xserver-multidpi/glamor/glamor_priv.h

995 lines
27 KiB
C
Raw Normal View History

/*
* Copyright © 2008 Intel Corporation
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice (including the next
* paragraph) shall be included in all copies or substantial portions of the
* Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
* IN THE SOFTWARE.
*
* Authors:
* Eric Anholt <eric@anholt.net>
*
*/
#ifndef GLAMOR_PRIV_H
#define GLAMOR_PRIV_H
#ifdef HAVE_DIX_CONFIG_H
#include <dix-config.h>
#include <xorg-config.h>
#endif
#include "glamor.h"
#ifdef GLAMOR_GLES2
#define GL_BGRA GL_BGRA_EXT
#define GL_COLOR_INDEX 0x1900
#define GL_BITMAP 0x1A00
#define GL_UNSIGNED_INT_8_8_8_8 0x8035
#define GL_UNSIGNED_INT_8_8_8_8_REV 0x8367
#define GL_UNSIGNED_INT_2_10_10_10_REV 0x8368
#define GL_UNSIGNED_INT_10_10_10_2 0x8036
#define GL_UNSIGNED_SHORT_5_6_5_REV 0x8364
#define GL_UNSIGNED_SHORT_1_5_5_5_REV 0x8366
#define GL_UNSIGNED_SHORT_4_4_4_4_REV 0x8365
#define GL_PIXEL_PACK_BUFFER 0x88EB
#define GL_PIXEL_UNPACK_BUFFER 0x88EC
#define GL_CLAMP_TO_BORDER 0x812D
#define GL_READ_WRITE 0x88BA
#define GL_READ_ONLY 0x88B8
#define GL_WRITE_ONLY 0x88B9
#define GL_STREAM_DRAW 0x88E0
#define GL_STREAM_READ 0x88E1
#define GL_PACK_ROW_LENGTH 0x0D02
#define GL_UNPACK_ROW_LENGTH 0x0CF2
#define GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER 0x8CDB
#define GL_FRAMEBUFFER_INCOMPLETE_READ_BUFFER 0x8CDC
#define GL_FRAMEBUFFER_INCOMPLETE_MULTISAMPLE 0x8D56
#define GL_PACK_INVERT_MESA 0x8758
#define glMapBuffer(x, y) NULL
#define glUnmapBuffer(x)
#define glRasterPos2i(x,y)
#define glDrawPixels(x,y,z,a,b)
#endif
#define GL_GLEXT_PROTOTYPES
#ifdef GLAMOR_GLES2
#include <GLES2/gl2.h>
#include <GLES2/gl2ext.h>
#define GLAMOR_DEFAULT_PRECISION \
"precision mediump float;\n"
#else
#include <GL/gl.h>
#include <GL/glext.h>
#define GLAMOR_DEFAULT_PRECISION
#endif
#ifdef RENDER
#include "glyphstr.h"
#endif
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
#include "glamor_debug.h"
#define glamor_check_fbo_size(_glamor_,_w_, _h_) ((_w_) > 0 && (_h_) > 0 \
&& (_w_) < _glamor_->max_fbo_size \
&& (_h_) < _glamor_->max_fbo_size)
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
#define glamor_check_fbo_depth(_depth_) ( \
_depth_ == 8 \
|| _depth_ == 15 \
|| _depth_ == 16 \
|| _depth_ == 24 \
|| _depth_ == 30 \
|| _depth_ == 32)
#define GLAMOR_PIXMAP_PRIV_IS_PICTURE(pixmap_priv) (pixmap_priv->is_picture == 1)
#define GLAMOR_PIXMAP_PRIV_HAS_FBO(pixmap_priv) (pixmap_priv->gl_fbo == 1)
2009-08-20 21:32:53 +02:00
typedef struct glamor_transform_uniforms {
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
GLint x_bias;
GLint x_scale;
GLint y_bias;
GLint y_scale;
2009-08-20 21:32:53 +02:00
} glamor_transform_uniforms;
typedef struct glamor_composite_shader {
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
GLuint prog;
GLint dest_to_dest_uniform_location;
GLint dest_to_source_uniform_location;
GLint dest_to_mask_uniform_location;
GLint source_uniform_location;
GLint mask_uniform_location;
} glamor_composite_shader;
typedef struct {
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
INT16 x_src;
INT16 y_src;
INT16 x_mask;
INT16 y_mask;
INT16 x_dst;
INT16 y_dst;
INT16 width;
INT16 height;
} glamor_composite_rect_t;
#define GLAMOR_NUM_GLYPH_CACHES 4
#define GLAMOR_NUM_GLYPH_CACHE_FORMATS 2
typedef struct {
PicturePtr picture; /* Where the glyphs of the cache are stored */
GlyphPtr *glyphs;
uint16_t count;
uint16_t evict;
} glamor_glyph_cache_t;
enum glamor_vertex_type {
GLAMOR_VERTEX_POS,
