[client] egl: only use a single texture but multiple PBOs

This commit is contained in:
Geoffrey McRae 2020-12-31 14:18:38 +11:00
parent 678ba0f484
commit 1e4e582f67

View File

@ -32,18 +32,17 @@ Place, Suite 330, Boston, MA 02111-1307 USA
#include <SDL2/SDL_egl.h>
/* this must be a multiple of 2 */
#define TEXTURE_COUNT 2
#define BUFFER_COUNT 2
struct Tex
struct Buffer
{
GLuint t;
bool hasPBO;
GLuint pbo;
void * map;
GLsync sync;
};
struct TexState
struct BufferState
{
_Atomic(uint8_t) w, u, s, d;
};
@ -66,9 +65,10 @@ struct EGL_Texture
unsigned int fourcc;
size_t pboBufferSize;
struct TexState state;
int textureCount;
struct Tex tex[TEXTURE_COUNT];
struct BufferState state;
int bufferCount;
GLuint tex;
struct Buffer buf[BUFFER_COUNT];
};
bool egl_texture_init(EGL_Texture ** texture, EGLDisplay * display)
@ -92,25 +92,24 @@ void egl_texture_free(EGL_Texture ** texture)
glDeleteSamplers(1, &(*texture)->sampler);
for(int i = 0; i < (*texture)->textureCount; ++i)
for(int i = 0; i < (*texture)->bufferCount; ++i)
{
struct Tex * t = &(*texture)->tex[i];
if (t->hasPBO)
struct Buffer * b = &(*texture)->buf[i];
if (b->hasPBO)
{
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, t->pbo);
if ((*texture)->tex[i].map)
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, b->pbo);
if ((*texture)->buf[i].map)
{
glUnmapBuffer(GL_PIXEL_UNPACK_BUFFER);
(*texture)->tex[i].map = NULL;
(*texture)->buf[i].map = NULL;
}
glDeleteBuffers(1, &t->pbo);
if (t->sync)
glDeleteSync(t->sync);
glDeleteBuffers(1, &b->pbo);
if (b->sync)
glDeleteSync(b->sync);
}
glDeleteTextures(1, &t->t);
}
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0);
glDeleteTextures(1, &(*texture)->tex);
free(*texture);
*texture = NULL;
@ -118,18 +117,19 @@ void egl_texture_free(EGL_Texture ** texture)
static bool egl_texture_map(EGL_Texture * texture, uint8_t i)
{
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, texture->tex[i].pbo);
texture->tex[i].map = glMapBufferRange(
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, texture->buf[i].pbo);
texture->buf[i].map = glMapBufferRange(
GL_PIXEL_UNPACK_BUFFER,
0,
texture->pboBufferSize,
GL_MAP_WRITE_BIT |
GL_MAP_UNSYNCHRONIZED_BIT |
GL_MAP_INVALIDATE_BUFFER_BIT
GL_MAP_INVALIDATE_BUFFER_BIT |
GL_MAP_PERSISTENT_BIT
);
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0);
if (!texture->tex[i].map)
if (!texture->buf[i].map)
{
EGL_ERROR("glMapBufferRange failed for %d of %lu bytes", i, texture->pboBufferSize);
return false;
@ -140,37 +140,37 @@ static bool egl_texture_map(EGL_Texture * texture, uint8_t i)
static void egl_texture_unmap(EGL_Texture * texture, uint8_t i)
{
if (!texture->tex[i].map)
if (!texture->buf[i].map)
return;
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, texture->tex[i].pbo);
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, texture->buf[i].pbo);
glUnmapBuffer(GL_PIXEL_UNPACK_BUFFER);
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0);
texture->tex[i].map = NULL;
texture->buf[i].map = NULL;
}
bool egl_texture_setup(EGL_Texture * texture, enum EGL_PixelFormat pixFmt, size_t width, size_t height, size_t stride, bool streaming, bool useDMA)
{
if (texture->streaming && !useDMA)
{
for(int i = 0; i < texture->textureCount; ++i)
for(int i = 0; i < texture->bufferCount; ++i)
{
egl_texture_unmap(texture, i);
if (texture->tex[i].hasPBO)
if (texture->buf[i].hasPBO)
{
