AuroraMiddleware/gtk
1
0
Fork 1
mirror of https://gitlab.gnome.org/GNOME/gtk.git synced 2024-12-27 06:00:22 +00:00
Code Issues Projects Releases Wiki Activity
6d564075f3
gtk/gsk/gl/gskglcompiler.c

704 lines
20 KiB
C
Raw Normal View History

Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
/* gskglcompiler.c
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
*
* Copyright 2020 Christian Hergert <chergert@redhat.com>
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this program. If not, see <http://www.gnu.org/licenses/>.
*
* SPDX-License-Identifier: LGPL-2.1-or-later
*/
#include "config.h"
#include <gsk/gskdebugprivate.h>
#include <gio/gio.h>
#include <string.h>
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
#include "gskglcommandqueueprivate.h"
#include "gskglcompilerprivate.h"
#include "gskglprogramprivate.h"
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
gskglcompiler.c: Force GLSL version 300 es as needed For libANGLE to work with our shaders, we must use "300 es" for the #version directive in our shaders, as well as using the non-legacy/ non-GLES codepath in the shaders. In order to check whether we are using the GLSL 300 es shaders, we check whether we are using a GLES 3.0+ context. As a result, make ->glsl_version a const char* and make sure the existing shader version macros are defined apprpriately, and add a new macro for the "300 es" shader version string. This will allow the gtk4 programs to run under Windows using EGL via libANGLE. Some of the GL demos won't work for now, but at least this makes things a lot better for using GL-accelerated graphics under Windows for those that want to or need to use libANGLE (such as those with graphics drivers that aren't capable of our Desktop (W)GL requirements in GTK.
2021-11-23 10:22:55 +00:00
#define SHADER_VERSION_GLES "100"
#define SHADER_VERSION_GLES3 "300 es"
#define SHADER_VERSION_GL2_LEGACY "110"
#define SHADER_VERSION_GL3_LEGACY "130"
#define SHADER_VERSION_GL3 "150"
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
struct _GskGLCompiler
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
GObject parent_instance;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
GskGLDriver *driver;
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
GBytes *all_preamble;
GBytes *fragment_preamble;
GBytes *vertex_preamble;
GBytes *fragment_source;
GBytes *fragment_suffix;
GBytes *vertex_source;
GBytes *vertex_suffix;
GArray *attrib_locations;
gskglcompiler.c: Force GLSL version 300 es as needed For libANGLE to work with our shaders, we must use "300 es" for the #version directive in our shaders, as well as using the non-legacy/ non-GLES codepath in the shaders. In order to check whether we are using the GLSL 300 es shaders, we check whether we are using a GLES 3.0+ context. As a result, make ->glsl_version a const char* and make sure the existing shader version macros are defined apprpriately, and add a new macro for the "300 es" shader version string. This will allow the gtk4 programs to run under Windows using EGL via libANGLE. Some of the GL demos won't work for now, but at least this makes things a lot better for using GL-accelerated graphics under Windows for those that want to or need to use libANGLE (such as those with graphics drivers that aren't capable of our Desktop (W)GL requirements in GTK.
2021-11-23 10:22:55 +00:00
const char *glsl_version;
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
guint gl3 : 1;
guint gles : 1;
gsk: Define GSK_GLES3 if we have gles3 The compiler defines GSK_GLES only for gles2. Add another define for gles3, so we can differentiate where we need to. This took a week to find!
