blob: e5b70701c4432b009f2dcf3dfb297a7ed2f04594 [file] [log] [blame]
/**
* \file texobj.c
* Texture object management.
*/
/*
* Mesa 3-D graphics library
*
* Copyright (C) 1999-2007 Brian Paul All Rights Reserved.
*
* 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 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.
*/
#include <stdio.h>
#include "bufferobj.h"
#include "context.h"
#include "enums.h"
#include "fbobject.h"
#include "formats.h"
#include "hash.h"
#include "imports.h"
#include "macros.h"
#include "shaderimage.h"
#include "teximage.h"
#include "texobj.h"
#include "texstate.h"
#include "mtypes.h"
#include "program/prog_instruction.h"
/**********************************************************************/
/** \name Internal functions */
/*@{*/
/**
* This function checks for all valid combinations of Min and Mag filters for
* Float types, when extensions like OES_texture_float and
* OES_texture_float_linear are supported. OES_texture_float mentions support
* for NEAREST, NEAREST_MIPMAP_NEAREST magnification and minification filters.
* Mag filters like LINEAR and min filters like NEAREST_MIPMAP_LINEAR,
* LINEAR_MIPMAP_NEAREST and LINEAR_MIPMAP_LINEAR are only valid in case
* OES_texture_float_linear is supported.
*
* Returns true in case the filter is valid for given Float type else false.
*/
static bool
valid_filter_for_float(const struct gl_context *ctx,
const struct gl_texture_object *obj)
{
switch (obj->Sampler.MagFilter) {
case GL_LINEAR:
if (obj->_IsHalfFloat && !ctx->Extensions.OES_texture_half_float_linear) {
return false;
} else if (obj->_IsFloat && !ctx->Extensions.OES_texture_float_linear) {
return false;
}
case GL_NEAREST:
case GL_NEAREST_MIPMAP_NEAREST:
break;
default:
unreachable("Invalid mag filter");
}
switch (obj->Sampler.MinFilter) {
case GL_LINEAR:
case GL_NEAREST_MIPMAP_LINEAR:
case GL_LINEAR_MIPMAP_NEAREST:
case GL_LINEAR_MIPMAP_LINEAR:
if (obj->_IsHalfFloat && !ctx->Extensions.OES_texture_half_float_linear) {
return false;
} else if (obj->_IsFloat && !ctx->Extensions.OES_texture_float_linear) {
return false;
}
case GL_NEAREST:
case GL_NEAREST_MIPMAP_NEAREST:
break;
default:
unreachable("Invalid min filter");
}
return true;
}
/**
* Return the gl_texture_object for a given ID.
*/
struct gl_texture_object *
_mesa_lookup_texture(struct gl_context *ctx, GLuint id)
{
return (struct gl_texture_object *)
_mesa_HashLookup(ctx->Shared->TexObjects, id);
}
/**
* Wrapper around _mesa_lookup_texture that throws GL_INVALID_OPERATION if id
* is not in the hash table. After calling _mesa_error, it returns NULL.
*/
struct gl_texture_object *
_mesa_lookup_texture_err(struct gl_context *ctx, GLuint id, const char* func)
{
struct gl_texture_object *texObj = NULL;
if (id > 0)
texObj = _mesa_lookup_texture(ctx, id); /* Returns NULL if not found. */
if (!texObj)
_mesa_error(ctx, GL_INVALID_OPERATION, "%s(texture)", func);
return texObj;
}
void
_mesa_begin_texture_lookups(struct gl_context *ctx)
{
_mesa_HashLockMutex(ctx->Shared->TexObjects);
}
void
_mesa_end_texture_lookups(struct gl_context *ctx)
{
_mesa_HashUnlockMutex(ctx->Shared->TexObjects);
}
struct gl_texture_object *
_mesa_lookup_texture_locked(struct gl_context *ctx, GLuint id)
{
return (struct gl_texture_object *)
_mesa_HashLookupLocked(ctx->Shared->TexObjects, id);
}
/**
* Return a pointer to the current texture object for the given target
* on the current texture unit.
* Note: all <target> error checking should have been done by this point.
*/
struct gl_texture_object *
_mesa_get_current_tex_object(struct gl_context *ctx, GLenum target)
{
struct gl_texture_unit *texUnit = _mesa_get_current_tex_unit(ctx);
const GLboolean arrayTex = ctx->Extensions.EXT_texture_array;
switch (target) {
case GL_TEXTURE_1D:
return texUnit->CurrentTex[TEXTURE_1D_INDEX];
case GL_PROXY_TEXTURE_1D:
return ctx->Texture.ProxyTex[TEXTURE_1D_INDEX];
case GL_TEXTURE_2D:
return texUnit->CurrentTex[TEXTURE_2D_INDEX];
case GL_PROXY_TEXTURE_2D:
return ctx->Texture.ProxyTex[TEXTURE_2D_INDEX];
case GL_TEXTURE_3D:
return texUnit->CurrentTex[TEXTURE_3D_INDEX];
case GL_PROXY_TEXTURE_3D:
return ctx->Texture.ProxyTex[TEXTURE_3D_INDEX];
case GL_TEXTURE_CUBE_MAP_POSITIVE_X:
case GL_TEXTURE_CUBE_MAP_NEGATIVE_X:
case GL_TEXTURE_CUBE_MAP_POSITIVE_Y:
case GL_TEXTURE_CUBE_MAP_NEGATIVE_Y:
case GL_TEXTURE_CUBE_MAP_POSITIVE_Z:
case GL_TEXTURE_CUBE_MAP_NEGATIVE_Z:
case GL_TEXTURE_CUBE_MAP:
return ctx->Extensions.ARB_texture_cube_map
? texUnit->CurrentTex[TEXTURE_CUBE_INDEX] : NULL;
case GL_PROXY_TEXTURE_CUBE_MAP:
return ctx->Extensions.ARB_texture_cube_map
? ctx->Texture.ProxyTex[TEXTURE_CUBE_INDEX] : NULL;
case GL_TEXTURE_CUBE_MAP_ARRAY:
return _mesa_has_texture_cube_map_array(ctx)
? texUnit->CurrentTex[TEXTURE_CUBE_ARRAY_INDEX] : NULL;
case GL_PROXY_TEXTURE_CUBE_MAP_ARRAY:
return _mesa_has_texture_cube_map_array(ctx)
? ctx->Texture.ProxyTex[TEXTURE_CUBE_ARRAY_INDEX] : NULL;
case GL_TEXTURE_RECTANGLE_NV:
return ctx->Extensions.NV_texture_rectangle
? texUnit->CurrentTex[TEXTURE_RECT_INDEX] : NULL;
case GL_PROXY_TEXTURE_RECTANGLE_NV:
return ctx->Extensions.NV_texture_rectangle
? ctx->Texture.ProxyTex[TEXTURE_RECT_INDEX] : NULL;
case GL_TEXTURE_1D_ARRAY_EXT:
return arrayTex ? texUnit->CurrentTex[TEXTURE_1D_ARRAY_INDEX] : NULL;
case GL_PROXY_TEXTURE_1D_ARRAY_EXT:
return arrayTex ? ctx->Texture.ProxyTex[TEXTURE_1D_ARRAY_INDEX] : NULL;
case GL_TEXTURE_2D_ARRAY_EXT:
return arrayTex ? texUnit->CurrentTex[TEXTURE_2D_ARRAY_INDEX] : NULL;
case GL_PROXY_TEXTURE_2D_ARRAY_EXT:
return arrayTex ? ctx->Texture.ProxyTex[TEXTURE_2D_ARRAY_INDEX] : NULL;
case GL_TEXTURE_BUFFER:
return (_mesa_has_ARB_texture_buffer_object(ctx) ||
_mesa_has_OES_texture_buffer(ctx)) ?
texUnit->CurrentTex[TEXTURE_BUFFER_INDEX] : NULL;
case GL_TEXTURE_EXTERNAL_OES:
return _mesa_is_gles(ctx) && ctx->Extensions.OES_EGL_image_external
? texUnit->CurrentTex[TEXTURE_EXTERNAL_INDEX] : NULL;
case GL_TEXTURE_2D_MULTISAMPLE:
return ctx->Extensions.ARB_texture_multisample
? texUnit->CurrentTex[TEXTURE_2D_MULTISAMPLE_INDEX] : NULL;
case GL_PROXY_TEXTURE_2D_MULTISAMPLE:
return ctx->Extensions.ARB_texture_multisample
? ctx->Texture.ProxyTex[TEXTURE_2D_MULTISAMPLE_INDEX] : NULL;
case GL_TEXTURE_2D_MULTISAMPLE_ARRAY:
return ctx->Extensions.ARB_texture_multisample
? texUnit->CurrentTex[TEXTURE_2D_MULTISAMPLE_ARRAY_INDEX] : NULL;
case GL_PROXY_TEXTURE_2D_MULTISAMPLE_ARRAY:
return ctx->Extensions.ARB_texture_multisample
? ctx->Texture.ProxyTex[TEXTURE_2D_MULTISAMPLE_ARRAY_INDEX] : NULL;
default:
_mesa_problem(NULL, "bad target in _mesa_get_current_tex_object()");
return NULL;
}
}
/**
* Allocate and initialize a new texture object. But don't put it into the
* texture object hash table.