glDeleteBuffers(1, &texture->tex[i].pbo);
texture->tex[i].hasPBO = false;
glDeleteBuffers(1, &texture->buf[i].pbo);
texture->buf[i].hasPBO = false;
}
}
}
texture->pixFmt = pixFmt;
texture->width = width;
texture->height = height;
texture->stride = stride;
texture->streaming = streaming;
texture->textureCount = streaming ? TEXTURE_COUNT : 1;
texture->ready = false;
texture->pixFmt = pixFmt;
texture->width = width;
texture->height = height;
texture->stride = stride;
texture->streaming = streaming;
texture->bufferCount = streaming ? BUFFER_COUNT : 1;
texture->ready = false;
atomic_store_explicit(&texture->state.w, 0, memory_order_relaxed);
atomic_store_explicit(&texture->state.u, 0, memory_order_relaxed);
@ -222,12 +222,9 @@ bool egl_texture_setup(EGL_Texture * texture, enum EGL_PixelFormat pixFmt, size_
texture->pitch = stride / texture->bpp;
for(int i = 0; i < texture->textureCount; ++i)
{
if (texture->tex[0].t)
glDeleteTextures(1, &texture->tex[i].t);
glGenTextures(1, &texture->tex[i].t);
}
if (texture->tex)
glDeleteTextures(1, &texture->tex);
glGenTextures(1, &texture->tex);
if (!texture->sampler)
{
@ -241,29 +238,30 @@ bool egl_texture_setup(EGL_Texture * texture, enum EGL_PixelFormat pixFmt, size_
if (useDMA)
return true;
for(int i = 0; i < texture->textureCount; ++i)
{
glBindTexture(GL_TEXTURE_2D, texture->tex[i].t);
glTexImage2D(GL_TEXTURE_2D, 0, texture->intFormat, texture->width,
texture->height, 0, texture->format, texture->dataType, NULL);
}
glBindTexture(GL_TEXTURE_2D, texture->tex);
glTexImage2D(GL_TEXTURE_2D, 0, texture->intFormat, texture->width,
texture->height, 0, texture->format, texture->dataType, NULL);
glBindTexture(GL_TEXTURE_2D, 0);
if (!streaming)
return true;
for(int i = 0; i < texture->textureCount; ++i)
for(int i = 0; i < texture->bufferCount; ++i)
{
glGenBuffers(1, &texture->tex[i].pbo);
texture->tex[i].hasPBO = true;
glGenBuffers(1, &texture->buf[i].pbo);
texture->buf[i].hasPBO = true;
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, texture->tex[i].pbo);
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, texture->buf[i].pbo);
glBufferStorage(
GL_PIXEL_UNPACK_BUFFER,
texture->pboBufferSize,
NULL,
GL_MAP_WRITE_BIT
GL_MAP_WRITE_BIT |
GL_MAP_PERSISTENT_BIT
);
if (!egl_texture_map(texture, i))
return false;
}
return true;
@ -295,17 +293,13 @@ bool egl_texture_update(EGL_Texture * texture, const uint8_t * buffer)
return true;
}
const uint8_t t = sw % TEXTURE_COUNT;
if (!egl_texture_map(texture, t))
return EGL_TEX_STATUS_ERROR;
memcpy(texture->tex[t].map, buffer, texture->pboBufferSize);
const uint8_t b = sw % BUFFER_COUNT;
memcpy(texture->buf[b].map, buffer, texture->pboBufferSize);
atomic_fetch_add_explicit(&texture->state.w, 1, memory_order_release);
egl_texture_unmap(texture, t);
}
else
{
glBindTexture(GL_TEXTURE_2D, texture->tex[0].t);
glBindTexture(GL_TEXTURE_2D, texture->tex);
glPixelStorei(GL_UNPACK_ROW_LENGTH, texture->pitch);
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, texture->width, texture->height,
texture->format, texture->dataType, buffer);
@ -328,13 +322,11 @@ bool egl_texture_update_from_frame(EGL_Texture * texture, const FrameBuffer * fr
return true;
}
const uint8_t t = sw % TEXTURE_COUNT;
if (!egl_texture_map(texture, t))
return EGL_TEX_STATUS_ERROR;
const uint8_t b = sw % BUFFER_COUNT;
framebuffer_read(
frame,
texture->tex[t].map,
texture->buf[b].map,