2023-10-22 15:00:52 +00:00
guint gles3 : 1;
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
guint legacy : 1;
guint debug_shaders : 1;
};
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
typedef struct _GskGLProgramAttrib
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
const char *name;
guint location;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
} GskGLProgramAttrib;
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
static GBytes *empty_bytes;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
G_DEFINE_TYPE (GskGLCompiler, gsk_gl_compiler, G_TYPE_OBJECT)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
static void
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_finalize (GObject *object)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
GskGLCompiler *self = (GskGLCompiler *)object;
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_clear_pointer (&self->all_preamble, g_bytes_unref);
g_clear_pointer (&self->fragment_preamble, g_bytes_unref);
g_clear_pointer (&self->vertex_preamble, g_bytes_unref);
g_clear_pointer (&self->vertex_suffix, g_bytes_unref);
g_clear_pointer (&self->fragment_source, g_bytes_unref);
g_clear_pointer (&self->fragment_suffix, g_bytes_unref);
g_clear_pointer (&self->vertex_source, g_bytes_unref);
g_clear_pointer (&self->attrib_locations, g_array_unref);
g_clear_object (&self->driver);
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
G_OBJECT_CLASS (gsk_gl_compiler_parent_class)->finalize (object);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
}
static void
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_class_init (GskGLCompilerClass *klass)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
GObjectClass *object_class = G_OBJECT_CLASS (klass);
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
object_class->finalize = gsk_gl_compiler_finalize;
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
empty_bytes = g_bytes_new (NULL, 0);
}
static void
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_init (GskGLCompiler *self)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
gskglcompiler.c: Force GLSL version 300 es as needed For libANGLE to work with our shaders, we must use "300 es" for the #version directive in our shaders, as well as using the non-legacy/ non-GLES codepath in the shaders. In order to check whether we are using the GLSL 300 es shaders, we check whether we are using a GLES 3.0+ context. As a result, make ->glsl_version a const char* and make sure the existing shader version macros are defined apprpriately, and add a new macro for the "300 es" shader version string. This will allow the gtk4 programs to run under Windows using EGL via libANGLE. Some of the GL demos won't work for now, but at least this makes things a lot better for using GL-accelerated graphics under Windows for those that want to or need to use libANGLE (such as those with graphics drivers that aren't capable of our Desktop (W)GL requirements in GTK.
2021-11-23 10:22:55 +00:00
self->glsl_version = "150";
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
self->attrib_locations = g_array_new (FALSE, FALSE, sizeof (GskGLProgramAttrib));
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
self->all_preamble = g_bytes_ref (empty_bytes);
self->vertex_preamble = g_bytes_ref (empty_bytes);
self->fragment_preamble = g_bytes_ref (empty_bytes);
self->vertex_source = g_bytes_ref (empty_bytes);
self->vertex_suffix = g_bytes_ref (empty_bytes);
self->fragment_source = g_bytes_ref (empty_bytes);
self->fragment_suffix = g_bytes_ref (empty_bytes);
}
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
GskGLCompiler *
gsk_gl_compiler_new (GskGLDriver *driver,
gboolean debug_shaders)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
GskGLCompiler *self;
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
GdkGLContext *context;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
g_return_val_if_fail (GSK_IS_GL_DRIVER (driver), NULL);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_return_val_if_fail (driver->shared_command_queue != NULL, NULL);
self = g_object_new (GSK_TYPE_GL_COMPILER, NULL);
self->driver = g_object_ref (driver);
self->debug_shaders = !!debug_shaders;
gsk: Cosmetics Get the context from the driver.
2023-05-16 18:53:20 +00:00
context = gsk_gl_driver_get_context (self->driver);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
if (gdk_gl_context_get_use_es (context))
{
gskglcompiler.c: Force GLSL version 300 es as needed For libANGLE to work with our shaders, we must use "300 es" for the #version directive in our shaders, as well as using the non-legacy/ non-GLES codepath in the shaders. In order to check whether we are using the GLSL 300 es shaders, we check whether we are using a GLES 3.0+ context. As a result, make ->glsl_version a const char* and make sure the existing shader version macros are defined apprpriately, and add a new macro for the "300 es" shader version string. This will allow the gtk4 programs to run under Windows using EGL via libANGLE. Some of the GL demos won't work for now, but at least this makes things a lot better for using GL-accelerated graphics under Windows for those that want to or need to use libANGLE (such as those with graphics drivers that aren't capable of our Desktop (W)GL requirements in GTK.
2021-11-23 10:22:55 +00:00
int maj, min;
/* for OpenGL/ES 3.0+, use "300 es" as our shader version */
gdk_gl_context_get_version (context, &maj, &min);
if (maj >= 3)
gsk: Define GSK_GLES3 if we have gles3 The compiler defines GSK_GLES only for gles2. Add another define for gles3, so we can differentiate where we need to. This took a week to find!
2023-10-22 15:00:52 +00:00
{
self->glsl_version = SHADER_VERSION_GLES3;
self->gles3 = TRUE;
}
gskglcompiler.c: Force GLSL version 300 es as needed For libANGLE to work with our shaders, we must use "300 es" for the #version directive in our shaders, as well as using the non-legacy/ non-GLES codepath in the shaders. In order to check whether we are using the GLSL 300 es shaders, we check whether we are using a GLES 3.0+ context. As a result, make ->glsl_version a const char* and make sure the existing shader version macros are defined apprpriately, and add a new macro for the "300 es" shader version string. This will allow the gtk4 programs to run under Windows using EGL via libANGLE. Some of the GL demos won't work for now, but at least this makes things a lot better for using GL-accelerated graphics under Windows for those that want to or need to use libANGLE (such as those with graphics drivers that aren't capable of our Desktop (W)GL requirements in GTK.