*
* Called via ctx->Driver.NewTextureObject, unless overridden by a device
* driver.
*
* \param shared the shared GL state structure to contain the texture object
* \param name integer name for the texture object
* \param target either GL_TEXTURE_1D, GL_TEXTURE_2D, GL_TEXTURE_3D,
* GL_TEXTURE_CUBE_MAP or GL_TEXTURE_RECTANGLE_NV. zero is ok for the sake
* of GenTextures()
*
* \return pointer to new texture object.
*/
struct gl_texture_object *
_mesa_new_texture_object( struct gl_context *ctx, GLuint name, GLenum target )
{
struct gl_texture_object *obj;
(void) ctx;
obj = MALLOC_STRUCT(gl_texture_object);
_mesa_initialize_texture_object(ctx, obj, name, target);
return obj;
}
/**
* Initialize a new texture object to default values.
* \param obj the texture object
* \param name the texture name
* \param target the texture target
*/
void
_mesa_initialize_texture_object( struct gl_context *ctx,
struct gl_texture_object *obj,
GLuint name, GLenum target )
{
assert(target == 0 ||
target == GL_TEXTURE_1D ||
target == GL_TEXTURE_2D ||
target == GL_TEXTURE_3D ||
target == GL_TEXTURE_CUBE_MAP ||
target == GL_TEXTURE_RECTANGLE_NV ||
target == GL_TEXTURE_1D_ARRAY_EXT ||
target == GL_TEXTURE_2D_ARRAY_EXT ||
target == GL_TEXTURE_EXTERNAL_OES ||
target == GL_TEXTURE_CUBE_MAP_ARRAY ||
target == GL_TEXTURE_BUFFER ||
target == GL_TEXTURE_2D_MULTISAMPLE ||
target == GL_TEXTURE_2D_MULTISAMPLE_ARRAY);
memset(obj, 0, sizeof(*obj));
/* init the non-zero fields */
mtx_init(&obj->Mutex, mtx_plain);
obj->RefCount = 1;
obj->Name = name;
obj->Target = target;
if (target != 0) {
obj->TargetIndex = _mesa_tex_target_to_index(ctx, target);
}
else {
obj->TargetIndex = NUM_TEXTURE_TARGETS; /* invalid/error value */
}
obj->Priority = 1.0F;
obj->BaseLevel = 0;
obj->MaxLevel = 1000;
/* must be one; no support for (YUV) planes in separate buffers */
obj->RequiredTextureImageUnits = 1;
/* sampler state */
if (target == GL_TEXTURE_RECTANGLE_NV ||
target == GL_TEXTURE_EXTERNAL_OES) {
obj->Sampler.WrapS = GL_CLAMP_TO_EDGE;
obj->Sampler.WrapT = GL_CLAMP_TO_EDGE;
obj->Sampler.WrapR = GL_CLAMP_TO_EDGE;
obj->Sampler.MinFilter = GL_LINEAR;
}
else {
obj->Sampler.WrapS = GL_REPEAT;
obj->Sampler.WrapT = GL_REPEAT;
obj->Sampler.WrapR = GL_REPEAT;
obj->Sampler.MinFilter = GL_NEAREST_MIPMAP_LINEAR;
}
obj->Sampler.MagFilter = GL_LINEAR;
obj->Sampler.MinLod = -1000.0;
obj->Sampler.MaxLod = 1000.0;
obj->Sampler.LodBias = 0.0;
obj->Sampler.MaxAnisotropy = 1.0;
obj->Sampler.CompareMode = GL_NONE; /* ARB_shadow */
obj->Sampler.CompareFunc = GL_LEQUAL; /* ARB_shadow */
obj->DepthMode = ctx->API == API_OPENGL_CORE ? GL_RED : GL_LUMINANCE;
obj->StencilSampling = false;
obj->Sampler.CubeMapSeamless = GL_FALSE;
obj->Swizzle[0] = GL_RED;
obj->Swizzle[1] = GL_GREEN;
obj->Swizzle[2] = GL_BLUE;
obj->Swizzle[3] = GL_ALPHA;
obj->_Swizzle = SWIZZLE_NOOP;
obj->Sampler.sRGBDecode = GL_DECODE_EXT;
obj->BufferObjectFormat = GL_R8;
obj->_BufferObjectFormat = MESA_FORMAT_R_UNORM8;
obj->ImageFormatCompatibilityType = GL_IMAGE_FORMAT_COMPATIBILITY_BY_SIZE;
}
/**
* Some texture initialization can't be finished until we know which
* target it's getting bound to (GL_TEXTURE_1D/2D/etc).
*/
static void
finish_texture_init(struct gl_context *ctx, GLenum target,
struct gl_texture_object *obj)
{
GLenum filter = GL_LINEAR;
assert(obj->Target == 0);
obj->Target = target;
obj->TargetIndex = _mesa_tex_target_to_index(ctx, target);
assert(obj->TargetIndex < NUM_TEXTURE_TARGETS);
switch (target) {
case GL_TEXTURE_2D_MULTISAMPLE:
case GL_TEXTURE_2D_MULTISAMPLE_ARRAY:
filter = GL_NEAREST;
/* fallthrough */
case GL_TEXTURE_RECTANGLE_NV:
case GL_TEXTURE_EXTERNAL_OES:
/* have to init wrap and filter state here - kind of klunky */
obj->Sampler.WrapS = GL_CLAMP_TO_EDGE;
obj->Sampler.WrapT = GL_CLAMP_TO_EDGE;
obj->Sampler.WrapR = GL_CLAMP_TO_EDGE;
obj->Sampler.MinFilter = filter;
obj->Sampler.MagFilter = filter;
if (ctx->Driver.TexParameter) {
/* XXX we probably don't need to make all these calls */
ctx->Driver.TexParameter(ctx, obj, GL_TEXTURE_WRAP_S);
ctx->Driver.TexParameter(ctx, obj, GL_TEXTURE_WRAP_T);
ctx->Driver.TexParameter(ctx, obj, GL_TEXTURE_WRAP_R);
ctx->Driver.TexParameter(ctx, obj, GL_TEXTURE_MIN_FILTER);
ctx->Driver.TexParameter(ctx, obj, GL_TEXTURE_MAG_FILTER);
}
break;
default:
/* nothing needs done */
break;
}
}
/**
* Deallocate a texture object struct. It should have already been
* removed from the texture object pool.
* Called via ctx->Driver.DeleteTexture() if not overriden by a driver.
*
* \param shared the shared GL state to which the object belongs.
* \param texObj the texture object to delete.
*/
void
_mesa_delete_texture_object(struct gl_context *ctx,
struct gl_texture_object *texObj)
{
GLuint i, face;
/* Set Target to an invalid value. With some assertions elsewhere
* we can try to detect possible use of deleted textures.
*/
texObj->Target = 0x99;
/* free the texture images */
for (face = 0; face < 6; face++) {
for (i = 0; i < MAX_TEXTURE_LEVELS; i++) {
if (texObj->Image[face][i]) {
ctx->Driver.DeleteTextureImage(ctx, texObj->Image[face][i]);
}
}
}
_mesa_reference_buffer_object(ctx, &texObj->BufferObject, NULL);
/* destroy the mutex -- it may have allocated memory (eg on bsd) */
mtx_destroy(&texObj->Mutex);
free(texObj->Label);
/* free this object */
free(texObj);
}
/**
* Copy texture object state from one texture object to another.
* Use for glPush/PopAttrib.
*
* \param dest destination texture object.
* \param src source texture object.
*/
void
_mesa_copy_texture_object( struct gl_texture_object *dest,
const struct gl_texture_object *src )
{
dest->Target = src->Target;
dest->TargetIndex = src->TargetIndex;
dest->Name = src->Name;
dest->Priority = src->Priority;
dest->Sampler.BorderColor.f[0] = src->Sampler.BorderColor.f[0];
dest->Sampler.BorderColor.f[1] = src->Sampler.BorderColor.f[1];
dest->Sampler.BorderColor.f[2] = src->Sampler.BorderColor.f[2];
dest->Sampler.BorderColor.f[3] = src->Sampler.BorderColor.f[3];
dest->Sampler.WrapS = src->Sampler.WrapS;
dest->Sampler.WrapT = src->Sampler.WrapT;
dest->Sampler.WrapR = src->Sampler.WrapR;
dest->Sampler.MinFilter = src->Sampler.MinFilter;
dest->Sampler.MagFilter = src->Sampler.MagFilter;
dest->Sampler.MinLod = src->Sampler.MinLod;
dest->Sampler.MaxLod = src->Sampler.MaxLod;
dest->Sampler.LodBias = src->Sampler.LodBias;
dest->BaseLevel = src->BaseLevel;
dest->MaxLevel = src->MaxLevel;
dest->Sampler.MaxAnisotropy = src->Sampler.MaxAnisotropy;
dest->Sampler.CompareMode = src->Sampler.CompareMode;
dest->Sampler.CompareFunc = src->Sampler.CompareFunc;
dest->Sampler.CubeMapSeamless = src->Sampler.CubeMapSeamless;
dest->DepthMode = src->DepthMode;
dest->StencilSampling = src->StencilSampling;
dest->Sampler.sRGBDecode = src->Sampler.sRGBDecode;
dest->_MaxLevel = src->_MaxLevel;
dest->_MaxLambda = src->_MaxLambda;
dest->GenerateMipmap = src->GenerateMipmap;
dest->_BaseComplete = src->_BaseComplete;
dest->_MipmapComplete = src->_MipmapComplete;
COPY_4V(dest->Swizzle, src->Swizzle);
dest->_Swizzle = src->_Swizzle;
dest->_IsHalfFloat = src->_IsHalfFloat;
dest->_IsFloat = src->_IsFloat;
dest->RequiredTextureImageUnits = src->RequiredTextureImageUnits;
}
/**
* Free all texture images of the given texture object.