texture->stride,
texture->height,
texture->width,
@ -343,7 +335,6 @@ bool egl_texture_update_from_frame(EGL_Texture * texture, const FrameBuffer * fr
);
atomic_fetch_add_explicit(&texture->state.w, 1, memory_order_release);
egl_texture_unmap(texture, t);
return true;
}
@ -362,7 +353,6 @@ bool egl_texture_update_from_dma(EGL_Texture * texture, const FrameBuffer * fram
return true;
}
const uint8_t t = sw % TEXTURE_COUNT;
EGLAttrib const attribs[] =
{
EGL_WIDTH , texture->width,
@ -390,7 +380,7 @@ bool egl_texture_update_from_dma(EGL_Texture * texture, const FrameBuffer * fram
}
/* bind the texture and initiate the transfer */
glBindTexture(GL_TEXTURE_2D, texture->tex[t].t);
glBindTexture(GL_TEXTURE_2D, texture->tex);
glEGLImageTargetTexture2DOES(GL_TEXTURE_2D, image);
/* wait for completion */
@ -418,20 +408,20 @@ enum EGL_TexStatus egl_texture_process(EGL_Texture * texture)
nextu == atomic_load_explicit(&texture->state.d, memory_order_acquire))
return texture->ready ? EGL_TEX_STATUS_OK : EGL_TEX_STATUS_NOTREADY;
const uint8_t t = su % TEXTURE_COUNT;
const uint8_t b = su % BUFFER_COUNT;
/* update the texture */
if (!texture->dma)
{
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, texture->tex[t].pbo);
glBindTexture(GL_TEXTURE_2D, texture->tex[t].t);
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, texture->buf[b].pbo);
glBindTexture(GL_TEXTURE_2D, texture->tex);
glPixelStorei(GL_UNPACK_ROW_LENGTH, texture->pitch);
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, texture->width, texture->height,
texture->format, texture->dataType, (const void *)0);
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, 0);
/* create a fence to prevent usage before the update is complete */
texture->tex[t].sync = glFenceSync(GL_SYNC_GPU_COMMANDS_COMPLETE, 0);
texture->buf[b].sync = glFenceSync(GL_SYNC_GPU_COMMANDS_COMPLETE, 0);
/* we must flush to ensure the sync is in the command buffer */
glFlush();
@ -453,20 +443,20 @@ enum EGL_TexStatus egl_texture_bind(EGL_Texture * texture)
if (!texture->ready)
return EGL_TEX_STATUS_NOTREADY;
const uint8_t t = ss % TEXTURE_COUNT;
const uint8_t b = ss % BUFFER_COUNT;
if (texture->dma)
{
ss = atomic_fetch_add_explicit(&texture->state.s, 1,
memory_order_release) + 1;
}
else if (texture->tex[t].sync != 0)
else if (texture->buf[b].sync != 0)
{
switch(glClientWaitSync(texture->tex[t].sync, 0, 20000000)) // 20ms
switch(glClientWaitSync(texture->buf[b].sync, 0, 20000000)) // 20ms
{
case GL_ALREADY_SIGNALED:
case GL_CONDITION_SATISFIED:
glDeleteSync(texture->tex[t].sync);
texture->tex[t].sync = 0;
glDeleteSync(texture->buf[b].sync);
texture->buf[b].sync = 0;
ss = atomic_fetch_add_explicit(&texture->state.s, 1,
memory_order_release) + 1;
@ -477,8 +467,8 @@ enum EGL_TexStatus egl_texture_bind(EGL_Texture * texture)
case GL_WAIT_FAILED:
case GL_INVALID_VALUE:
glDeleteSync(texture->tex[t].sync);
texture->tex[t].sync = 0;
glDeleteSync(texture->buf[b].sync);
texture->buf[b].sync = 0;
EGL_ERROR("glClientWaitSync failed");
return EGL_TEX_STATUS_ERROR;
}
@ -489,9 +479,8 @@ enum EGL_TexStatus egl_texture_bind(EGL_Texture * texture)
memory_order_release) + 1;
}
const uint8_t t = sd % TEXTURE_COUNT;
glActiveTexture(GL_TEXTURE0);
glBindTexture(GL_TEXTURE_2D, texture->tex[t].t);
glBindTexture(GL_TEXTURE_2D, texture->tex);
glBindSampler(0, texture->sampler);
return EGL_TEX_STATUS_OK;