2021-11-23 10:22:55 +00:00
else
{
self->glsl_version = SHADER_VERSION_GLES;
self->gles = TRUE;
}
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
}
else if (gdk_gl_context_is_legacy (context))
{
int maj, min;
gdk_gl_context_get_version (context, &maj, &min);
gskglcompiler.c: Relax check for GL 3.x+ legacy contexts On Windows with nVidia drivers at least, when we create a legacy context via wglCreateContext(), we may still get a (W)GL 4.x context. Allow such contexts to also use GLSL version 130 instead of 110, so that things do continue to work.
2021-11-22 10:41:58 +00:00
/* On Windows, legacy contexts can give us a GL 4.x context */
if (maj >= 3)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
self->glsl_version = SHADER_VERSION_GL3_LEGACY;
else
self->glsl_version = SHADER_VERSION_GL2_LEGACY;
self->legacy = TRUE;
}
else
{
self->glsl_version = SHADER_VERSION_GL3;
self->gl3 = TRUE;
}
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_command_queue_make_current (self->driver->shared_command_queue);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
return g_steal_pointer (&self);
}
void
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_bind_attribute (GskGLCompiler *self,
const char *name,
guint location)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
GskGLProgramAttrib attrib;
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
g_return_if_fail (GSK_IS_GL_COMPILER (self));
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_return_if_fail (name != NULL);
g_return_if_fail (location < 32);
attrib.name = g_intern_string (name);
attrib.location = location;
g_array_append_val (self->attrib_locations, attrib);
}
void
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_clear_attributes (GskGLCompiler *self)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
g_return_if_fail (GSK_IS_GL_COMPILER (self));
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_array_set_size (self->attrib_locations, 0);
}
void
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_set_preamble (GskGLCompiler *self,
GskGLCompilerKind kind,
GBytes *preamble_bytes)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
GBytes **loc = NULL;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
g_return_if_fail (GSK_IS_GL_COMPILER (self));
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_return_if_fail (preamble_bytes != NULL);
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
if (kind == GSK_GL_COMPILER_ALL)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
loc = &self->all_preamble;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
else if (kind == GSK_GL_COMPILER_FRAGMENT)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
loc = &self->fragment_preamble;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
else if (kind == GSK_GL_COMPILER_VERTEX)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
loc = &self->vertex_preamble;
else
g_return_if_reached ();
g_assert (loc != NULL);
if (*loc != preamble_bytes)
{
g_clear_pointer (loc, g_bytes_unref);
*loc = preamble_bytes ? g_bytes_ref (preamble_bytes) : NULL;
}
}
void
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_set_preamble_from_resource (GskGLCompiler *self,
GskGLCompilerKind kind,
const char *resource_path)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
GError *error = NULL;
GBytes *bytes;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
g_return_if_fail (GSK_IS_GL_COMPILER (self));
g_return_if_fail (kind == GSK_GL_COMPILER_ALL ||
kind == GSK_GL_COMPILER_VERTEX ||
kind == GSK_GL_COMPILER_FRAGMENT);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_return_if_fail (resource_path != NULL);
bytes = g_resources_lookup_data (resource_path,
G_RESOURCE_LOOKUP_FLAGS_NONE,
&error);
if (bytes == NULL)
g_warning ("Cannot set shader from resource: %s", error->message);
else
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_set_preamble (self, kind, bytes);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_clear_pointer (&bytes, g_bytes_unref);
g_clear_error (&error);
}
void
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_set_source (GskGLCompiler *self,
GskGLCompilerKind kind,
GBytes *source_bytes)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
GBytes **loc = NULL;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
g_return_if_fail (GSK_IS_GL_COMPILER (self));
g_return_if_fail (kind == GSK_GL_COMPILER_ALL ||
kind == GSK_GL_COMPILER_VERTEX ||
kind == GSK_GL_COMPILER_FRAGMENT);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
if (source_bytes == NULL)
source_bytes = empty_bytes;
/* If kind is ALL, then we need to split the fragment and
* vertex shaders from the bytes and assign them individually.