*
* \param ctx GL context.
* \param t texture object.
*
* \sa _mesa_clear_texture_image().
*/
void
_mesa_clear_texture_object(struct gl_context *ctx,
struct gl_texture_object *texObj)
{
GLuint i, j;
if (texObj->Target == 0)
return;
for (i = 0; i < MAX_FACES; i++) {
for (j = 0; j < MAX_TEXTURE_LEVELS; j++) {
struct gl_texture_image *texImage = texObj->Image[i][j];
if (texImage)
_mesa_clear_texture_image(ctx, texImage);
}
}
}
/**
* Check if the given texture object is valid by examining its Target field.
* For debugging only.
*/
static GLboolean
valid_texture_object(const struct gl_texture_object *tex)
{
switch (tex->Target) {
case 0:
case GL_TEXTURE_1D:
case GL_TEXTURE_2D:
case GL_TEXTURE_3D:
case GL_TEXTURE_CUBE_MAP:
case GL_TEXTURE_RECTANGLE_NV:
case GL_TEXTURE_1D_ARRAY_EXT:
case GL_TEXTURE_2D_ARRAY_EXT:
case GL_TEXTURE_BUFFER:
case GL_TEXTURE_EXTERNAL_OES:
case GL_TEXTURE_CUBE_MAP_ARRAY:
case GL_TEXTURE_2D_MULTISAMPLE:
case GL_TEXTURE_2D_MULTISAMPLE_ARRAY:
return GL_TRUE;
case 0x99:
_mesa_problem(NULL, "invalid reference to a deleted texture object");
return GL_FALSE;
default:
_mesa_problem(NULL, "invalid texture object Target 0x%x, Id = %u",
tex->Target, tex->Name);
return GL_FALSE;
}
}
/**
* Reference (or unreference) a texture object.
* If '*ptr', decrement *ptr's refcount (and delete if it becomes zero).
* If 'tex' is non-null, increment its refcount.
* This is normally only called from the _mesa_reference_texobj() macro
* when there's a real pointer change.
*/
void
_mesa_reference_texobj_(struct gl_texture_object **ptr,
struct gl_texture_object *tex)
{
assert(ptr);
if (*ptr) {
/* Unreference the old texture */
GLboolean deleteFlag = GL_FALSE;
struct gl_texture_object *oldTex = *ptr;
assert(valid_texture_object(oldTex));
(void) valid_texture_object; /* silence warning in release builds */
mtx_lock(&oldTex->Mutex);
assert(oldTex->RefCount > 0);
oldTex->RefCount--;
deleteFlag = (oldTex->RefCount == 0);
mtx_unlock(&oldTex->Mutex);
if (deleteFlag) {
/* Passing in the context drastically changes the driver code for
* framebuffer deletion.
*/
GET_CURRENT_CONTEXT(ctx);
if (ctx)
ctx->Driver.DeleteTexture(ctx, oldTex);
else
_mesa_problem(NULL, "Unable to delete texture, no context");
}
*ptr = NULL;
}
assert(!*ptr);
if (tex) {
/* reference new texture */
assert(valid_texture_object(tex));
mtx_lock(&tex->Mutex);
if (tex->RefCount == 0) {
/* this texture's being deleted (look just above) */
/* Not sure this can every really happen. Warn if it does. */
_mesa_problem(NULL, "referencing deleted texture object");
*ptr = NULL;
}
else {
tex->RefCount++;
*ptr = tex;
}
mtx_unlock(&tex->Mutex);
}
}
enum base_mipmap { BASE, MIPMAP };
/**
* Mark a texture object as incomplete. There are actually three kinds of
* (in)completeness:
* 1. "base incomplete": the base level of the texture is invalid so no
* texturing is possible.
* 2. "mipmap incomplete": a non-base level of the texture is invalid so
* mipmap filtering isn't possible, but non-mipmap filtering is.
* 3. "texture incompleteness": some combination of texture state and
* sampler state renders the texture incomplete.
*
* \param t texture object
* \param bm either BASE or MIPMAP to indicate what's incomplete
* \param fmt... string describing why it's incomplete (for debugging).
*/
static void
incomplete(struct gl_texture_object *t, enum base_mipmap bm,
const char *fmt, ...)
{
if (MESA_DEBUG_FLAGS & DEBUG_INCOMPLETE_TEXTURE) {
va_list args;
char s[100];
va_start(args, fmt);
vsnprintf(s, sizeof(s), fmt, args);
va_end(args);
_mesa_debug(NULL, "Texture Obj %d incomplete because: %s\n", t->Name, s);
}
if (bm == BASE)
t->_BaseComplete = GL_FALSE;
t->_MipmapComplete = GL_FALSE;
}
/**
* Examine a texture object to determine if it is complete.
*
* The gl_texture_object::Complete flag will be set to GL_TRUE or GL_FALSE
* accordingly.
*
* \param ctx GL context.
* \param t texture object.
*
* According to the texture target, verifies that each of the mipmaps is
* present and has the expected size.
*/
void
_mesa_test_texobj_completeness( const struct gl_context *ctx,
struct gl_texture_object *t )
{
const GLint baseLevel = t->BaseLevel;
const struct gl_texture_image *baseImage;
GLint maxLevels = 0;
/* We'll set these to FALSE if tests fail below */
t->_BaseComplete = GL_TRUE;
t->_MipmapComplete = GL_TRUE;
if (t->Target == GL_TEXTURE_BUFFER) {
/* Buffer textures are always considered complete. The obvious case where
* they would be incomplete (no BO attached) is actually specced to be
* undefined rendering results.
*/
return;
}
/* Detect cases where the application set the base level to an invalid
* value.
*/
if ((baseLevel < 0) || (baseLevel >= MAX_TEXTURE_LEVELS)) {
incomplete(t, BASE, "base level = %d is invalid", baseLevel);
return;
}
if (t->MaxLevel < baseLevel) {
incomplete(t, MIPMAP, "MAX_LEVEL (%d) < BASE_LEVEL (%d)",
t->MaxLevel, baseLevel);
return;
}
baseImage = t->Image[0][baseLevel];
/* Always need the base level image */
if (!baseImage) {
incomplete(t, BASE, "Image[baseLevel=%d] == NULL", baseLevel);
return;
}
/* Check width/height/depth for zero */
if (baseImage->Width == 0 ||
baseImage->Height == 0 ||
baseImage->Depth == 0) {
incomplete(t, BASE, "texture width or height or depth = 0");
return;
}
/* Check if the texture values are integer */
{
GLenum datatype = _mesa_get_format_datatype(baseImage->TexFormat);
t->_IsIntegerFormat = datatype == GL_INT || datatype == GL_UNSIGNED_INT;
}
/* Check if the texture type is Float or HalfFloatOES and ensure Min and Mag
* filters are supported in this case.
*/
if (_mesa_is_gles(ctx) && !valid_filter_for_float(ctx, t)) {
incomplete(t, BASE, "Filter is not supported with Float types.");
return;
}
/* Compute _MaxLevel (the maximum mipmap level we'll sample from given the
* mipmap image sizes and GL_TEXTURE_MAX_LEVEL state).
*/
switch (t->Target) {
case GL_TEXTURE_1D:
case GL_TEXTURE_1D_ARRAY_EXT:
maxLevels = ctx->Const.MaxTextureLevels;
break;
case GL_TEXTURE_2D:
case GL_TEXTURE_2D_ARRAY_EXT:
maxLevels = ctx->Const.MaxTextureLevels;
break;
case GL_TEXTURE_3D:
maxLevels = ctx->Const.Max3DTextureLevels;
break;
case GL_TEXTURE_CUBE_MAP:
case GL_TEXTURE_CUBE_MAP_ARRAY:
maxLevels = ctx->Const.MaxCubeTextureLevels;
break;
case GL_TEXTURE_RECTANGLE_NV:
case GL_TEXTURE_BUFFER:
case GL_TEXTURE_EXTERNAL_OES:
case GL_TEXTURE_2D_MULTISAMPLE:
case GL_TEXTURE_2D_MULTISAMPLE_ARRAY:
maxLevels = 1; /* no mipmapping */
break;
default:
_mesa_problem(ctx, "Bad t->Target in _mesa_test_texobj_completeness");
return;
}
assert(maxLevels > 0);
t->_MaxLevel = MIN3(t->MaxLevel,
/* 'p' in the GL spec */
(int) (baseLevel + baseImage->MaxNumLevels - 1),
/* 'q' in the GL spec */
maxLevels - 1);
if (t->Immutable) {
/* Adjust max level for views: the data store may have more levels than
* the view exposes.