GLAMOR_VERTEX_SOURCE,
GLAMOR_VERTEX_MASK
};
enum shader_source {
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
SHADER_SOURCE_SOLID,
SHADER_SOURCE_TEXTURE,
SHADER_SOURCE_TEXTURE_ALPHA,
SHADER_SOURCE_COUNT,
};
enum shader_mask {
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
SHADER_MASK_NONE,
SHADER_MASK_SOLID,
SHADER_MASK_TEXTURE,
SHADER_MASK_TEXTURE_ALPHA,
SHADER_MASK_COUNT,
};
enum shader_in {
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
SHADER_IN_SOURCE_ONLY,
SHADER_IN_NORMAL,
SHADER_IN_CA_SOURCE,
SHADER_IN_CA_ALPHA,
SHADER_IN_COUNT,
};
#include "glamor_gl_dispatch.h"
struct glamor_screen_private;
struct glamor_pixmap_private;
typedef void (*glamor_pixmap_validate_function_t)(struct glamor_screen_private*,
struct glamor_pixmap_private*);
enum glamor_gl_flavor {
GLAMOR_GL_DESKTOP, // OPENGL API
GLAMOR_GL_ES2 // OPENGL ES2.0 API
};
#define GLAMOR_CREATE_PIXMAP_CPU 0x100
typedef struct glamor_screen_private {
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
CloseScreenProcPtr saved_close_screen;
CreateGCProcPtr saved_create_gc;
CreatePixmapProcPtr saved_create_pixmap;
DestroyPixmapProcPtr saved_destroy_pixmap;
GetSpansProcPtr saved_get_spans;
GetImageProcPtr saved_get_image;
CompositeProcPtr saved_composite;
TrapezoidsProcPtr saved_trapezoids;
GlyphsProcPtr saved_glyphs;
ChangeWindowAttributesProcPtr saved_change_window_attributes;
CopyWindowProcPtr saved_copy_window;
BitmapToRegionProcPtr saved_bitmap_to_region;
TrianglesProcPtr saved_triangles;
CreatePictureProcPtr saved_create_picture;
DestroyPictureProcPtr saved_destroy_picture;
UnrealizeGlyphProcPtr saved_unrealize_glyph;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
int yInverted;
int screen_fbo;
GLuint vbo;
int vbo_offset;
int vbo_size;
char *vb;
int vb_stride;
enum glamor_gl_flavor gl_flavor;
int has_pack_invert;
int has_fbo_blit;
int max_fbo_size;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
/* glamor_finishaccess */
GLint finish_access_prog[2];
GLint finish_access_no_revert[2];
GLint finish_access_swap_rb[2];
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
/* glamor_solid */
GLint solid_prog;
GLint solid_color_uniform_location;
/* glamor_tile */
GLint tile_prog;
/* glamor_putimage */
GLint put_image_xybitmap_prog;
glamor_transform_uniforms put_image_xybitmap_transform;
GLint put_image_xybitmap_fg_uniform_location;
GLint put_image_xybitmap_bg_uniform_location;
/* glamor_composite */
glamor_composite_shader composite_shader[SHADER_SOURCE_COUNT]
[SHADER_MASK_COUNT]
[SHADER_IN_COUNT];
Bool has_source_coords, has_mask_coords;
int render_nr_verts;
glamor_pixmap_validate_function_t *pixmap_validate_funcs;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
glamor_glyph_cache_t glyph_caches[GLAMOR_NUM_GLYPH_CACHES];
char delayed_fallback_string[GLAMOR_DELAYED_STRING_MAX + 1];
int delayed_fallback_pending;
glamor_glyph_cache_t glyphCaches[GLAMOR_NUM_GLYPH_CACHE_FORMATS];
Bool glyph_cache_initialized;
struct glamor_gl_dispatch dispatch;
} glamor_screen_private;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
typedef enum glamor_access {
GLAMOR_ACCESS_RO,
GLAMOR_ACCESS_RW,
GLAMOR_ACCESS_WO,
} glamor_access_t;
/*
* glamor_pixmap_private - glamor pixmap's private structure.
* @gl_fbo: The pixmap is attached to a fbo originally.
* @gl_tex: The pixmap is in a gl texture originally.
* @pbo_valid: The pbo has a valid copy of the pixmap's data.
* @is_picture: The drawable is attached to a picture.
* @tex: attached texture.
* @fb: attached fbo.
* @pbo: attached pbo.