* This safely scans for FRAGMENT_SHADER and VERTEX_SHADER as
* specified within the GLSL resources. Some care is taken to
* use GBytes which reference the original bytes instead of
* copying them.
*/
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
if (kind == GSK_GL_COMPILER_ALL)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
gsize len = 0;
const char *source;
const char *vertex_shader_start;
const char *fragment_shader_start;
const char *endpos;
GBytes *fragment_bytes;
GBytes *vertex_bytes;
g_clear_pointer (&self->fragment_source, g_bytes_unref);
g_clear_pointer (&self->vertex_source, g_bytes_unref);
source = g_bytes_get_data (source_bytes, &len);
endpos = source + len;
vertex_shader_start = g_strstr_len (source, len, "VERTEX_SHADER");
fragment_shader_start = g_strstr_len (source, len, "FRAGMENT_SHADER");
if (vertex_shader_start == NULL)
{
g_warning ("Failed to locate VERTEX_SHADER in shader source");
return;
}
if (fragment_shader_start == NULL)
{
g_warning ("Failed to locate FRAGMENT_SHADER in shader source");
return;
}
if (vertex_shader_start > fragment_shader_start)
{
g_warning ("VERTEX_SHADER must come before FRAGMENT_SHADER");
return;
}
/* Locate next newlines */
while (vertex_shader_start < endpos && vertex_shader_start[0] != '\n')
vertex_shader_start++;
while (fragment_shader_start < endpos && fragment_shader_start[0] != '\n')
fragment_shader_start++;
vertex_bytes = g_bytes_new_from_bytes (source_bytes,
vertex_shader_start - source,
fragment_shader_start - vertex_shader_start);
fragment_bytes = g_bytes_new_from_bytes (source_bytes,
fragment_shader_start - source,
endpos - fragment_shader_start);
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_set_source (self, GSK_GL_COMPILER_VERTEX, vertex_bytes);
gsk_gl_compiler_set_source (self, GSK_GL_COMPILER_FRAGMENT, fragment_bytes);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_bytes_unref (fragment_bytes);
g_bytes_unref (vertex_bytes);
return;
}
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
if (kind == GSK_GL_COMPILER_FRAGMENT)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
loc = &self->fragment_source;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
else if (kind == GSK_GL_COMPILER_VERTEX)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
loc = &self->vertex_source;
else
g_return_if_reached ();
if (*loc != source_bytes)
{
g_clear_pointer (loc, g_bytes_unref);
*loc = g_bytes_ref (source_bytes);
}
}
void
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_set_source_from_resource (GskGLCompiler *self,
GskGLCompilerKind kind,
const char *resource_path)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
GError *error = NULL;
GBytes *bytes;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
g_return_if_fail (GSK_IS_GL_COMPILER (self));
g_return_if_fail (kind == GSK_GL_COMPILER_ALL ||
kind == GSK_GL_COMPILER_VERTEX ||
kind == GSK_GL_COMPILER_FRAGMENT);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_return_if_fail (resource_path != NULL);
bytes = g_resources_lookup_data (resource_path,
G_RESOURCE_LOOKUP_FLAGS_NONE,
&error);
if (bytes == NULL)
g_warning ("Cannot set shader from resource: %s", error->message);
else
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_set_source (self, kind, bytes);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_clear_pointer (&bytes, g_bytes_unref);
g_clear_error (&error);
}
void
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_set_suffix (GskGLCompiler *self,
GskGLCompilerKind kind,
GBytes *suffix_bytes)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
GBytes **loc;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
g_return_if_fail (GSK_IS_GL_COMPILER (self));
g_return_if_fail (kind == GSK_GL_COMPILER_VERTEX ||
kind == GSK_GL_COMPILER_FRAGMENT);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_return_if_fail (suffix_bytes != NULL);