*/
t->_MaxLevel = MIN2(t->_MaxLevel, t->NumLevels - 1);
}
/* Compute _MaxLambda = q - p in the spec used during mipmapping */
t->_MaxLambda = (GLfloat) (t->_MaxLevel - baseLevel);
if (t->Immutable) {
/* This texture object was created with glTexStorage1/2/3D() so we
* know that all the mipmap levels are the right size and all cube
* map faces are the same size.
* We don't need to do any of the additional checks below.
*/
return;
}
if (t->Target == GL_TEXTURE_CUBE_MAP) {
/* Make sure that all six cube map level 0 images are the same size and
* format.
* Note: we know that the image's width==height (we enforce that
* at glTexImage time) so we only need to test the width here.
*/
GLuint face;
assert(baseImage->Width2 == baseImage->Height);
for (face = 1; face < 6; face++) {
assert(t->Image[face][baseLevel] == NULL ||
t->Image[face][baseLevel]->Width2 ==
t->Image[face][baseLevel]->Height2);
if (t->Image[face][baseLevel] == NULL ||
t->Image[face][baseLevel]->Width2 != baseImage->Width2) {
incomplete(t, BASE, "Cube face missing or mismatched size");
return;
}
if (t->Image[face][baseLevel]->InternalFormat !=
baseImage->InternalFormat) {
incomplete(t, BASE, "Cube face format mismatch");
return;
}
if (t->Image[face][baseLevel]->Border != baseImage->Border) {
incomplete(t, BASE, "Cube face border size mismatch");
return;
}
}
}
/*
* Do mipmap consistency checking.
* Note: we don't care about the current texture sampler state here.
* To determine texture completeness we'll either look at _BaseComplete
* or _MipmapComplete depending on the current minification filter mode.
*/
{
GLint i;
const GLint minLevel = baseLevel;
const GLint maxLevel = t->_MaxLevel;
const GLuint numFaces = _mesa_num_tex_faces(t->Target);
GLuint width, height, depth, face;
if (minLevel > maxLevel) {
incomplete(t, MIPMAP, "minLevel > maxLevel");
return;
}
/* Get the base image's dimensions */
width = baseImage->Width2;
height = baseImage->Height2;
depth = baseImage->Depth2;
/* Note: this loop will be a no-op for RECT, BUFFER, EXTERNAL,
* MULTISAMPLE and MULTISAMPLE_ARRAY textures
*/
for (i = baseLevel + 1; i < maxLevels; i++) {
/* Compute the expected size of image at level[i] */
if (width > 1) {
width /= 2;
}
if (height > 1 && t->Target != GL_TEXTURE_1D_ARRAY) {
height /= 2;
}
if (depth > 1 && t->Target != GL_TEXTURE_2D_ARRAY
&& t->Target != GL_TEXTURE_CUBE_MAP_ARRAY) {
depth /= 2;
}
/* loop over cube faces (or single face otherwise) */
for (face = 0; face < numFaces; face++) {
if (i >= minLevel && i <= maxLevel) {
const struct gl_texture_image *img = t->Image[face][i];
if (!img) {
incomplete(t, MIPMAP, "TexImage[%d] is missing", i);
return;
}
if (img->InternalFormat != baseImage->InternalFormat) {
incomplete(t, MIPMAP, "Format[i] != Format[baseLevel]");
return;
}
if (img->Border != baseImage->Border) {
incomplete(t, MIPMAP, "Border[i] != Border[baseLevel]");
return;
}
if (img->Width2 != width) {
incomplete(t, MIPMAP, "TexImage[%d] bad width %u", i,
img->Width2);
return;
}
if (img->Height2 != height) {
incomplete(t, MIPMAP, "TexImage[%d] bad height %u", i,
img->Height2);
return;
}
if (img->Depth2 != depth) {
incomplete(t, MIPMAP, "TexImage[%d] bad depth %u", i,
img->Depth2);
return;
}
}
}
if (width == 1 && height == 1 && depth == 1) {
return; /* found smallest needed mipmap, all done! */
}
}
}
}
GLboolean
_mesa_cube_level_complete(const struct gl_texture_object *texObj,
const GLint level)
{
const struct gl_texture_image *img0, *img;
GLuint face;
if (texObj->Target != GL_TEXTURE_CUBE_MAP)
return GL_FALSE;
if ((level < 0) || (level >= MAX_TEXTURE_LEVELS))
return GL_FALSE;
/* check first face */
img0 = texObj->Image[0][level];
if (!img0 ||
img0->Width < 1 ||
img0->Width != img0->Height)
return GL_FALSE;
/* check remaining faces vs. first face */
for (face = 1; face < 6; face++) {
img = texObj->Image[face][level];
if (!img ||
img->Width != img0->Width ||
img->Height != img0->Height ||
img->TexFormat != img0->TexFormat)
return GL_FALSE;
}
return GL_TRUE;
}
/**
* Check if the given cube map texture is "cube complete" as defined in
* the OpenGL specification.
*/
GLboolean
_mesa_cube_complete(const struct gl_texture_object *texObj)
{
return _mesa_cube_level_complete(texObj, texObj->BaseLevel);
}
/**
* Mark a texture object dirty. It forces the object to be incomplete
* and forces the context to re-validate its state.
*
* \param ctx GL context.
* \param texObj texture object.
*/
void
_mesa_dirty_texobj(struct gl_context *ctx, struct gl_texture_object *texObj)
{
texObj->_BaseComplete = GL_FALSE;
texObj->_MipmapComplete = GL_FALSE;
ctx->NewState |= _NEW_TEXTURE;
}
/**
* Return pointer to a default/fallback texture of the given type/target.
* The texture is an RGBA texture with all texels = (0,0,0,1).
* That's the value a GLSL sampler should get when sampling from an
* incomplete texture.
*/
struct gl_texture_object *
_mesa_get_fallback_texture(struct gl_context *ctx, gl_texture_index tex)
{
if (!ctx->Shared->FallbackTex[tex]) {
/* create fallback texture now */
const GLsizei width = 1, height = 1;
GLsizei depth = 1;
GLubyte texel[24];
struct gl_texture_object *texObj;
struct gl_texture_image *texImage;
mesa_format texFormat;
GLuint dims, face, numFaces = 1;
GLenum target;
for (face = 0; face < 6; face++) {
texel[4*face + 0] =
texel[4*face + 1] =
texel[4*face + 2] = 0x0;
texel[4*face + 3] = 0xff;
}
switch (tex) {
case TEXTURE_2D_ARRAY_INDEX:
dims = 3;
target = GL_TEXTURE_2D_ARRAY;
break;
case TEXTURE_1D_ARRAY_INDEX:
dims = 2;
target = GL_TEXTURE_1D_ARRAY;
break;
case TEXTURE_CUBE_INDEX:
dims = 2;
target = GL_TEXTURE_CUBE_MAP;
numFaces = 6;
break;
case TEXTURE_3D_INDEX:
dims = 3;
target = GL_TEXTURE_3D;
break;
case TEXTURE_RECT_INDEX:
dims = 2;
target = GL_TEXTURE_RECTANGLE;
break;
case TEXTURE_2D_INDEX:
dims = 2;
target = GL_TEXTURE_2D;
break;
case TEXTURE_1D_INDEX:
dims = 1;
target = GL_TEXTURE_1D;
break;
case TEXTURE_BUFFER_INDEX:
dims = 0;
target = GL_TEXTURE_BUFFER;
break;
case TEXTURE_CUBE_ARRAY_INDEX:
dims = 3;
target = GL_TEXTURE_CUBE_MAP_ARRAY;
depth = 6;
break;
case TEXTURE_EXTERNAL_INDEX:
dims = 2;
target = GL_TEXTURE_EXTERNAL_OES;
break;
case TEXTURE_2D_MULTISAMPLE_INDEX:
dims = 2;
target = GL_TEXTURE_2D_MULTISAMPLE;
break;
case TEXTURE_2D_MULTISAMPLE_ARRAY_INDEX:
dims = 3;
target = GL_TEXTURE_2D_MULTISAMPLE_ARRAY;
break;
default:
/* no-op */
return NULL;
}
/* create texture object */
texObj = ctx->Driver.NewTextureObject(ctx, 0, target);
if (!texObj)
return NULL;
assert(texObj->RefCount == 1);
texObj->Sampler.MinFilter = GL_NEAREST;
texObj->Sampler.MagFilter = GL_NEAREST;
texFormat = ctx->Driver.ChooseTextureFormat(ctx, target,
GL_RGBA, GL_RGBA,
GL_UNSIGNED_BYTE);
/* need a loop here just for cube maps */
for (face = 0; face < numFaces; face++) {
const GLenum faceTarget = _mesa_cube_face_target(target, face);
/* initialize level[0] texture image */
texImage = _mesa_get_tex_image(ctx, texObj, faceTarget, 0);
_mesa_init_teximage_fields(ctx, texImage,
width,
(dims > 1) ? height : 1,
(dims > 2) ? depth : 1,
0, /* border */
GL_RGBA, texFormat);
ctx->Driver.TexImage(ctx, dims, texImage,
GL_RGBA, GL_UNSIGNED_BYTE, texel,
&ctx->DefaultPacking);
}
_mesa_test_texobj_completeness(ctx, texObj);
assert(texObj->_BaseComplete);
assert(texObj->_MipmapComplete);
ctx->Shared->FallbackTex[tex] = texObj;
}
return ctx->Shared->FallbackTex[tex];
}
/**
* Compute the size of the given texture object, in bytes.