* @access_mode: access mode during the prepare/finish pair.
* @pict_format: the corresponding picture's format.
* #pending_op: currently only support pending filling.
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
* @container: The corresponding pixmap's pointer.
**/
#define GLAMOR_PIXMAP_PRIV_NEED_VALIDATE(pixmap_priv) \
(GLAMOR_PIXMAP_PRIV_HAS_FBO(pixmap_priv) \
&& (pixmap_priv->pending_op.type != GLAMOR_PENDING_NONE))
#define GLAMOR_PIXMAP_PRIV_NO_PENDING(pixmap_priv) \
(pixmap_priv->pending_op.type == GLAMOR_PENDING_NONE)
enum _glamor_pending_op_type{
GLAMOR_PENDING_NONE,
GLAMOR_PENDING_FILL
};
typedef struct _glamor_pending_fill {
unsigned int type;
GLfloat color4fv[4];
CARD32 colori;
} glamor_pending_fill;
typedef union _glamor_pending_op {
unsigned int type;
glamor_pending_fill fill;
} glamor_pending_op;
2009-08-21 00:28:13 +02:00
typedef struct glamor_pixmap_private {
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
unsigned char gl_fbo:1;
unsigned char gl_tex:1;
unsigned char pbo_valid:1;
unsigned char is_picture:1;
GLuint tex;
GLuint fb;
GLuint pbo;
glamor_access_t access_mode;
PictFormatShort pict_format;
glamor_pending_op pending_op;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
PixmapPtr container;
glamor_screen_private *glamor_priv;
2009-08-21 00:28:13 +02:00
} glamor_pixmap_private;
#define GLAMOR_CHECK_PENDING_FILL(_dispatch_, _glamor_priv_, _pixmap_priv_) do \
{ \
if (_pixmap_priv_->pending_op.type == GLAMOR_PENDING_FILL) { \
_dispatch_->glUseProgram(_glamor_priv_->solid_prog); \
_dispatch_->glUniform4fv(_glamor_priv_->solid_color_uniform_location, 1, \
_pixmap_priv_->pending_op.fill.color4fv); \
} \
} while(0)
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
/*
* Pixmap dynamic status, used by dynamic upload feature.
*
* GLAMOR_NONE: initial status, don't need to do anything.
* GLAMOR_UPLOAD_PENDING: marked as need to be uploaded to gl texture.
* GLAMOR_UPLOAD_DONE: the pixmap has been uploaded successfully.
* GLAMOR_UPLOAD_FAILED: fail to upload the pixmap.
*
* */
typedef enum glamor_pixmap_status {
GLAMOR_NONE,
GLAMOR_UPLOAD_PENDING,
GLAMOR_UPLOAD_DONE,
GLAMOR_UPLOAD_FAILED
} glamor_pixmap_status_t;
extern DevPrivateKey glamor_screen_private_key;
2009-08-21 00:28:13 +02:00
extern DevPrivateKey glamor_pixmap_private_key;
static inline glamor_screen_private *
glamor_get_screen_private(ScreenPtr screen)
{
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
return (glamor_screen_private *)dixLookupPrivate(&screen->devPrivates,
glamor_screen_private_key);
}
2009-08-21 00:28:13 +02:00
static inline glamor_pixmap_private *
glamor_get_pixmap_private(PixmapPtr pixmap)
{
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
return dixLookupPrivate(&pixmap->devPrivates, glamor_pixmap_private_key);
2009-08-21 00:28:13 +02:00
}
#define ARRAY_SIZE(x) (sizeof(x) / sizeof(x[0]))
#define ALIGN(i,m) (((i) + (m) - 1) & ~((m) - 1))
#define MIN(a,b) ((a) < (b) ? (a) : (b))
/**
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
* Borrow from uxa.
*/
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
static inline CARD32
format_for_depth(int depth)
{
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
switch (depth) {
case 1: return PICT_a1;
case 4: return PICT_a4;
case 8: return PICT_a8;
case 15: return PICT_x1r5g5b5;
case 16: return PICT_r5g6b5;
default:
case 24: return PICT_x8r8g8b8;
#if XORG_VERSION_CURRENT >= 10699900
case 30: return PICT_x2r10g10b10;
#endif
case 32: return PICT_a8r8g8b8;
}
}
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
static inline CARD32
format_for_pixmap(PixmapPtr pixmap)
{
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
glamor_pixmap_private *pixmap_priv;
PictFormatShort pict_format;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
pixmap_priv = glamor_get_pixmap_private(pixmap);
if (GLAMOR_PIXMAP_PRIV_IS_PICTURE(pixmap_priv))
pict_format = pixmap_priv->pict_format;
else
pict_format = format_for_depth(pixmap->drawable.depth);
return pict_format;
}
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
/*
* Map picture's format to the correct gl texture format and type.