if (suffix_bytes == NULL)
suffix_bytes = empty_bytes;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
if (kind == GSK_GL_COMPILER_FRAGMENT)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
loc = &self->fragment_suffix;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
else if (kind == GSK_GL_COMPILER_VERTEX)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
loc = &self->vertex_suffix;
else
g_return_if_reached ();
if (*loc != suffix_bytes)
{
g_clear_pointer (loc, g_bytes_unref);
*loc = g_bytes_ref (suffix_bytes);
}
}
void
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_set_suffix_from_resource (GskGLCompiler *self,
GskGLCompilerKind kind,
const char *resource_path)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
GError *error = NULL;
GBytes *bytes;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
g_return_if_fail (GSK_IS_GL_COMPILER (self));
g_return_if_fail (kind == GSK_GL_COMPILER_VERTEX ||
kind == GSK_GL_COMPILER_FRAGMENT);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_return_if_fail (resource_path != NULL);
bytes = g_resources_lookup_data (resource_path,
G_RESOURCE_LOOKUP_FLAGS_NONE,
&error);
if (bytes == NULL)
g_warning ("Cannot set suffix from resource: %s", error->message);
else
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_compiler_set_suffix (self, kind, bytes);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_clear_pointer (&bytes, g_bytes_unref);
g_clear_error (&error);
}
static void
prepend_line_numbers (char *code,
GString *s)
{
char *p;
int line;
p = code;
line = 1;
while (*p)
{
char *end = strchr (p, '\n');
if (end)
end = end + 1; /* Include newline */
else
end = p + strlen (p);
g_string_append_printf (s, "%3d| ", line++);
g_string_append_len (s, p, end - p);
p = end;
}
}
static gboolean
check_shader_error (int shader_id,
GError **error)
{
GLint status;
GLint log_len;
GLint code_len;
char *buffer;
char *code;
GString *s;
glGetShaderiv (shader_id, GL_COMPILE_STATUS, &status);
if G_LIKELY (status == GL_TRUE)
return TRUE;
glGetShaderiv (shader_id, GL_INFO_LOG_LENGTH, &log_len);
buffer = g_malloc0 (log_len + 1);
glGetShaderInfoLog (shader_id, log_len, NULL, buffer);
glGetShaderiv (shader_id, GL_SHADER_SOURCE_LENGTH, &code_len);
code = g_malloc0 (code_len + 1);
glGetShaderSource (shader_id, code_len, NULL, code);
s = g_string_new ("");
prepend_line_numbers (code, s);
g_set_error (error,
GDK_GL_ERROR,
GDK_GL_ERROR_COMPILATION_FAILED,
"Compilation failure in shader.\n"
"Source Code: %s\n"
"\n"
"Error Message:\n"
"%s\n"
"\n",
s->str,
buffer);
g_string_free (s, TRUE);
g_free (buffer);
g_free (code);
return FALSE;
}
static void
print_shader_info (const char *prefix,
int shader_id,
const char *name)
{
if (GSK_DEBUG_CHECK(SHADERS))
{
int code_len;
glGetShaderiv (shader_id, GL_SHADER_SOURCE_LENGTH, &code_len);
if (code_len > 0)
{
char *code;
GString *s;
code = g_malloc0 (code_len + 1);
glGetShaderSource (shader_id, code_len, NULL, code);
s = g_string_new (NULL);
prepend_line_numbers (code, s);
g_message ("%s %d, %s:\n%s",
prefix, shader_id,
name ? name : "unnamed",
s->str);
g_string_free (s, TRUE);
g_free (code);
}
}
}
static const char *
get_shader_string (GBytes *bytes)
{
/* 0 length bytes will give us NULL back */
const char *str = g_bytes_get_data (bytes, NULL);
return str ? str : "";
}
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
GskGLProgram *
gsk_gl_compiler_compile (GskGLCompiler *self,
const char *name,
const char *clip,
GError **error)
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
{
char version[32];
const char *debug = "";
const char *legacy = "";
const char *gl3 = "";
const char *gles = "";
gsk: Define GSK_GLES3 if we have gles3 The compiler defines GSK_GLES only for gles2. Add another define for gles3, so we can differentiate where we need to. This took a week to find!