*/
static GLuint
texture_size(const struct gl_texture_object *texObj)
{
const GLuint numFaces = _mesa_num_tex_faces(texObj->Target);
GLuint face, level, size = 0;
for (face = 0; face < numFaces; face++) {
for (level = 0; level < MAX_TEXTURE_LEVELS; level++) {
const struct gl_texture_image *img = texObj->Image[face][level];
if (img) {
GLuint sz = _mesa_format_image_size(img->TexFormat, img->Width,
img->Height, img->Depth);
size += sz;
}
}
}
return size;
}
/**
* Callback called from _mesa_HashWalk()
*/
static void
count_tex_size(GLuint key, void *data, void *userData)
{
const struct gl_texture_object *texObj =
(const struct gl_texture_object *) data;
GLuint *total = (GLuint *) userData;
(void) key;
*total = *total + texture_size(texObj);
}
/**
* Compute total size (in bytes) of all textures for the given context.
* For debugging purposes.
*/
GLuint
_mesa_total_texture_memory(struct gl_context *ctx)
{
GLuint tgt, total = 0;
_mesa_HashWalk(ctx->Shared->TexObjects, count_tex_size, &total);
/* plus, the default texture objects */
for (tgt = 0; tgt < NUM_TEXTURE_TARGETS; tgt++) {
total += texture_size(ctx->Shared->DefaultTex[tgt]);
}
return total;
}
/**
* Return the base format for the given texture object by looking
* at the base texture image.
* \return base format (such as GL_RGBA) or GL_NONE if it can't be determined
*/
GLenum
_mesa_texture_base_format(const struct gl_texture_object *texObj)
{
const struct gl_texture_image *texImage = _mesa_base_tex_image(texObj);
return texImage ? texImage->_BaseFormat : GL_NONE;
}
static struct gl_texture_object *
invalidate_tex_image_error_check(struct gl_context *ctx, GLuint texture,
GLint level, const char *name)
{
/* The GL_ARB_invalidate_subdata spec says:
*
* "If <texture> is zero or is not the name of a texture, the error
* INVALID_VALUE is generated."
*
* This performs the error check in a different order than listed in the
* spec. We have to get the texture object before we can validate the
* other parameters against values in the texture object.
*/
struct gl_texture_object *const t = _mesa_lookup_texture(ctx, texture);
if (texture == 0 || t == NULL) {
_mesa_error(ctx, GL_INVALID_VALUE, "%s(texture)", name);
return NULL;
}
/* The GL_ARB_invalidate_subdata spec says:
*
* "If <level> is less than zero or greater than the base 2 logarithm
* of the maximum texture width, height, or depth, the error
* INVALID_VALUE is generated."
*/
if (level < 0 || level > t->MaxLevel) {
_mesa_error(ctx, GL_INVALID_VALUE, "%s(level)", name);
return NULL;
}
/* The GL_ARB_invalidate_subdata spec says:
*
* "If the target of <texture> is TEXTURE_RECTANGLE, TEXTURE_BUFFER,
* TEXTURE_2D_MULTISAMPLE, or TEXTURE_2D_MULTISAMPLE_ARRAY, and <level>
* is not zero, the error INVALID_VALUE is generated."
*/
if (level != 0) {
switch (t->Target) {
case GL_TEXTURE_RECTANGLE:
case GL_TEXTURE_BUFFER:
case GL_TEXTURE_2D_MULTISAMPLE:
case GL_TEXTURE_2D_MULTISAMPLE_ARRAY:
_mesa_error(ctx, GL_INVALID_VALUE, "%s(level)", name);
return NULL;
default:
break;
}
}
return t;
}
/**
* Helper function for glCreateTextures and glGenTextures. Need this because
* glCreateTextures should throw errors if target = 0. This is not exposed to
* the rest of Mesa to encourage Mesa internals to use nameless textures,
* which do not require expensive hash lookups.
* \param target either 0 or a valid / error-checked texture target enum
*/
static void
create_textures(struct gl_context *ctx, GLenum target,
GLsizei n, GLuint *textures, const char *caller)
{
GLuint first;
GLint i;
if (MESA_VERBOSE & (VERBOSE_API|VERBOSE_TEXTURE))
_mesa_debug(ctx, "%s %d\n", caller, n);
if (n < 0) {
_mesa_error(ctx, GL_INVALID_VALUE, "%s(n < 0)", caller);
return;
}
if (!textures)
return;
/*
* This must be atomic (generation and allocation of texture IDs)
*/
_mesa_HashLockMutex(ctx->Shared->TexObjects);
first = _mesa_HashFindFreeKeyBlock(ctx->Shared->TexObjects, n);
/* Allocate new, empty texture objects */
for (i = 0; i < n; i++) {
struct gl_texture_object *texObj;
GLuint name = first + i;
texObj = ctx->Driver.NewTextureObject(ctx, name, target);
if (!texObj) {
_mesa_HashUnlockMutex(ctx->Shared->TexObjects);
_mesa_error(ctx, GL_OUT_OF_MEMORY, "gl%sTextures", caller);
return;
}
/* insert into hash table */
_mesa_HashInsertLocked(ctx->Shared->TexObjects, texObj->Name, texObj);
textures[i] = name;
}
_mesa_HashUnlockMutex(ctx->Shared->TexObjects);
}
/*@}*/
/***********************************************************************/
/** \name API functions */
/*@{*/
/**
* Generate texture names.
*
* \param n number of texture names to be generated.
* \param textures an array in which will hold the generated texture names.
*
* \sa glGenTextures(), glCreateTextures().
*
* Calls _mesa_HashFindFreeKeyBlock() to find a block of free texture
* IDs which are stored in \p textures. Corresponding empty texture
* objects are also generated.
*/
void GLAPIENTRY
_mesa_GenTextures(GLsizei n, GLuint *textures)
{
GET_CURRENT_CONTEXT(ctx);
create_textures(ctx, 0, n, textures, "glGenTextures");
}
/**
* Create texture objects.
*
* \param target the texture target for each name to be generated.
* \param n number of texture names to be generated.
* \param textures an array in which will hold the generated texture names.
*
* \sa glCreateTextures(), glGenTextures().
*
* Calls _mesa_HashFindFreeKeyBlock() to find a block of free texture
* IDs which are stored in \p textures. Corresponding empty texture
* objects are also generated.
*/
void GLAPIENTRY
_mesa_CreateTextures(GLenum target, GLsizei n, GLuint *textures)
{
GLint targetIndex;
GET_CURRENT_CONTEXT(ctx);
/*
* The 4.5 core profile spec (30.10.2014) doesn't specify what
* glCreateTextures should do with invalid targets, which was probably an
* oversight. This conforms to the spec for glBindTexture.
*/
targetIndex = _mesa_tex_target_to_index(ctx, target);
if (targetIndex < 0) {
_mesa_error(ctx, GL_INVALID_ENUM, "glCreateTextures(target)");
return;
}
create_textures(ctx, target, n, textures, "glCreateTextures");
}
/**
* Check if the given texture object is bound to the current draw or
* read framebuffer. If so, Unbind it.
*/
static void
unbind_texobj_from_fbo(struct gl_context *ctx,
struct gl_texture_object *texObj)
{
bool progress = false;
/* Section 4.4.2 (Attaching Images to Framebuffer Objects), subsection
* "Attaching Texture Images to a Framebuffer," of the OpenGL 3.1 spec
* says:
*
* "If a texture object is deleted while its image is attached to one
* or more attachment points in the currently bound framebuffer, then
* it is as if FramebufferTexture* had been called, with a texture of
* zero, for each attachment point to which this image was attached in
* the currently bound framebuffer. In other words, this texture image
* is first detached from all attachment points in the currently bound
* framebuffer. Note that the texture image is specifically not
* detached from any other framebuffer objects. Detaching the texture
* image from any other framebuffer objects is the responsibility of
* the application."