* no_alpha is used to indicate whehter we need to wire alpha to 1.
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
*
* Return 0 if find a matched texture type. Otherwise return -1.
**/
#ifndef GLAMOR_GLES2
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
static inline int
glamor_get_tex_format_type_from_pictformat(PictFormatShort format,
GLenum *tex_format,
GLenum *tex_type,
int *no_alpha,
int *no_revert)
{
*no_alpha = 0;
*no_revert = 1;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
switch (format) {
case PICT_a1:
*tex_format = GL_COLOR_INDEX;
*tex_type = GL_BITMAP;
break;
case PICT_b8g8r8x8:
*no_alpha = 1;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
case PICT_b8g8r8a8:
*tex_format = GL_BGRA;
*tex_type = GL_UNSIGNED_INT_8_8_8_8;
break;
case PICT_x8r8g8b8:
*no_alpha = 1;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
case PICT_a8r8g8b8:
*tex_format = GL_BGRA;
*tex_type = GL_UNSIGNED_INT_8_8_8_8_REV;
break;
case PICT_x8b8g8r8:
*no_alpha = 1;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
case PICT_a8b8g8r8:
*tex_format = GL_RGBA;
*tex_type = GL_UNSIGNED_INT_8_8_8_8_REV;
break;
case PICT_x2r10g10b10:
*no_alpha = 1;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
case PICT_a2r10g10b10:
*tex_format = GL_BGRA;
*tex_type = GL_UNSIGNED_INT_2_10_10_10_REV;
break;
case PICT_x2b10g10r10:
*no_alpha = 1;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
case PICT_a2b10g10r10:
*tex_format = GL_RGBA;
*tex_type = GL_UNSIGNED_INT_2_10_10_10_REV;
break;
case PICT_r5g6b5:
*tex_format = GL_RGB;
*tex_type = GL_UNSIGNED_SHORT_5_6_5;
break;
case PICT_b5g6r5:
*tex_format = GL_RGB;
*tex_type = GL_UNSIGNED_SHORT_5_6_5_REV;
break;
case PICT_x1b5g5r5:
*no_alpha = 1;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
case PICT_a1b5g5r5:
*tex_format = GL_RGBA;
*tex_type = GL_UNSIGNED_SHORT_1_5_5_5_REV;
break;
case PICT_x1r5g5b5:
*no_alpha = 1;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
case PICT_a1r5g5b5:
*tex_format = GL_BGRA;
*tex_type = GL_UNSIGNED_SHORT_1_5_5_5_REV;
break;
case PICT_a8:
*tex_format = GL_ALPHA;
*tex_type = GL_UNSIGNED_BYTE;
break;
case PICT_x4r4g4b4:
*no_alpha = 1;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
case PICT_a4r4g4b4:
*tex_format = GL_BGRA;
*tex_type = GL_UNSIGNED_SHORT_4_4_4_4_REV;
break;
case PICT_x4b4g4r4:
*no_alpha = 1;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
case PICT_a4b4g4r4:
*tex_format = GL_RGBA;
*tex_type = GL_UNSIGNED_SHORT_4_4_4_4_REV;
break;
default:
LogMessageVerb(X_INFO, 0, "fail to get matched format for %x \n", format);
return -1;
}
return 0;
}
#else
#define IS_LITTLE_ENDIAN (IMAGE_BYTE_ORDER == LSBFirst)
static inline int
glamor_get_tex_format_type_from_pictformat(PictFormatShort format,
GLenum *tex_format,
GLenum *tex_type,
int *no_alpha,
int *no_revert)
{
*no_alpha = 0;
*no_revert = IS_LITTLE_ENDIAN;
switch (format) {
case PICT_b8g8r8x8:
*no_alpha = 1;
case PICT_b8g8r8a8:
*tex_format = GL_BGRA;
*tex_type = GL_UNSIGNED_BYTE;
*no_revert = !IS_LITTLE_ENDIAN;
break;
case PICT_x8r8g8b8:
*no_alpha = 1;
case PICT_a8r8g8b8:
*tex_format = GL_BGRA;
*tex_type = GL_UNSIGNED_BYTE;
break;
case PICT_x8b8g8r8:
*no_alpha = 1;
case PICT_a8b8g8r8:
*tex_format = GL_RGBA;
*tex_type = GL_UNSIGNED_BYTE;
break;
case PICT_x2r10g10b10:
*no_alpha = 1;
case PICT_a2r10g10b10:
*tex_format = GL_BGRA;
*tex_type = GL_UNSIGNED_INT_10_10_10_2;
*no_revert = TRUE;
break;
case PICT_x2b10g10r10:
*no_alpha = 1;
case PICT_a2b10g10r10:
