Update the GL blit code to emit correct shaders when integer color
formats are involved.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
When bounding a blit within the source resource extents, whether a
coordinate bound is inclusive or exclusive depends on the direction of
the read. Take this direction into account and bound the coordinates
accordingly.
Fixes: piglit fbo-blit-stretch
Reviewed-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Ensure we don't blit pixels originating from outside the extents of the
source resource. To achieve this we clip the source blit region, and
update the destination blit region accordingly, while respecting scaled
blits.
This fix will allow us to introduce swizzles for some formats in
upcoming commits, and thus use the GL blitter for them, without
introducing any regressions
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
Signed-off-by: Jakob Bornecrantz <jakob@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Tested-by: Gurchetan Singh <gurchetansingh@chromium.org>
Update the blit shader creation code to take the destination format
swizzle into account. This involves inverting the swizzle information
and using the inverted swizzle to reorder the texel components in the
shader. The new swizzle code also removes the need for having a special
case if the destination format uses emulated alpha.
This commit temporarily removes the VIRGL_FORMAT_R8G8B8X8_UNORM special
case. Improved support for it, and other similar "ignore alpha" cases,
is added back in an upcoming commit using appropriate swizzles.
At the moment we always recreate the shader in case we have a swizzle,
but a more efficient shader caching scheme could be devised to improve
this.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
Signed-off-by: Jakob Bornecrantz <jakob@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Tested-by: Gurchetan Singh <gurchetansingh@chromium.org>
Use the source format swizzle information to set the
GL_TEXTURE_SWIZZLE_* parameters for the GL blit operation. This also
removes the need for the emulated alpha special case, since when using
emulated alpha the source format already has proper swizzle information.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
Signed-off-by: Jakob Bornecrantz <jakob@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Tested-by: Gurchetan Singh <gurchetansingh@chromium.org>
Ignores alpha channel while blitting R8G8B8X8_UNIFORM colorspace.
framebuffer It would be nice to perform some code refactoring in
vrend_blitter to minimize dup code. Fixes following test:
dEQP-GLES3.functional.fbo.blit.conversion.rgb8_to_rgba8']
dEQP-GLES3.functional.fbo.blit.conversion.rgb8_to_rgb10_a2
dEQP-GLES3.functional.fbo.blit.conversion.rgb8_to_rgba4']
Tested-by: Jakob Bornecrantz <jakob@collabora.com>
Reviewed-by: Jakob Bornecrantz <jakob@collabora.com>
Signed-off-by: Jakob Bornecrantz <jakob@collabora.com>
In the case of PIPE_TEXTURE_CUBE, util_gen_mipmap (in Mesa)
encodes the target face in blit.{src,dst}.box.z. We need to
bind to the correct texture target before we draw.
Also removed extraneous whitespace.
Fixes:
dEQP-GLES2.functional.texture.mipmap.cube.generate.a8_fastest
dEQP-GLES2.functional.texture.mipmap.cube.generate.a8_nicest
dEQP-GLES2.functional.texture.specification.basic_copyteximage2d.cube_alpha
dEQP-GLES2.functional.texture.specification.basic_copytexsubimage2d.cube_alpha
[airlied: also fixes piglit:
spec@arb_texture_cube_map@copyteximage cube]
v2: Set layer to dst_z for cases other than GL_CUBE_MAP
Signed-off-by: Dave Airlie <airlied@redhat.com>
When emulating GL_ALPHA8 with GL_R8, we should make the alpha
component equal to the red color channel.
Fixes:
['dEQP-GLES2.functional.texture.mipmap.2d.generate.a8_fastest']
['dEQP-GLES2.functional.texture.mipmap.2d.generate.a8_nicest']
on OpenGL 4.5 (core).
Tested-by: Jakob Bornecrantz <jakob@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Currently, we always try to create an OpenGL 3.1 context. Some
dEQP tests require an OpenGL 3.2 context (specifically, ones
that use glGetInteger64v). Let's try to create the highest
version context we can, and iterate to lower versions, i.e:
https://developer.android.com/guide/topics/graphics/opengl.html#version-check
The return code for (*create_gl_context) is a little unclear.
This patch assumes NULL is returned on failure. This should work
for GLX and EGL.
GLX:
"On failure glXCreateContextAttribsARB returns NULL and generates
an X error with extended error information"
https://www.khronos.org/registry/OpenGL/extensions/ARB/GLX_ARB_create_context.txt
EGL:
"#define EGL_NO_CONTEXT ((EGLContext)0)"
https://www.khronos.org/registry/EGL/api/1.1/EGL/egl.h
The semantics of rcbs->create_gl_context may be different, though.
Fixes:
dEQP-GLES3.functional.state_query.integers.max_vertex_output_components_getinteger64
dEQP-GLES3.functional.state_query.integers.max_vertex_output_components_getfloat
Signed-off-by: Dave Airlie <airlied@redhat.com>
OpenGL ES does not have glDrawBuffer, but it does have glDrawBuffers. It has
been available in Desktop OpenGL since 2.0 and was exposed in OpenGL ES 3.0.
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Jakob Bornecrantz <jakob.bornecrantz@collabora.com>
The 'blit_ctx->initialised' is not setted to true.
Every time init blit context, it will create a new
'blit_ctx->gl_context' thus causing a memory leak.
This patch avoid this.
Signed-off-by: Li Qiang <liq3ea@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Alexandros Frantzis
Robert Tarasov
Gurchetan Singh
Jakob Bornecrantz
Jakob Bornecrantz
Li Qiang
Dave Airlie
Marc-André Lureau
Dave Airlie
Marc-André Lureau