*/
if (_mesa_is_user_fbo(ctx->DrawBuffer)) {
progress = _mesa_detach_renderbuffer(ctx, ctx->DrawBuffer, texObj);
}
if (_mesa_is_user_fbo(ctx->ReadBuffer)
&& ctx->ReadBuffer != ctx->DrawBuffer) {
progress = _mesa_detach_renderbuffer(ctx, ctx->ReadBuffer, texObj)
|| progress;
}
if (progress)
/* Vertices are already flushed by _mesa_DeleteTextures */
ctx->NewState |= _NEW_BUFFERS;
}
/**
* Check if the given texture object is bound to any texture image units and
* unbind it if so (revert to default textures).
*/
static void
unbind_texobj_from_texunits(struct gl_context *ctx,
struct gl_texture_object *texObj)
{
const gl_texture_index index = texObj->TargetIndex;
GLuint u;
if (texObj->Target == 0) {
/* texture was never bound */
return;
}
assert(index < NUM_TEXTURE_TARGETS);
for (u = 0; u < ctx->Texture.NumCurrentTexUsed; u++) {
struct gl_texture_unit *unit = &ctx->Texture.Unit[u];
if (texObj == unit->CurrentTex[index]) {
/* Bind the default texture for this unit/target */
_mesa_reference_texobj(&unit->CurrentTex[index],
ctx->Shared->DefaultTex[index]);
unit->_BoundTextures &= ~(1 << index);
}
}
}
/**
* Check if the given texture object is bound to any shader image unit
* and unbind it if that's the case.
*/
static void
unbind_texobj_from_image_units(struct gl_context *ctx,
struct gl_texture_object *texObj)
{
GLuint i;
for (i = 0; i < ctx->Const.MaxImageUnits; i++) {
struct gl_image_unit *unit = &ctx->ImageUnits[i];
if (texObj == unit->TexObj) {
_mesa_reference_texobj(&unit->TexObj, NULL);
*unit = _mesa_default_image_unit(ctx);
}
}
}
/**
* Unbinds all textures bound to the given texture image unit.
*/
static void
unbind_textures_from_unit(struct gl_context *ctx, GLuint unit)
{
struct gl_texture_unit *texUnit = &ctx->Texture.Unit[unit];
while (texUnit->_BoundTextures) {
const GLuint index = ffs(texUnit->_BoundTextures) - 1;
struct gl_texture_object *texObj = ctx->Shared->DefaultTex[index];
_mesa_reference_texobj(&texUnit->CurrentTex[index], texObj);
/* Pass BindTexture call to device driver */
if (ctx->Driver.BindTexture)
ctx->Driver.BindTexture(ctx, unit, 0, texObj);
texUnit->_BoundTextures &= ~(1 << index);
ctx->NewState |= _NEW_TEXTURE;
}
}
/**
* Delete named textures.
*
* \param n number of textures to be deleted.
* \param textures array of texture IDs to be deleted.
*
* \sa glDeleteTextures().
*
* If we're about to delete a texture that's currently bound to any
* texture unit, unbind the texture first. Decrement the reference
* count on the texture object and delete it if it's zero.
* Recall that texture objects can be shared among several rendering
* contexts.
*/
void GLAPIENTRY
_mesa_DeleteTextures( GLsizei n, const GLuint *textures)
{
GET_CURRENT_CONTEXT(ctx);
GLint i;
if (MESA_VERBOSE & (VERBOSE_API|VERBOSE_TEXTURE))
_mesa_debug(ctx, "glDeleteTextures %d\n", n);
if (n < 0) {
_mesa_error(ctx, GL_INVALID_VALUE, "glDeleteTextures(n < 0)");
return;
}
FLUSH_VERTICES(ctx, 0); /* too complex */
if (n < 0) {
_mesa_error(ctx, GL_INVALID_VALUE, "glDeleteTextures(n)");
return;
}
if (!textures)
return;
for (i = 0; i < n; i++) {
if (textures[i] > 0) {
struct gl_texture_object *delObj
= _mesa_lookup_texture(ctx, textures[i]);
if (delObj) {
_mesa_lock_texture(ctx, delObj);
/* Check if texture is bound to any framebuffer objects.
* If so, unbind.
* See section 4.4.2.3 of GL_EXT_framebuffer_object.
*/
unbind_texobj_from_fbo(ctx, delObj);
/* Check if this texture is currently bound to any texture units.
* If so, unbind it.
*/
unbind_texobj_from_texunits(ctx, delObj);
/* Check if this texture is currently bound to any shader
* image unit. If so, unbind it.
* See section 3.9.X of GL_ARB_shader_image_load_store.
*/
unbind_texobj_from_image_units(ctx, delObj);
_mesa_unlock_texture(ctx, delObj);
ctx->NewState |= _NEW_TEXTURE;
/* The texture _name_ is now free for re-use.
* Remove it from the hash table now.
*/
_mesa_HashRemove(ctx->Shared->TexObjects, delObj->Name);
/* Unreference the texobj. If refcount hits zero, the texture
* will be deleted.
*/
_mesa_reference_texobj(&delObj, NULL);
}
}
}
}
/**
* This deletes a texObj without altering the hash table.
*/
void
_mesa_delete_nameless_texture(struct gl_context *ctx,
struct gl_texture_object *texObj)
{
if (!texObj)
return;
FLUSH_VERTICES(ctx, 0);
_mesa_lock_texture(ctx, texObj);
{
/* Check if texture is bound to any framebuffer objects.
* If so, unbind.
* See section 4.4.2.3 of GL_EXT_framebuffer_object.
*/
unbind_texobj_from_fbo(ctx, texObj);
/* Check if this texture is currently bound to any texture units.
* If so, unbind it.
*/
unbind_texobj_from_texunits(ctx, texObj);
/* Check if this texture is currently bound to any shader
* image unit. If so, unbind it.
* See section 3.9.X of GL_ARB_shader_image_load_store.
*/
unbind_texobj_from_image_units(ctx, texObj);
}
_mesa_unlock_texture(ctx, texObj);
ctx->NewState |= _NEW_TEXTURE;
/* Unreference the texobj. If refcount hits zero, the texture
* will be deleted.
*/
_mesa_reference_texobj(&texObj, NULL);
}
/**
* Convert a GL texture target enum such as GL_TEXTURE_2D or GL_TEXTURE_3D
* into the corresponding Mesa texture target index.
* Note that proxy targets are not valid here.
* \return TEXTURE_x_INDEX or -1 if target is invalid
*/
int
_mesa_tex_target_to_index(const struct gl_context *ctx, GLenum target)
{
switch (target) {
case GL_TEXTURE_1D:
return _mesa_is_desktop_gl(ctx) ? TEXTURE_1D_INDEX : -1;
case GL_TEXTURE_2D:
return TEXTURE_2D_INDEX;
case GL_TEXTURE_3D:
return ctx->API != API_OPENGLES ? TEXTURE_3D_INDEX : -1;
case GL_TEXTURE_CUBE_MAP:
return ctx->Extensions.ARB_texture_cube_map
? TEXTURE_CUBE_INDEX : -1;
case GL_TEXTURE_RECTANGLE:
return _mesa_is_desktop_gl(ctx) && ctx->Extensions.NV_texture_rectangle
? TEXTURE_RECT_INDEX : -1;
case GL_TEXTURE_1D_ARRAY:
return _mesa_is_desktop_gl(ctx) && ctx->Extensions.EXT_texture_array
? TEXTURE_1D_ARRAY_INDEX : -1;
case GL_TEXTURE_2D_ARRAY:
return (_mesa_is_desktop_gl(ctx) && ctx->Extensions.EXT_texture_array)
|| _mesa_is_gles3(ctx)
? TEXTURE_2D_ARRAY_INDEX : -1;
case GL_TEXTURE_BUFFER:
return (_mesa_has_ARB_texture_buffer_object(ctx) ||
_mesa_has_OES_texture_buffer(ctx)) ?
TEXTURE_BUFFER_INDEX : -1;
case GL_TEXTURE_EXTERNAL_OES:
return _mesa_is_gles(ctx) && ctx->Extensions.OES_EGL_image_external
? TEXTURE_EXTERNAL_INDEX : -1;
case GL_TEXTURE_CUBE_MAP_ARRAY:
return _mesa_has_texture_cube_map_array(ctx)
? TEXTURE_CUBE_ARRAY_INDEX : -1;
case GL_TEXTURE_2D_MULTISAMPLE:
return ((_mesa_is_desktop_gl(ctx) && ctx->Extensions.ARB_texture_multisample) ||
_mesa_is_gles31(ctx)) ? TEXTURE_2D_MULTISAMPLE_INDEX: -1;
case GL_TEXTURE_2D_MULTISAMPLE_ARRAY:
return ((_mesa_is_desktop_gl(ctx) && ctx->Extensions.ARB_texture_multisample) ||
_mesa_is_gles31(ctx))
? TEXTURE_2D_MULTISAMPLE_ARRAY_INDEX: -1;
default:
return -1;
}
}
/**
* Do actual texture binding. All error checking should have been done prior
* to calling this function. Note that the texture target (1D, 2D, etc) is
* always specified by the texObj->TargetIndex.