*tex_format = GL_RGBA;
*tex_type = GL_UNSIGNED_INT_10_10_10_2;
*no_revert = TRUE;
break;
case PICT_r5g6b5:
*tex_format = GL_RGB;
*tex_type = GL_UNSIGNED_SHORT_5_6_5;
*no_revert = TRUE;
break;
case PICT_b5g6r5:
*tex_format = GL_RGB;
*tex_type = GL_UNSIGNED_SHORT_5_6_5;
*no_revert = FALSE;
break;
case PICT_x1b5g5r5:
*no_alpha = 1;
case PICT_a1b5g5r5:
*tex_format = GL_RGBA;
*tex_type = GL_UNSIGNED_SHORT_1_5_5_5_REV;
*no_revert = TRUE;
break;
case PICT_x1r5g5b5:
*no_alpha = 1;
case PICT_a1r5g5b5:
*tex_format = GL_BGRA;
*tex_type = GL_UNSIGNED_SHORT_1_5_5_5_REV;
*no_revert = TRUE;
break;
case PICT_a8:
*tex_format = GL_ALPHA;
*tex_type = GL_UNSIGNED_BYTE;
*no_revert = TRUE;
break;
case PICT_x4r4g4b4:
*no_alpha = 1;
case PICT_a4r4g4b4:
*tex_format = GL_BGRA;
*tex_type = GL_UNSIGNED_SHORT_4_4_4_4_REV;
*no_revert = TRUE;
break;
case PICT_x4b4g4r4:
*no_alpha = 1;
case PICT_a4b4g4r4:
*tex_format = GL_RGBA;
*tex_type = GL_UNSIGNED_SHORT_4_4_4_4_REV;
*no_revert = TRUE;
break;
default:
LogMessageVerb(X_INFO, 0, "fail to get matched format for %x \n", format);
return -1;
}
return 0;
}
#endif
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
static inline int
glamor_get_tex_format_type_from_pixmap(PixmapPtr pixmap,
GLenum *format,
GLenum *type,
int *no_alpha,
int *no_revert)
{
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
glamor_pixmap_private *pixmap_priv;
PictFormatShort pict_format;
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
pixmap_priv = glamor_get_pixmap_private(pixmap);
if (GLAMOR_PIXMAP_PRIV_IS_PICTURE(pixmap_priv))
pict_format = pixmap_priv->pict_format;
else
pict_format = format_for_depth(pixmap->drawable.depth);
return glamor_get_tex_format_type_from_pictformat(pict_format,
format, type,
no_alpha, no_revert);
}
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
/* borrowed from uxa */
static inline Bool
glamor_get_rgba_from_pixel(CARD32 pixel,
float * red,
float * green,
float * blue,
float * alpha,
CARD32 format)
{
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
int rbits, bbits, gbits, abits;
int rshift, bshift, gshift, ashift;
rbits = PICT_FORMAT_R(format);
gbits = PICT_FORMAT_G(format);
bbits = PICT_FORMAT_B(format);
abits = PICT_FORMAT_A(format);
if (PICT_FORMAT_TYPE(format) == PICT_TYPE_A) {
rshift = gshift = bshift = ashift = 0;
} else if (PICT_FORMAT_TYPE(format) == PICT_TYPE_ARGB) {
bshift = 0;
gshift = bbits;
rshift = gshift + gbits;
ashift = rshift + rbits;
} else if (PICT_FORMAT_TYPE(format) == PICT_TYPE_ABGR) {
rshift = 0;
gshift = rbits;
bshift = gshift + gbits;
ashift = bshift + bbits;
#if XORG_VERSION_CURRENT >= 10699900
} else if (PICT_FORMAT_TYPE(format) == PICT_TYPE_BGRA) {
ashift = 0;
rshift = abits;
if (abits == 0)
rshift = PICT_FORMAT_BPP(format) - (rbits+gbits+bbits);
gshift = rshift + rbits;
bshift = gshift + gbits;
#endif
} else {
return FALSE;
}
#define COLOR_INT_TO_FLOAT(_fc_, _p_, _s_, _bits_) \
*_fc_ = (((_p_) >> (_s_)) & (( 1 << (_bits_)) - 1)) \
/ (float)((1<<(_bits_)) - 1)
if (rbits)
COLOR_INT_TO_FLOAT(red, pixel, rshift, rbits);
else
*red = 0;
if (gbits)
COLOR_INT_TO_FLOAT(green, pixel, gshift, gbits);
else
*green = 0;
if (bbits)
COLOR_INT_TO_FLOAT(blue, pixel, bshift, bbits);
else
*blue = 0;
if (abits)
COLOR_INT_TO_FLOAT(alpha, pixel, ashift, abits);
else
*alpha = 1;
return TRUE;
}
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
/**
* Returns TRUE if the given planemask covers all the significant bits in the
* pixel values for pDrawable.
