summaryrefslogtreecommitdiff
path: root/src/mesa/main/texobj.c
diff options
context:
space:
mode:
Diffstat (limited to 'src/mesa/main/texobj.c')
-rw-r--r--src/mesa/main/texobj.c1052
1 files changed, 1052 insertions, 0 deletions
diff --git a/src/mesa/main/texobj.c b/src/mesa/main/texobj.c
new file mode 100644
index 00000000000..cfe9229cdb1
--- /dev/null
+++ b/src/mesa/main/texobj.c
@@ -0,0 +1,1052 @@
+/**
+ * \file texobj.c
+ * Texture object management.
+ */
+
+/*
+ * Mesa 3-D graphics library
+ * Version: 6.5
+ *
+ * Copyright (C) 1999-2006 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
+ * BRIAN PAUL 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 "glheader.h"
+#include "colortab.h"
+#include "context.h"
+#include "enums.h"
+#include "fbobject.h"
+#include "hash.h"
+#include "imports.h"
+#include "macros.h"
+#include "teximage.h"
+#include "texstate.h"
+#include "texobj.h"
+#include "mtypes.h"
+
+
+#ifdef __VMS
+#define _mesa_sprintf sprintf
+#endif
+
+/**********************************************************************/
+/** \name Internal functions */
+/*@{*/
+
+
+/**
+ * Return the gl_texture_object for a given ID.
+ */
+struct gl_texture_object *
+_mesa_lookup_texture(GLcontext *ctx, GLuint id)
+{
+ return (struct gl_texture_object *)
+ _mesa_HashLookup(ctx->Shared->TexObjects, id);
+}
+
+
+
+/**
+ * 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_ARB 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( GLcontext *ctx, GLuint name, GLenum target )
+{
+ struct gl_texture_object *obj;
+ (void) ctx;
+ obj = MALLOC_STRUCT(gl_texture_object);
+ _mesa_initialize_texture_object(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_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_ARB ||
+ target == GL_TEXTURE_RECTANGLE_NV);
+
+ _mesa_bzero(obj, sizeof(*obj));
+ /* init the non-zero fields */
+ _glthread_INIT_MUTEX(obj->Mutex);
+ obj->RefCount = 1;
+ obj->Name = name;
+ obj->Target = target;
+ obj->Priority = 1.0F;
+ if (target == GL_TEXTURE_RECTANGLE_NV) {
+ obj->WrapS = GL_CLAMP_TO_EDGE;
+ obj->WrapT = GL_CLAMP_TO_EDGE;
+ obj->WrapR = GL_CLAMP_TO_EDGE;
+ obj->MinFilter = GL_LINEAR;
+ }
+ else {
+ obj->WrapS = GL_REPEAT;
+ obj->WrapT = GL_REPEAT;
+ obj->WrapR = GL_REPEAT;
+ obj->MinFilter = GL_NEAREST_MIPMAP_LINEAR;
+ }
+ obj->MagFilter = GL_LINEAR;
+ obj->MinLod = -1000.0;
+ obj->MaxLod = 1000.0;
+ obj->LodBias = 0.0;
+ obj->BaseLevel = 0;
+ obj->MaxLevel = 1000;
+ obj->MaxAnisotropy = 1.0;
+ obj->CompareFlag = GL_FALSE; /* SGIX_shadow */
+ obj->CompareOperator = GL_TEXTURE_LEQUAL_R_SGIX; /* SGIX_shadow */
+ obj->CompareMode = GL_NONE; /* ARB_shadow */
+ obj->CompareFunc = GL_LEQUAL; /* ARB_shadow */
+ obj->DepthMode = GL_LUMINANCE; /* ARB_depth_texture */
+ obj->ShadowAmbient = 0.0F; /* ARB/SGIX_shadow_ambient */
+ _mesa_init_colortable(&obj->Palette);
+}
+
+
+/**
+ * Deallocate a texture object struct. It should have already been
+ * removed from the texture object pool.
+ *
+ * \param shared the shared GL state to which the object belongs.
+ * \param texOjb the texture object to delete.
