GitHub: chapulina).

---

Prerequisites

• [X] Put an X between the brackets on this line if you have done all of the following:
  • Checked the Q&A board for common solutions: http://answers.gazebosim.org
  • Checked that your issue isn't already filed.
  • Checked that there is not already an Ignition package that provides the described functionality: https://ignitionrobotics.org/libs

Description

GPU laser crashes with more than 16k horizontal samples.

Steps to Reproduce

Try this SDF:

            <sensor name='front_laser' type='gpu_ray'>
              <pose>0.08 0 0.3 0 0 0</pose>
              <update_rate>15</update_rate>
              <topic>/x1/points</topic>
              <lidar>
                <scan>
                  <horizontal>
                    <samples>18000</samples>
                    <resolution>0.1</resolution>
                    <min_angle>-3.1459</min_angle>
                    <max_angle>3.1459</max_angle>
                  </horizontal>
                  <vertical>
                    <samples>16</samples>
                    <resolution>1</resolution>
                    <min_angle>-0.261799</min_angle>
                    <max_angle>0.261799</max_angle>
                  </vertical>
                </scan>
                <range>
                  <min>0.05</min>
                  <max>75</max>
                  <resolution>0.01</resolution>
                </range>
                <noise>
                  <type>gaussian</type>
                  <mean>0</mean>
                  <stddev>0.01</stddev>
                </noise>
              </lidar>
            </sensor>

Expected behavior:

Sensor is properly created with no crash. If you change the 18000 above to 16000 there's no crash.

Actual behavior:

Crash with:

terminate called after throwing an instance of 'Ogre::RenderingAPIException'
  what():  OGRE EXCEPTION(3:RenderingAPIException): Zero sized texture surface on texture x1::base::front_laser_samplerTex face 0 mipmap 0. The GL driver probably refused to create the texture. in GL3PlusTexture::_createSurfaceList at /var/lib/jenkins/workspace/ogre-2.1-debbuilder/repo/RenderSystems/GL3Plus/src/OgreGL3PlusTexture.cpp (line 516)

Reproduces how often:

100%

Versions

ign-sensors2

Additional Information

I talked to @iche033 and apparently large sample counts are redundant anyway, since the underlying computations use a cube with 1024x1024 sized walls.

We should at least prevent the crash, but a better solution may involve interpolating the samples.

[osrf-migration]

Original comment by Louise Poubel (Bitbucket: chapulina, GitHub: chapulina).

---

• Edited issue description