*/
static inline Bool
glamor_pm_is_solid(DrawablePtr drawable, unsigned long planemask)
{
return (planemask & FbFullMask(drawable->depth)) ==
FbFullMask(drawable->depth);
}
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
extern int glamor_debug_level;
/* glamor.c */
PixmapPtr glamor_get_drawable_pixmap(DrawablePtr drawable);
Bool glamor_close_screen(int idx, ScreenPtr screen);
/* glamor_copyarea.c */
RegionPtr
glamor_copy_area(DrawablePtr src, DrawablePtr dst, GCPtr gc,
int srcx, int srcy, int width, int height, int dstx, int dsty);
void
glamor_copy_n_to_n(DrawablePtr src,
DrawablePtr dst,
GCPtr gc,
BoxPtr box,
int nbox,
int dx,
int dy,
Bool reverse,
Bool upsidedown,
Pixel bitplane,
void *closure);
/* glamor_copywindow.c */
void glamor_copy_window(WindowPtr win, DDXPointRec old_origin,
RegionPtr src_region);
/* glamor_core.c */
Bool glamor_prepare_access(DrawablePtr drawable, glamor_access_t access);
void glamor_finish_access(DrawablePtr drawable);
Bool glamor_prepare_access_window(WindowPtr window);
void glamor_finish_access_window(WindowPtr window);
Bool glamor_prepare_access_gc(GCPtr gc);
void glamor_finish_access_gc(GCPtr gc);
void glamor_init_finish_access_shaders(ScreenPtr screen);
const Bool glamor_get_drawable_location(const DrawablePtr drawable);
void glamor_get_drawable_deltas(DrawablePtr drawable, PixmapPtr pixmap,
int *x, int *y);
Bool glamor_create_gc(GCPtr gc);
Bool glamor_stipple(PixmapPtr pixmap, PixmapPtr stipple,
int x, int y, int width, int height,
unsigned char alu, unsigned long planemask,
unsigned long fg_pixel, unsigned long bg_pixel,
int stipple_x, int stipple_y);
GLint glamor_compile_glsl_prog(glamor_gl_dispatch *dispatch, GLenum type, const char *source);
void glamor_link_glsl_prog(glamor_gl_dispatch *dispatch, GLint prog);
2009-08-20 21:32:53 +02:00
void glamor_get_color_4f_from_pixel(PixmapPtr pixmap, unsigned long fg_pixel,
GLfloat *color);
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
int glamor_set_destination_pixmap(PixmapPtr pixmap);
int glamor_set_destination_pixmap_priv(glamor_pixmap_private *pixmap_priv);
/* nc means no check. caller must ensure this pixmap has valid fbo.
* usually use the GLAMOR_PIXMAP_PRIV_HAS_FBO firstly.
* */
void glamor_set_destination_pixmap_priv_nc(glamor_pixmap_private *pixmap_priv);
PixmapPtr
glamor_es2_pixmap_read_prepare(PixmapPtr source, GLenum *format,
GLenum *type, int no_alpha, int no_revert);
void glamor_set_alu(struct glamor_gl_dispatch * dispatch, unsigned char alu);
Bool glamor_set_planemask(PixmapPtr pixmap, unsigned long planemask);
2009-08-20 21:32:53 +02:00
void glamor_get_transform_uniform_locations(GLint prog,
glamor_transform_uniforms *uniform_locations);
void glamor_set_transform_for_pixmap(PixmapPtr pixmap,
glamor_transform_uniforms *uniform_locations);
Bool glamor_change_window_attributes(WindowPtr pWin, unsigned long mask);
RegionPtr glamor_bitmap_to_region(PixmapPtr pixmap);
Bool glamor_gl_has_extension(char *extension);
int glamor_gl_get_version(void);
#define GLAMOR_GL_VERSION_ENCODE(major, minor) ( \
((major) * 256) \
+ ((minor) * 1))
/* glamor_fill.c */
void glamor_fill(DrawablePtr drawable,
GCPtr gc,
int x,
int y,
int width,
int height);
Bool glamor_solid(PixmapPtr pixmap, int x, int y, int width, int height,
2009-08-20 21:32:53 +02:00
unsigned char alu, unsigned long planemask,
unsigned long fg_pixel);
void glamor_solid_fail_region(PixmapPtr pixmap,
int x, int y, int width, int height);
/* glamor_fillspans.c */
void glamor_fill_spans(DrawablePtr drawable,
GCPtr gc,
int n,
DDXPointPtr points,
int *widths,
int sorted);