*
* \param unit index of texture unit to update
* \param texObj the new texture object (cannot be NULL)
*/
static void
bind_texture(struct gl_context *ctx,
unsigned unit,
struct gl_texture_object *texObj)
{
struct gl_texture_unit *texUnit;
int targetIndex;
assert(unit < ARRAY_SIZE(ctx->Texture.Unit));
texUnit = &ctx->Texture.Unit[unit];
assert(texObj);
assert(valid_texture_object(texObj));
targetIndex = texObj->TargetIndex;
assert(targetIndex >= 0);
assert(targetIndex < NUM_TEXTURE_TARGETS);
/* Check if this texture is only used by this context and is already bound.
* If so, just return.
*/
{
bool early_out;
mtx_lock(&ctx->Shared->Mutex);
early_out = ((ctx->Shared->RefCount == 1)
&& (texObj == texUnit->CurrentTex[targetIndex]));
mtx_unlock(&ctx->Shared->Mutex);
if (early_out) {
return;
}
}
/* flush before changing binding */
FLUSH_VERTICES(ctx, _NEW_TEXTURE);
/* If the refcount on the previously bound texture is decremented to
* zero, it'll be deleted here.
*/
_mesa_reference_texobj(&texUnit->CurrentTex[targetIndex], texObj);
ctx->Texture.NumCurrentTexUsed = MAX2(ctx->Texture.NumCurrentTexUsed,
unit + 1);
if (texObj->Name != 0)
texUnit->_BoundTextures |= (1 << targetIndex);
else
texUnit->_BoundTextures &= ~(1 << targetIndex);
/* Pass BindTexture call to device driver */
if (ctx->Driver.BindTexture) {
ctx->Driver.BindTexture(ctx, unit, texObj->Target, texObj);
}
}
/**
* Implement glBindTexture(). Do error checking, look-up or create a new
* texture object, then bind it in the current texture unit.
*
* \param target texture target.
* \param texName texture name.
*/
void GLAPIENTRY
_mesa_BindTexture( GLenum target, GLuint texName )
{
GET_CURRENT_CONTEXT(ctx);
struct gl_texture_object *newTexObj = NULL;
GLint targetIndex;
if (MESA_VERBOSE & (VERBOSE_API|VERBOSE_TEXTURE))
_mesa_debug(ctx, "glBindTexture %s %d\n",
_mesa_enum_to_string(target), (GLint) texName);
targetIndex = _mesa_tex_target_to_index(ctx, target);
if (targetIndex < 0) {
_mesa_error(ctx, GL_INVALID_ENUM, "glBindTexture(target)");
return;
}
assert(targetIndex < NUM_TEXTURE_TARGETS);
/*
* Get pointer to new texture object (newTexObj)
*/
if (texName == 0) {
/* Use a default texture object */
newTexObj = ctx->Shared->DefaultTex[targetIndex];
}
else {
/* non-default texture object */
newTexObj = _mesa_lookup_texture(ctx, texName);
if (newTexObj) {
/* error checking */
if (newTexObj->Target != 0 && newTexObj->Target != target) {
/* The named texture object's target doesn't match the
* given target
*/
_mesa_error( ctx, GL_INVALID_OPERATION,
"glBindTexture(target mismatch)" );
return;
}
if (newTexObj->Target == 0) {
finish_texture_init(ctx, target, newTexObj);
}
}
else {
if (ctx->API == API_OPENGL_CORE) {
_mesa_error(ctx, GL_INVALID_OPERATION,
"glBindTexture(non-gen name)");
return;
}
/* if this is a new texture id, allocate a texture object now */
newTexObj = ctx->Driver.NewTextureObject(ctx, texName, target);
if (!newTexObj) {
_mesa_error(ctx, GL_OUT_OF_MEMORY, "glBindTexture");
return;
}
/* and insert it into hash table */
_mesa_HashInsert(ctx->Shared->TexObjects, texName, newTexObj);
}
}
assert(newTexObj->Target == target);
assert(newTexObj->TargetIndex == targetIndex);
bind_texture(ctx, ctx->Texture.CurrentUnit, newTexObj);
}
/**
* OpenGL 4.5 / GL_ARB_direct_state_access glBindTextureUnit().
*
* \param unit texture unit.
* \param texture texture name.
*
* \sa glBindTexture().
*
* If the named texture is 0, this will reset each target for the specified
* texture unit to its default texture.
* If the named texture is not 0 or a recognized texture name, this throws
* GL_INVALID_OPERATION.
*/
void GLAPIENTRY
_mesa_BindTextureUnit(GLuint unit, GLuint texture)
{
GET_CURRENT_CONTEXT(ctx);
struct gl_texture_object *texObj;
if (unit >= _mesa_max_tex_unit(ctx)) {
_mesa_error(ctx, GL_INVALID_VALUE, "glBindTextureUnit(unit=%u)", unit);
return;
}
if (MESA_VERBOSE & (VERBOSE_API|VERBOSE_TEXTURE))
_mesa_debug(ctx, "glBindTextureUnit %s %d\n",
_mesa_enum_to_string(GL_TEXTURE0+unit), (GLint) texture);
/* Section 8.1 (Texture Objects) of the OpenGL 4.5 core profile spec
* (20141030) says:
* "When texture is zero, each of the targets enumerated at the
* beginning of this section is reset to its default texture for the
* corresponding texture image unit."
*/
if (texture == 0) {
unbind_textures_from_unit(ctx, unit);
return;
}
/* Get the non-default texture object */
texObj = _mesa_lookup_texture(ctx, texture);
/* Error checking */
if (!texObj) {
_mesa_error(ctx, GL_INVALID_OPERATION,
"glBindTextureUnit(non-gen name)");
return;
}
if (texObj->Target == 0) {
/* Texture object was gen'd but never bound so the target is not set */
_mesa_error(ctx, GL_INVALID_OPERATION, "glBindTextureUnit(target)");
return;
}
assert(valid_texture_object(texObj));
bind_texture(ctx, unit, texObj);
}
/**
* OpenGL 4.4 / GL_ARB_multi_bind glBindTextures().
*/
void GLAPIENTRY
_mesa_BindTextures(GLuint first, GLsizei count, const GLuint *textures)
{
GET_CURRENT_CONTEXT(ctx);
GLint i;
/* The ARB_multi_bind spec says:
*
* "An INVALID_OPERATION error is generated if <first> + <count>
* is greater than the number of texture image units supported
* by the implementation."
*/
if (first + count > ctx->Const.MaxCombinedTextureImageUnits) {
_mesa_error(ctx, GL_INVALID_OPERATION,
"glBindTextures(first=%u + count=%d > the value of "
"GL_MAX_COMBINED_TEXTURE_IMAGE_UNITS=%u)",
first, count, ctx->Const.MaxCombinedTextureImageUnits);
return;
}
if (textures) {
/* Note that the error semantics for multi-bind commands differ from
* those of other GL commands.
*
* The issues section in the ARB_multi_bind spec says:
*
* "(11) Typically, OpenGL specifies that if an error is generated by
* a command, that command has no effect. This is somewhat
* unfortunate for multi-bind commands, because it would require
* a first pass to scan the entire list of bound objects for
* errors and then a second pass to actually perform the
* bindings. Should we have different error semantics?
*
* RESOLVED: Yes. In this specification, when the parameters for
* one of the <count> binding points are invalid, that binding
* point is not updated and an error will be generated. However,
* other binding points in the same command will be updated if
* their parameters are valid and no other error occurs."
*/
_mesa_begin_texture_lookups(ctx);
for (i = 0; i < count; i++) {
if (textures[i] != 0) {
struct gl_texture_unit *texUnit = &ctx->Texture.Unit[first + i];
struct gl_texture_object *current = texUnit->_Current;
struct gl_texture_object *texObj;
if (current && current->Name == textures[i])
texObj = current;
else
texObj = _mesa_lookup_texture_locked(ctx, textures[i]);
if (texObj && texObj->Target != 0) {
bind_texture(ctx, first + i, texObj);
} else {
/* The ARB_multi_bind spec says:
*
* "An INVALID_OPERATION error is generated if any value
* in <textures> is not zero or the name of an existing
* texture object (per binding)."
*/
_mesa_error(ctx, GL_INVALID_OPERATION,
"glBindTextures(textures[%d]=%u is not zero "
"or the name of an existing texture object)",
i, textures[i]);
}
} else {
unbind_textures_from_unit(ctx, first + i);
}
}
_mesa_end_texture_lookups(ctx);
} else {
/* Unbind all textures in the range <first> through <first>+<count>-1 */
for (i = 0; i < count; i++)
unbind_textures_from_unit(ctx, first + i);
}
}
/**
* Set texture priorities.
*
* \param n number of textures.
* \param texName texture names.
* \param priorities corresponding texture priorities.
*
* \sa glPrioritizeTextures().
*
* Looks up each texture in the hash, clamps the corresponding priority between
* 0.0 and 1.0, and calls dd_function_table::PrioritizeTexture.