+ */
+void
+_mesa_delete_texture_object( GLcontext *ctx, struct gl_texture_object *texObj )
+{
+ GLuint i, face;
+
+ (void) ctx;
+
+ _mesa_free_colortable_data(&texObj->Palette);
+
+ /* free the texture images */
+ for (face = 0; face < 6; face++) {
+ for (i = 0; i < MAX_TEXTURE_LEVELS; i++) {
+ if (texObj->Image[face][i]) {
+ _mesa_delete_texture_image( ctx, texObj->Image[face][i] );
+ }
+ }
+ }
+
+ /* destroy the mutex -- it may have allocated memory (eg on bsd) */
+ _glthread_DESTROY_MUTEX(texObj->Mutex);
+
+ /* free this object */
+ _mesa_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->Name = src->Name;
+ dest->Priority = src->Priority;
+ dest->BorderColor[0] = src->BorderColor[0];
+ dest->BorderColor[1] = src->BorderColor[1];
+ dest->BorderColor[2] = src->BorderColor[2];
+ dest->BorderColor[3] = src->BorderColor[3];
+ dest->WrapS = src->WrapS;
+ dest->WrapT = src->WrapT;
+ dest->WrapR = src->WrapR;
+ dest->MinFilter = src->MinFilter;
+ dest->MagFilter = src->MagFilter;
+ dest->MinLod = src->MinLod;
+ dest->MaxLod = src->MaxLod;
+ dest->LodBias = src->LodBias;
+ dest->BaseLevel = src->BaseLevel;
+ dest->MaxLevel = src->MaxLevel;
+ dest->MaxAnisotropy = src->MaxAnisotropy;
+ dest->CompareFlag = src->CompareFlag;
+ dest->CompareOperator = src->CompareOperator;
+ dest->ShadowAmbient = src->ShadowAmbient;
+ dest->CompareMode = src->CompareMode;
+ dest->CompareFunc = src->CompareFunc;
+ dest->DepthMode = src->DepthMode;
+ dest->_MaxLevel = src->_MaxLevel;
+ dest->_MaxLambda = src->_MaxLambda;
+ dest->GenerateMipmap = src->GenerateMipmap;
+ dest->Palette = src->Palette;
+ dest->Complete = src->Complete;
+}
+
+
+/**
+ * Report why a texture object is incomplete.
+ *
+ * \param t texture object.
+ * \param why string describing why it's incomplete.
+ *
+ * \note For debug purposes only.
+ */
+#if 0
+static void
+incomplete(const struct gl_texture_object *t, const char *why)
+{
+ _mesa_printf("Texture Obj %d incomplete because: %s\n", t->Name, why);
+}
+#else
+#define incomplete(t, why)
+#endif
+
+
+/**
+ * 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 GLcontext *ctx,
+ struct gl_texture_object *t )
+{
+ const GLint baseLevel = t->BaseLevel;
+ GLint maxLog2 = 0, maxLevels = 0;
+
+ t->Complete = GL_TRUE; /* be optimistic */
+
+ /* Always need the base level image */
+ if (!t->Image[0][baseLevel]) {
+ char s[100];
+ _mesa_sprintf(s, "obj %p (%d) Image[baseLevel=%d] == NULL",
+ (void *) t, t->Name, baseLevel);
+ incomplete(t, s);
+ t->Complete = GL_FALSE;
+ return;
+ }
+
+ /* Check width/height/depth for zero */
+ if (t->Image[0][baseLevel]->Width == 0 ||
+ t->Image[0][baseLevel]->Height == 0 ||
+ t->Image[0][baseLevel]->Depth == 0) {
+ incomplete(t, "texture width = 0");
+ t->Complete = GL_FALSE;
+ return;
+ }
+
+ /* Compute _MaxLevel */
+ if (t->Target == GL_TEXTURE_1D) {
+ maxLog2 = t->Image[0][baseLevel]->WidthLog2;
+ maxLevels = ctx->Const.MaxTextureLevels;
+ }
+ else if (t->Target == GL_TEXTURE_2D) {
+ maxLog2 = MAX2(t->Image[0][baseLevel]->WidthLog2,
+ t->Image[0][baseLevel]->HeightLog2);
+ maxLevels = ctx->Const.MaxTextureLevels;
+ }
+ else if (t->Target == GL_TEXTURE_3D) {
+ GLint max = MAX2(t->Image[0][baseLevel]->WidthLog2,
+ t->Image[0][baseLevel]->HeightLog2);
+ maxLog2 = MAX2(max, (GLint)(t->Image[0][baseLevel]->DepthLog2));
+ maxLevels = ctx->Const.Max3DTextureLevels;