2009-08-20 21:32:53 +02:00
void glamor_init_solid_shader(ScreenPtr screen);
2009-08-20 23:43:43 +02:00
/* glamor_getspans.c */
void
glamor_get_spans(DrawablePtr drawable,
int wmax,
DDXPointPtr points,
int *widths,
int nspans,
char *dst_start);
/* glamor_glyphs.c */
void glamor_glyphs_fini(ScreenPtr screen);
void glamor_glyphs(CARD8 op,
PicturePtr pSrc,
PicturePtr pDst,
PictFormatPtr maskFormat,
INT16 xSrc,
INT16 ySrc, int nlist, GlyphListPtr list, GlyphPtr * glyphs);
void
glamor_glyph_unrealize (ScreenPtr screen, GlyphPtr glyph);
/* glamor_setspans.c */
void glamor_set_spans(DrawablePtr drawable, GCPtr gc, char *src,
DDXPointPtr points, int *widths, int n, int sorted);
/* glamor_polyfillrect.c */
void
glamor_poly_fill_rect(DrawablePtr drawable,
GCPtr gc,
int nrect,
xRectangle *prect);
/* glamor_polylines.c */
void
glamor_poly_lines(DrawablePtr drawable, GCPtr gc, int mode, int n,
DDXPointPtr points);
2009-08-23 00:54:24 +02:00
/* glamor_putimage.c */
void
glamor_put_image(DrawablePtr drawable, GCPtr gc, int depth, int x, int y,
int w, int h, int leftPad, int format, char *bits);
void glamor_init_putimage_shaders(ScreenPtr screen);
2009-08-23 00:54:24 +02:00
2009-08-21 01:43:22 +02:00
/* glamor_render.c */
void glamor_composite(CARD8 op,
PicturePtr pSrc,
PicturePtr pMask,
PicturePtr pDst,
INT16 xSrc,
INT16 ySrc,
INT16 xMask,
INT16 yMask,
INT16 xDst,
INT16 yDst,
CARD16 width,
CARD16 height);
void glamor_trapezoids(CARD8 op,
PicturePtr src, PicturePtr dst,
PictFormatPtr mask_format, INT16 x_src, INT16 y_src,
int ntrap, xTrapezoid *traps);
void glamor_init_composite_shaders(ScreenPtr screen);
void glamor_composite_rects(CARD8 op,
PicturePtr src, PicturePtr mask, PicturePtr dst,
int nrect, glamor_composite_rect_t *rects);
2009-08-21 01:43:22 +02:00
/* glamor_tile.c */
Bool glamor_tile(PixmapPtr pixmap, PixmapPtr tile,
int x, int y, int width, int height,
unsigned char alu, unsigned long planemask,
int tile_x, int tile_y);
void glamor_init_tile_shader(ScreenPtr screen);
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
/* glamor_triangles.c */
void
glamor_triangles (CARD8 op,
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
PicturePtr pSrc,
PicturePtr pDst,
PictFormatPtr maskFormat,
INT16 xSrc,
INT16 ySrc,
int ntris,
xTriangle *tris);
/* glamor_pixmap.c */
void
glamor_pixmap_init(ScreenPtr screen);
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
/**
* Download a pixmap's texture to cpu memory. If success,
* One copy of current pixmap's texture will be put into
* the pixmap->devPrivate.ptr. Will use pbo to map to
* the pointer if possible.
* The pixmap must be a gl texture pixmap. gl_fbo and
* gl_tex must be 1. Used by glamor_prepare_access.
*
*/
Bool
glamor_download_pixmap_to_cpu(PixmapPtr pixmap, glamor_access_t access);
/**
* Restore a pixmap's data which is downloaded by
* glamor_download_pixmap_to_cpu to its original
* gl texture. Used by glamor_finish_access.
*
* The pixmap must be
* in texture originally. In other word, the gl_fbo
* must be 1.
**/
void
glamor_restore_pixmap_to_texture(PixmapPtr pixmap);
/**
* Ensure to have a fbo attached to the pixmap.
* If the pixmap already has one fbo then do nothing.
* Otherwise, it will generate a new fbo, and bind
* the pixmap's texture to the fbo.
* The pixmap must has a valid texture before call this
* API, othersie, it will trigger a assert.
*/
void
glamor_pixmap_ensure_fb(PixmapPtr pixmap);
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
/**
* Upload a pixmap to gl texture. Used by dynamic pixmap
* uploading feature. The pixmap must be a software pixmap.