*/
void GLAPIENTRY
_mesa_PrioritizeTextures( GLsizei n, const GLuint *texName,
const GLclampf *priorities )
{
GET_CURRENT_CONTEXT(ctx);
GLint i;
if (MESA_VERBOSE & (VERBOSE_API|VERBOSE_TEXTURE))
_mesa_debug(ctx, "glPrioritizeTextures %d\n", n);
FLUSH_VERTICES(ctx, 0);
if (n < 0) {
_mesa_error( ctx, GL_INVALID_VALUE, "glPrioritizeTextures" );
return;
}
if (!priorities)
return;
for (i = 0; i < n; i++) {
if (texName[i] > 0) {
struct gl_texture_object *t = _mesa_lookup_texture(ctx, texName[i]);
if (t) {
t->Priority = CLAMP( priorities[i], 0.0F, 1.0F );
}
}
}
ctx->NewState |= _NEW_TEXTURE;
}
/**
* See if textures are loaded in texture memory.
*
* \param n number of textures to query.
* \param texName array with the texture names.
* \param residences array which will hold the residence status.
*
* \return GL_TRUE if all textures are resident and
* residences is left unchanged,
*
* Note: we assume all textures are always resident
*/
GLboolean GLAPIENTRY
_mesa_AreTexturesResident(GLsizei n, const GLuint *texName,
GLboolean *residences)
{
GET_CURRENT_CONTEXT(ctx);
GLboolean allResident = GL_TRUE;
GLint i;
ASSERT_OUTSIDE_BEGIN_END_WITH_RETVAL(ctx, GL_FALSE);
if (MESA_VERBOSE & (VERBOSE_API|VERBOSE_TEXTURE))
_mesa_debug(ctx, "glAreTexturesResident %d\n", n);
if (n < 0) {
_mesa_error(ctx, GL_INVALID_VALUE, "glAreTexturesResident(n)");
return GL_FALSE;
}
if (!texName || !residences)
return GL_FALSE;
/* We only do error checking on the texture names */
for (i = 0; i < n; i++) {
struct gl_texture_object *t;
if (texName[i] == 0) {
_mesa_error(ctx, GL_INVALID_VALUE, "glAreTexturesResident");
return GL_FALSE;
}
t = _mesa_lookup_texture(ctx, texName[i]);
if (!t) {
_mesa_error(ctx, GL_INVALID_VALUE, "glAreTexturesResident");
return GL_FALSE;
}
}
return allResident;
}
/**
* See if a name corresponds to a texture.
*
* \param texture texture name.
*
* \return GL_TRUE if texture name corresponds to a texture, or GL_FALSE
* otherwise.
*
* \sa glIsTexture().
*
* Calls _mesa_HashLookup().
*/
GLboolean GLAPIENTRY
_mesa_IsTexture( GLuint texture )
{
struct gl_texture_object *t;
GET_CURRENT_CONTEXT(ctx);
ASSERT_OUTSIDE_BEGIN_END_WITH_RETVAL(ctx, GL_FALSE);
if (MESA_VERBOSE & (VERBOSE_API|VERBOSE_TEXTURE))
_mesa_debug(ctx, "glIsTexture %d\n", texture);
if (!texture)
return GL_FALSE;
t = _mesa_lookup_texture(ctx, texture);
/* IsTexture is true only after object has been bound once. */
return t && t->Target;
}
/**
* Simplest implementation of texture locking: grab the shared tex
* mutex. Examine the shared context state timestamp and if there has
* been a change, set the appropriate bits in ctx->NewState.
*
* This is used to deal with synchronizing things when a texture object
* is used/modified by different contexts (or threads) which are sharing
* the texture.
*
* See also _mesa_lock/unlock_texture() in teximage.h
*/
void
_mesa_lock_context_textures( struct gl_context *ctx )
{
mtx_lock(&ctx->Shared->TexMutex);
if (ctx->Shared->TextureStateStamp != ctx->TextureStateTimestamp) {
ctx->NewState |= _NEW_TEXTURE;
ctx->TextureStateTimestamp = ctx->Shared->TextureStateStamp;
}
}
void
_mesa_unlock_context_textures( struct gl_context *ctx )
{
assert(ctx->Shared->TextureStateStamp == ctx->TextureStateTimestamp);
mtx_unlock(&ctx->Shared->TexMutex);
}
void GLAPIENTRY
_mesa_InvalidateTexSubImage(GLuint texture, GLint level, GLint xoffset,
GLint yoffset, GLint zoffset, GLsizei width,
GLsizei height, GLsizei depth)
{
struct gl_texture_object *t;
struct gl_texture_image *image;
GET_CURRENT_CONTEXT(ctx);
if (MESA_VERBOSE & (VERBOSE_API|VERBOSE_TEXTURE))
_mesa_debug(ctx, "glInvalidateTexSubImage %d\n", texture);
t = invalidate_tex_image_error_check(ctx, texture, level,
"glInvalidateTexSubImage");
/* The GL_ARB_invalidate_subdata spec says:
*
* "...the specified subregion must be between -<b> and <dim>+<b> where
* <dim> is the size of the dimension of the texture image, and <b> is
* the size of the border of that texture image, otherwise
* INVALID_VALUE is generated (border is not applied to dimensions that
* don't exist in a given texture target)."
*/
image = t->Image[0][level];
if (image) {
int xBorder;
int yBorder;
int zBorder;
int imageWidth;
int imageHeight;
int imageDepth;
/* The GL_ARB_invalidate_subdata spec says:
*
* "For texture targets that don't have certain dimensions, this
* command treats those dimensions as having a size of 1. For
* example, to invalidate a portion of a two-dimensional texture,
* the application would use <zoffset> equal to zero and <depth>
* equal to one."
*/
switch (t->Target) {
case GL_TEXTURE_BUFFER:
xBorder = 0;
yBorder = 0;
zBorder = 0;
imageWidth = 1;
imageHeight = 1;
imageDepth = 1;
break;
case GL_TEXTURE_1D:
xBorder = image->Border;
yBorder = 0;
zBorder = 0;
imageWidth = image->Width;
imageHeight = 1;
imageDepth = 1;
break;
case GL_TEXTURE_1D_ARRAY:
xBorder = image->Border;
yBorder = 0;
zBorder = 0;
imageWidth = image->Width;
imageHeight = image->Height;
imageDepth = 1;
break;
case GL_TEXTURE_2D:
case GL_TEXTURE_CUBE_MAP:
case GL_TEXTURE_RECTANGLE:
case GL_TEXTURE_2D_MULTISAMPLE:
xBorder = image->Border;
yBorder = image->Border;
zBorder = 0;
imageWidth = image->Width;
imageHeight = image->Height;
imageDepth = 1;
break;
case GL_TEXTURE_2D_ARRAY:
case GL_TEXTURE_CUBE_MAP_ARRAY:
case GL_TEXTURE_2D_MULTISAMPLE_ARRAY:
xBorder = image->Border;
yBorder = image->Border;
zBorder = 0;
imageWidth = image->Width;
imageHeight = image->Height;
imageDepth = image->Depth;
break;
case GL_TEXTURE_3D:
xBorder = image->Border;
yBorder = image->Border;
zBorder = image->Border;
imageWidth = image->Width;
imageHeight = image->Height;
imageDepth = image->Depth;
break;
default:
assert(!"Should not get here.");
xBorder = 0;
yBorder = 0;
zBorder = 0;
imageWidth = 0;
imageHeight = 0;
imageDepth = 0;
break;
}
if (xoffset < -xBorder) {
_mesa_error(ctx, GL_INVALID_VALUE, "glInvalidateSubTexImage(xoffset)");
return;
}
if (xoffset + width > imageWidth + xBorder) {
_mesa_error(ctx, GL_INVALID_VALUE,
"glInvalidateSubTexImage(xoffset+width)");
return;
}
if (yoffset < -yBorder) {
_mesa_error(ctx, GL_INVALID_VALUE, "glInvalidateSubTexImage(yoffset)");
return;
}
if (yoffset + height > imageHeight + yBorder) {
_mesa_error(ctx, GL_INVALID_VALUE,
"glInvalidateSubTexImage(yoffset+height)");
return;
}
if (zoffset < -zBorder) {
_mesa_error(ctx, GL_INVALID_VALUE,
"glInvalidateSubTexImage(zoffset)");
return;
}
if (zoffset + depth > imageDepth + zBorder) {
_mesa_error(ctx, GL_INVALID_VALUE,
"glInvalidateSubTexImage(zoffset+depth)");
return;
}
}
/* We don't actually do anything for this yet. Just return after
* validating the parameters and generating the required errors.
*/
return;
}
void GLAPIENTRY
_mesa_InvalidateTexImage(GLuint texture, GLint level)
{
GET_CURRENT_CONTEXT(ctx);
if (MESA_VERBOSE & (VERBOSE_API|VERBOSE_TEXTURE))
_mesa_debug(ctx, "glInvalidateTexImage(%d, %d)\n", texture, level);
invalidate_tex_image_error_check(ctx, texture, level,
"glInvalidateTexImage");
/* We don't actually do anything for this yet. Just return after
* validating the parameters and generating the required errors.
*/
return;
}
/*@}*/