On 10/30/2013 06:23 PM, Carl Worth wrote:
A recent inspection of the latest OpenGL specification revealed that
there is no requirement for errors to be generated if GenQueries or
DeleteQueries are called while a query is active.
Specifically, in Section 4.2 of OpenGL 4.4 (Core Profile) the only
Static glfw is to simplify platforms that do not have good support
for waffle. To make use of this feature, Simply download the latest
GLFW release ( for example 3.0.3 ) and follow the next steps.
1) decompress the source release.
2) rename the source release folder to glfw.
3) verify the file
This signal handler is built using dbghelp support. This is found in most
versions of windows, and should work great on Windows XP or newer. I
have tested this against Windows 7 using MSVC and there is no major
problems. Also, as a benefit, this will automatically generate the
backtrace
GLFW also works on linux so allow proper fallback handling of tests.
This means Waffle to GLFW to GLUT.
---
CMakeLists.txt | 56 +++-
1 file changed, 27 insertions(+), 29 deletions(-)
diff --git a/CMakeLists.txt b/CMakeLists.txt
index
---
tests/egl/egl-create-surface.c |2 ++
tests/egl/egl-nok-swap-region.c |2 ++
tests/egl/egl-nok-texture-from-pixmap.c |2 ++
tests/egl/egl-query-surface.c |2 ++
4 files changed, 8 insertions(+)
diff --git a/tests/egl/egl-create-surface.c
Add Crash to the list of reports possible for programs using the piglit
utility library.
---
tests/util/piglit-util.c |5 +
tests/util/piglit-util.h |3 ++-
2 files changed, 7 insertions(+), 1 deletion(-)
diff --git a/tests/util/piglit-util.c b/tests/util/piglit-util.c
index
---
tests/glx/glx-close-display.c |3 +++
tests/glx/glx-copy-sub-buffer.c|2 ++
tests/glx/glx-destroycontext-1.c |4 +++-
tests/glx/glx-destroycontext-2.c |2 ++
tests/glx/glx-dont-care-mask.c |2 ++
I know this is a long patch and contains some features that have already
been discussed earlier this year. However each feature serves an important
task.
Feature 1: Signal Handler
Signal handlers may not seem like a lot, but on some platforms they are
a critical task to ensure piglit will
The find_package(PythonInterp) for cmake version 2.8.6 and newer is
capable of properly handling multiple versions of python. This is
extremely useful on mixed systems where a person may have Python2 and
Python3 installed side by side.
---
CMakeLists.txt | 27 ++-
1 file
Working with paths is hard, and we don't always do it right. Instead of
trying to be perfect, just use python's os.path module to do all of the
hard work for us
Signed-off-by: Dylan Baker baker.dyla...@gmail.com
---
framework/core.py | 9 +
framework/exectest.py | 2 +-
The signal handler in general is really important for platforms like
windows where a program will generate some dialog automatically when
running a program.
---
tests/util/piglit-framework-gl.h | 10 ++
1 file changed, 10 insertions(+)
diff --git a/tests/util/piglit-framework-gl.h
Another library set that does not directly link gainst piglitutil.
---
tests/hiz/CMakeLists.gl.txt |1 +
1 file changed, 1 insertion(+)
diff --git a/tests/hiz/CMakeLists.gl.txt b/tests/hiz/CMakeLists.gl.txt
index d1615fc..7005880 100644
--- a/tests/hiz/CMakeLists.gl.txt
+++
* Non-Posix compliant platforms - Only a stub for a signal handler exists.
* Posix platforms - Add support for simple backtrace support when a crash is
detected. This support does not interfere with valgrind and gdb testing.
When running a program in the debugger ( for example, GDB ), the singal
it's a good idea to make sure the dependency trees are properly set to
where all utilities are based on what is found in the noapi (piglitutil)
debugger.
---
tests/util/CMakeLists.gl.txt|4
tests/util/CMakeLists.gles1.txt |1 +
tests/util/CMakeLists.gles2.txt |1 +
Sometimes a user just wants to be able to painlessly get a visual on what
crashed.