* This function will change current FBO and current shaders.
*/
enum glamor_pixmap_status
glamor_upload_pixmap_to_texture(PixmapPtr pixmap);
/**
* Upload a picture to gl texture. Similar to the
* glamor_upload_pixmap_to_texture. Used in rendering.
**/
enum glamor_pixmap_status
glamor_upload_picture_to_texture(PicturePtr picture);
/**
* Destroy all the resources allocated on the uploading
* phase, includs the tex and fbo.
**/
void
glamor_destroy_upload_pixmap(PixmapPtr pixmap);
void
glamor_validate_pixmap(PixmapPtr pixmap);
glamor : Add dynamic texture uploading feature. Major refactoring. 1. Rewrite the pixmap texture uploading and downloading functions. Add some new functions for both the prepare/finish access and the new performance feature dynamic texture uploading, which could download and upload the current image to/from a private texture/fbo. In the uploading or downloading phase, we need to handle two things: The first is the yInverted option, If it set, then we don't need to flip y. If not set, if it is from a dynamic texture uploading then we don't need to flip either if the current drawing process will flip it latter. If it is from finish_access, then we must flip the y axis. The second thing is the alpha channel hanlding, if the pixmap's format is something like x8a8r8g8, x1r5g5b5 which means it doesn't has alpha channel, but it do has those extra bits. Then we need to wire those bits to 1. 2. Add almost all the required picture format support. This is not as trivial as it looks like. The previous implementation only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, we have to fallback to cpu. The reason why we can't simply add those other color format is because the exists of picture. one drawable pixmap may has one or even more container pictures. The drawable pixmap's depth can't map to a specified color format, for example depth 16 can mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get get the color format just from the depth value. But the pixmap do not has a pict_format element. We have to make a new one in the pixmap private data structure. Reroute the CreatePicture to glamor_create_picture and then store the picture's format to the pixmap's private structure. This is not an ideal solution, as there may be more than one pictures refer to the same pixmap. Then we will have trouble. There is an example in glamor_composite_with_shader. The source and mask often share the same pixmap, but use different picture format. Our current solution is to combine those two different picture formats to one which will not lose any data. Then change the source's format to this new format and then upload the pixmap to texture once. It works. If we fail to find a matched new format then we fallback. There still is a potential problem, if two pictures refer to the same pixmap, and one of them destroy the picture, but the other still remained to be used latter. We don't handle that situation currently. To be fixed. 3. Dynamic texture uploading. This is a performance feature. Although we don't like the client to hold a pixmap data to shared memory and we can't accelerate it. And even worse, we may need to fallback all the required pixmaps to cpu memory and then process them on CPU. This feature is to mitigate this penalty. When the target pixmap has a valid gl fbo attached to it. But the other pixmaps are not. Then it will be more efficient to upload the other pixmaps to GPU and then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. To enable this feature, I experienced a significant performance improvement in the Game "Mines" :). 4. Debug facility. Modify the debug output mechanism. Now add a new macro: glamor_debug_output(_level_, _format_,...) to conditional output some messages according to the environment variable GLAMOR_DEBUG. We have the following levels currently. exports GLAMOR_DEBUG to 3 will enable all the above messages. 5. Changes in pixmap private data structure. Add some for the full color format supports and relate it to the pictures which already described. Also Add the following new elements: gl_fbo - to indicates whether this pixmap is on gpu only. gl_tex - to indicates whether the tex is valid and is containing the pixmap's image originally. As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may also has a valid fbo or tex attached to it. So we will have to use the above new element to check it true type. After this commit, we can pass the rendercheck testing for all the picture formats. And is much much fater than fallback to cpu when doing rendercheck testing. Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
2011-06-21 12:31:11 +02:00
int
glamor_create_picture(PicturePtr picture);
Bool
glamor_prepare_access_picture(PicturePtr picture, glamor_access_t access);
void
glamor_finish_access_picture(PicturePtr picture);
void
glamor_destroy_picture(PicturePtr picture);
enum glamor_pixmap_status
glamor_upload_picture_to_texture(PicturePtr picture);
void
glamor_picture_format_fixup(PicturePtr picture, glamor_pixmap_private *pixmap_priv);
/* Dynamic pixmap upload to texture if needed.
* Sometimes, the target is a gl texture pixmap/picture,
* but the source or mask is in cpu memory. In that case,
* upload the source/mask to gl texture and then avoid
* fallback the whole process to cpu. Most of the time,
* this will increase performance obviously. */
#define GLAMOR_PIXMAP_DYNAMIC_UPLOAD
#define GLAMOR_DELAYED_FILLING
#include"glamor_utils.h"
#endif /* GLAMOR_PRIV_H */