2023-10-22 15:00:52 +00:00
const char *gles3 = "";
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
int program_id;
int vertex_id;
int fragment_id;
int status;
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
g_return_val_if_fail (GSK_IS_GL_COMPILER (self), NULL);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_return_val_if_fail (self->all_preamble != NULL, NULL);
g_return_val_if_fail (self->fragment_preamble != NULL, NULL);
g_return_val_if_fail (self->vertex_preamble != NULL, NULL);
g_return_val_if_fail (self->fragment_source != NULL, NULL);
g_return_val_if_fail (self->vertex_source != NULL, NULL);
g_return_val_if_fail (self->driver != NULL, NULL);
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
gsk_gl_command_queue_make_current (self->driver->command_queue);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
gskglcompiler.c: Force GLSL version 300 es as needed For libANGLE to work with our shaders, we must use "300 es" for the #version directive in our shaders, as well as using the non-legacy/ non-GLES codepath in the shaders. In order to check whether we are using the GLSL 300 es shaders, we check whether we are using a GLES 3.0+ context. As a result, make ->glsl_version a const char* and make sure the existing shader version macros are defined apprpriately, and add a new macro for the "300 es" shader version string. This will allow the gtk4 programs to run under Windows using EGL via libANGLE. Some of the GL demos won't work for now, but at least this makes things a lot better for using GL-accelerated graphics under Windows for those that want to or need to use libANGLE (such as those with graphics drivers that aren't capable of our Desktop (W)GL requirements in GTK.
2021-11-23 10:22:55 +00:00
g_snprintf (version, sizeof version, "#version %s\n", self->glsl_version);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
if (self->debug_shaders)
debug = "#define GSK_DEBUG 1\n";
if (self->legacy)
legacy = "#define GSK_LEGACY 1\n";
if (self->gles)
ngl: fix typo from NGL rename
2021-03-01 18:28:26 +00:00
gles = "#define GSK_GLES 1\n";
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
gsk: Define GSK_GLES3 if we have gles3 The compiler defines GSK_GLES only for gles2. Add another define for gles3, so we can differentiate where we need to. This took a week to find!
2023-10-22 15:00:52 +00:00
if (self->gles3)
gles3 = "#define GSK_GLES3 1\n";
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
if (self->gl3)
ngl: fix typo from NGL rename
2021-03-01 18:28:26 +00:00
gl3 = "#define GSK_GL3 1\n";
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
vertex_id = glCreateShader (GL_VERTEX_SHADER);
glShaderSource (vertex_id,
gsk: Define GSK_GLES3 if we have gles3 The compiler defines GSK_GLES only for gles2. Add another define for gles3, so we can differentiate where we need to. This took a week to find!
2023-10-22 15:00:52 +00:00
11,
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
(const char *[]) {
gsk: Define GSK_GLES3 if we have gles3 The compiler defines GSK_GLES only for gles2. Add another define for gles3, so we can differentiate where we need to. This took a week to find!
2023-10-22 15:00:52 +00:00
version, debug, legacy, gl3, gles, gles3, clip,
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
get_shader_string (self->all_preamble),
get_shader_string (self->vertex_preamble),
get_shader_string (self->vertex_source),
get_shader_string (self->vertex_suffix),
},
(int[]) {
strlen (version),
strlen (debug),
strlen (legacy),
strlen (gl3),
strlen (gles),
gsk: Define GSK_GLES3 if we have gles3 The compiler defines GSK_GLES only for gles2. Add another define for gles3, so we can differentiate where we need to. This took a week to find!
2023-10-22 15:00:52 +00:00
strlen (gles3),
ngl: Add infrastructure for clip-variants of shaders Make ngl compile each shader three times. Once with #define NO_CLIP 1 once with #define RECT_CLIP 1 and once as before.
2021-03-07 22:50:19 +00:00
strlen (clip),
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_bytes_get_size (self->all_preamble),
g_bytes_get_size (self->vertex_preamble),
g_bytes_get_size (self->vertex_source),
g_bytes_get_size (self->vertex_suffix),
});
glCompileShader (vertex_id);
if (!check_shader_error (vertex_id, error))
{
glDeleteShader (vertex_id);
return NULL;
}
print_shader_info ("Vertex shader", vertex_id, name);
fragment_id = glCreateShader (GL_FRAGMENT_SHADER);
glShaderSource (fragment_id,
gsk: Define GSK_GLES3 if we have gles3 The compiler defines GSK_GLES only for gles2. Add another define for gles3, so we can differentiate where we need to. This took a week to find!
2023-10-22 15:00:52 +00:00
11,
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
(const char *[]) {
gsk: Define GSK_GLES3 if we have gles3 The compiler defines GSK_GLES only for gles2. Add another define for gles3, so we can differentiate where we need to. This took a week to find!
2023-10-22 15:00:52 +00:00
version, debug, legacy, gl3, gles, gles3, clip,
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
get_shader_string (self->all_preamble),
get_shader_string (self->fragment_preamble),
get_shader_string (self->fragment_source),
get_shader_string (self->fragment_suffix),
},
(int[]) {
strlen (version),
strlen (debug),
strlen (legacy),
strlen (gl3),
strlen (gles),
gsk: Define GSK_GLES3 if we have gles3 The compiler defines GSK_GLES only for gles2. Add another define for gles3, so we can differentiate where we need to. This took a week to find!
