Add error checking and fix handling of level (it should be negative).
Besides, always use the palette entry format/type when calling
_mesa_TexImage2D. It respects the base internal formats of the cpal
formats, and is simpler and faster, except for cases where the unpack
alignment needs to be changed.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
In GL_OES_framebuffer_object and OpenGL ES 2.0, GL_RGB565 is a valid
internal format. Since it is not supported by the core, map it to
GL_RGB5 as a workaround.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Unlike in OpenGL, GL_OES_texture_cube_map says that all coordinates are
changed the same time by the token GL_TEXTURE_GEN_STR_OES, and the
initial mode is GL_REFLECTION_MAP_OES.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
This commit adds definitions of
GL_EXT_texture_compression_dxt1
GL_EXT_texture_lod_bias
GL_EXT_blend_minmax
GL_EXT_multi_draw_arrays
to APIspec.xml and get_gen.py. Some of the enums are not avaiable in
the header files and the defining extensions are disabled.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
es_generator.py did not generate code for special functions. They were
supposed to be defined elsewhere. But as a result, parameter checking
was also skipped. This commit changes the way special functions are
called so that parameter checking is always performed.
When there is nothing to check, the check functions becomes macros
expanding to the real functions, as an optimization.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Have get_gen.py generate _mesa_GetFixedv. The generated function does
all the checkings and thus simplifies APIspec.xml a bit.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
APIspec.xml is based on APIspec.txt. The new format has less code
duplications and should be easier to read.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Build demos that require EGL_MESA_screen_surface, and build egl_i915.so
that supports EGL_MESA_screen_surface.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
This is done by stopping linking to libmesagallium.a and removing DRI
related stuff. The state tracker an application links to decides the
API supported.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Running make with -j for the first time might fail because glapi headers
haven't been generated. This commit should make it more reliable.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
drawtex.c was listed in LOCAL_ES1_SOURCES twice. My mistake when
merging the patches.
Also, run gl_apitemp.py with -c to silence warnings and add target
"install".
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
When the mode is on, the scripts would generate headers that are
suitable for OpenGL ES. There are two differences. One is that they
will generate function prototypes for OpenGL ES specific functions. The
other is that, when a function has multiple names, SET/GET/CALL macros
would be generated for each of names.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
The demo programs are written by Brian Paul, and cherry-picked from
opengl-es branch. Several minor issues such as a linking problem are
fixed.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
This is primitive support for OpenGL ES. It uses a subset of mesa
sources to build libesXgallium.a and libesXapi.a, where X is 1 for
OpenGL ES 1.x, 2 for OpenGL ES 2.x. The static libraries serve the same
purpose as libmesagallium.a and libglapi.a do for OpenGL.
This is based on the work of opengl-es branch.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
When GLX_INDIRECT_RENDERING is defined, some symbols are used in
libglapi.a but are not defined. Define them through the help of
glapitemp.h.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
When GLX_INDIRECT_RENDERING is defined, some symbols are used in
libglapi.a but are not defined. Define them through the help of
glapitemp.h.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reorganize glapitemp.h such that it is possible to skip normal entry
points or protocol entry points by defining
_GLAPI_SKIP_NORMAL_ENTRY_POINTS or _GLAPI_SKIP_PROTO_ENTRY_POINTS.
Protocol entry points are those with different GLX protocols. They are
skipped in libglapi.a when GLX_INDIRECT_RENDERING is defined.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Suggested by Joakim Sindholt.
Also, put flushing of colorbuffers _before_ the framebuffer state setup,
suggested by docs.
Signed-off-by: Corbin Simpson <MostAwesomeDude@gmail.com>
The exec->vtx.inputs[] array was being written past its end. This was
clobbering the following vbo_exec_context::eval state. Probably not noticed
since evaluators and immediate mode rendering don't happen at the same time.
Fixed the loop in vbo_exec_vtx_init().
Changed the size of the vbo_exec_context::vtx.arrays[] array.
Added a bunch of debug-build assertions.
Issue found by Vinson Lee.