---
framework/summary.py |9 +++--
1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/framework/summary.py b/framework/summary.py
index f5af627..0105bf0 100644
--- a/framework/summary.py
+++
Exercises an NVIDIA driver bug.
---
tests/all.tests |1 +
tests/texturing/CMakeLists.gl.txt |1 +
tests/texturing/cubemap-getteximage-pbo.c | 157 +
3 files changed, 159 insertions(+)
create mode 100644
On Thursday, October 31, 2013 09:48:09 AM Kenney Phillis wrote:
The find_package(PythonInterp) for cmake version 2.8.6 and newer is
capable of properly handling multiple versions of python. This is
extremely useful on mixed systems where a person may have Python2 and
Python3 installed side by
Exercises an NVIDIA driver bug related to glGetTexImage and sRGB
texture views.
---
tests/all.tests|1 +
tests/spec/arb_texture_view/CMakeLists.gl.txt |1 +
tests/spec/arb_texture_view/getteximage-srgb.c | 119
3 files changed,
On Thursday, October 31, 2013 09:48:07 AM Kenney Phillis wrote:
Sometimes a user just wants to be able to painlessly get a visual on what
crashed. ---
framework/summary.py |9 +++--
1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/framework/summary.py
On 10/29/2013 05:47 PM, Brian Paul wrote:
On 10/29/2013 03:55 PM, Jon Ashburn wrote:
Tests GL_ARB_viewport_array regarding the validity for the indices.
Use both valid and invalid parameters (index, first, count) for these
new
API entry points:
glViewportArrayv, glViewportIndexedf,
Sometimes we should handle scalar/vector types separately.
For example some builtin functions only support scalar type input,
and some relational builtin functions get different results between
scalar and vector type though with the same value.
Signed-off-by: Yi Sun yi@intel.com
diff --git
On Fri, Nov 01, 2013 at 12:20:47AM +0800, Yi Sun wrote:
According to the OpenCL spec version 1.2/1.1 chacpter 6.12.6:
The functions isequal, isnotequal, isgreater, isgreaterequal, isless,
islessequal, islessgreater, isfinite, isinf, isnan, isnormal,
isordered, isunordered and signbit described
On Friday, November 01, 2013 12:20:46 AM Yi Sun wrote:
Sometimes we should handle scalar/vector types separately.
For example some builtin functions only support scalar type input,
and some relational builtin functions get different results between
scalar and vector type though with the same
On 10/29/2013 05:47 PM, Brian Paul wrote:
On 10/29/2013 03:55 PM, Jon Ashburn wrote:
Test valid and invalid queries using glGetFLoati_v and glGetDoublei_v.
Also test correct queries of GL_SCISSOR_TEST default value and
settable values.
Tested on Nvidia Quadro 600 and all tests pass.
---
I have applied the suggestions that Dylan Baker suggested to the following
patches.
* framework/summary.py: add crashed page.
* cmake: use PythonInterp package and update version.
I will be submitting the updated patch set in a few days after I get more
feedback on the
other patches.
On Thu,
On 10/28/2013 04:53 PM, Zack Rusin wrote:
Increase the subpixel precision to 8 bits. This requires
changing some of the code to 64 bits to avoid overflows.
I'm not that familiar with this test. Does this change make this test
more strict? Does it make it more strict than the GL spec requires?
On 10/29/2013 02:55 PM, Jon Ashburn wrote:
Tests GL_ARB_viewport_array regarding the validity for the indices.
Use both valid and invalid parameters (index, first, count) for these new
API entry points:
glViewportArrayv, glViewportIndexedf, glViewportIndexedfv, glGetFloati_v.
Also test
Hi Tom,
Actually, I'm not sure if we can have two test under the same function name
'isnan'. The case maybe look like the following?
tests = {
'isnan' : {
'arg_types': [I, F],
'function_type': 'vvv',
'values': [
[0, -1,0, 0],#
Am 31.10.2013 20:34, schrieb Ian Romanick:
On 10/28/2013 04:53 PM, Zack Rusin wrote:
Increase the subpixel precision to 8 bits. This requires
changing some of the code to 64 bits to avoid overflows.
I'm not that familiar with this test. Does this change make this test
more strict? Does it
29 matches
Mail list logo