2023-10-22 15:00:52 +00:00
strlen (gles3),
ngl: Add infrastructure for clip-variants of shaders Make ngl compile each shader three times. Once with #define NO_CLIP 1 once with #define RECT_CLIP 1 and once as before.
2021-03-07 22:50:19 +00:00
strlen (clip),
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
g_bytes_get_size (self->all_preamble),
g_bytes_get_size (self->fragment_preamble),
g_bytes_get_size (self->fragment_source),
g_bytes_get_size (self->fragment_suffix),
});
glCompileShader (fragment_id);
if (!check_shader_error (fragment_id, error))
{
glDeleteShader (vertex_id);
glDeleteShader (fragment_id);
return NULL;
}
print_shader_info ("Fragment shader", fragment_id, name);
program_id = glCreateProgram ();
glAttachShader (program_id, vertex_id);
glAttachShader (program_id, fragment_id);
for (guint i = 0; i < self->attrib_locations->len; i++)
{
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
const GskGLProgramAttrib *attrib;
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
attrib = &g_array_index (self->attrib_locations, GskGLProgramAttrib, i);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
glBindAttribLocation (program_id, attrib->location, attrib->name);
}
glLinkProgram (program_id);
glGetProgramiv (program_id, GL_LINK_STATUS, &status);
glDetachShader (program_id, vertex_id);
glDeleteShader (vertex_id);
glDetachShader (program_id, fragment_id);
glDeleteShader (fragment_id);
if (status == GL_FALSE)
{
char *buffer = NULL;
int log_len = 0;
glGetProgramiv (program_id, GL_INFO_LOG_LENGTH, &log_len);
if (log_len > 0)
{
/* log_len includes NULL */
buffer = g_malloc0 (log_len);
glGetProgramInfoLog (program_id, log_len, NULL, buffer);
}
g_warning ("Linking failure in shader:\n%s",
buffer ? buffer : "");
g_set_error (error,
GDK_GL_ERROR,
GDK_GL_ERROR_LINK_FAILED,
"Linking failure in shader: %s",
buffer ? buffer : "");
g_free (buffer);
glDeleteProgram (program_id);
return NULL;
}
Rename ngl to gl We have only one gl renderer now, and it is a bit odd for it not be called gl.
2021-10-07 03:15:25 +00:00
return gsk_gl_program_new (self->driver, name, program_id);
gsk: add OpenGL based GskNglRenderer The primary goal here was to cleanup the current GL renderer to make maintenance easier going forward. Furthermore, it tracks state to allow us to implement more advanced renderer features going forward. Reordering This renderer will reorder batches by render target to reduce the number of times render targets are changed. In the future, we could also reorder by program within the render target if we can determine that vertices do not overlap. Uniform Snapshots To allow for reordering of batches all uniforms need to be tracked for the programs. This allows us to create the full uniform state when the batch has been moved into a new position. Some care was taken as it can be performance sensitive. Attachment Snapshots Similar to uniform snapshots, we need to know all of the texture attachments so that we can rebind them when necessary. Render Jobs To help isolate the process of creating GL commands from the renderer abstraction a render job abstraction was added. This could be extended in the future if we decided to do tiling. Command Queue Render jobs create batches using the command queue. The command queue will snapshot uniform and attachment state so that it can reorder batches right before executing them. Currently, the only reordering done is to ensure that we only visit each render target once. We could extend this by tracking vertices, attachments, and others. This code currently uses an inline array helper to reduce overhead from GArray which was showing up on profiles. It could be changed to use GdkArray without too much work, but had roughly double the instructions. Cycle counts have not yet been determined. GLSL Programs This was simplified to use XMACROS so that we can just extend one file (gskglprograms.defs) instead of multiple places. The programs are added as fields in the driver for easy access. Driver The driver manages textures, render targets, access to atlases, programs, and more. There is one driver per display, by using the shared GL context. Some work could be done here to batch uploads so that we make fewer calls to upload when sending icon theme data to the GPU. We'd need to keep a copy of the atlas data for such purposes.
2020-12-19 01:36:59 +00:00
}
Reference in New Issue Copy Permalink