+ }
+ else if (t->Target == GL_TEXTURE_CUBE_MAP_ARB) {
+ maxLog2 = MAX2(t->Image[0][baseLevel]->WidthLog2,
+ t->Image[0][baseLevel]->HeightLog2);
+ maxLevels = ctx->Const.MaxCubeTextureLevels;
+ }
+ else if (t->Target == GL_TEXTURE_RECTANGLE_NV) {
+ maxLog2 = 0; /* not applicable */
+ maxLevels = 1; /* no mipmapping */
+ }
+ else {
+ _mesa_problem(ctx, "Bad t->Target in _mesa_test_texobj_completeness");
+ return;
+ }
+
+ ASSERT(maxLevels > 0);
+
+ t->_MaxLevel = baseLevel + maxLog2;
+ t->_MaxLevel = MIN2(t->_MaxLevel, t->MaxLevel);
+ t->_MaxLevel = MIN2(t->_MaxLevel, maxLevels - 1);
+
+ /* Compute _MaxLambda = q - b (see the 1.2 spec) used during mipmapping */
+ t->_MaxLambda = (GLfloat) (t->_MaxLevel - t->BaseLevel);
+
+ if (t->Target == GL_TEXTURE_CUBE_MAP_ARB) {
+ /* make sure that all six cube map level 0 images are the same size */
+ const GLuint w = t->Image[0][baseLevel]->Width2;
+ const GLuint h = t->Image[0][baseLevel]->Height2;
+ GLuint face;
+ for (face = 1; face < 6; face++) {
+ if (t->Image[face][baseLevel] == NULL ||
+ t->Image[face][baseLevel]->Width2 != w ||
+ t->Image[face][baseLevel]->Height2 != h) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "Non-quare cubemap image");
+ return;
+ }
+ }
+ }
+
+ /* extra checking for mipmaps */
+ if (t->MinFilter != GL_NEAREST && t->MinFilter != GL_LINEAR) {
+ /*
+ * Mipmapping: determine if we have a complete set of mipmaps
+ */
+ GLint i;
+ GLint minLevel = baseLevel;
+ GLint maxLevel = t->_MaxLevel;
+
+ if (minLevel > maxLevel) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "minLevel > maxLevel");
+ return;
+ }
+
+ /* Test dimension-independent attributes */
+ for (i = minLevel; i <= maxLevel; i++) {
+ if (t->Image[0][i]) {
+ if (t->Image[0][i]->TexFormat != t->Image[0][baseLevel]->TexFormat) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "Format[i] != Format[baseLevel]");
+ return;
+ }
+ if (t->Image[0][i]->Border != t->Image[0][baseLevel]->Border) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "Border[i] != Border[baseLevel]");
+ return;
+ }
+ }
+ }
+
+ /* Test things which depend on number of texture image dimensions */
+ if (t->Target == GL_TEXTURE_1D) {
+ /* Test 1-D mipmaps */
+ GLuint width = t->Image[0][baseLevel]->Width2;
+ for (i = baseLevel + 1; i < maxLevels; i++) {
+ if (width > 1) {
+ width /= 2;
+ }
+ if (i >= minLevel && i <= maxLevel) {
+ if (!t->Image[0][i]) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "1D Image[0][i] == NULL");
+ return;
+ }
+ if (t->Image[0][i]->Width2 != width ) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "1D Image[0][i] bad width");
+ return;
+ }
+ }
+ if (width == 1) {
+ return; /* found smallest needed mipmap, all done! */
+ }
+ }
+ }
+ else if (t->Target == GL_TEXTURE_2D) {
+ /* Test 2-D mipmaps */
+ GLuint width = t->Image[0][baseLevel]->Width2;
+ GLuint height = t->Image[0][baseLevel]->Height2;
+ for (i = baseLevel + 1; i < maxLevels; i++) {
+ if (width > 1) {
+ width /= 2;
+ }
+ if (height > 1) {
+ height /= 2;
+ }
+ if (i >= minLevel && i <= maxLevel) {
+ if (!t->Image[0][i]) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "2D Image[0][i] == NULL");
+ return;
+ }
+ if (t->Image[0][i]->Width2 != width) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "2D Image[0][i] bad width");
+ return;
+ }
+ if (t->Image[0][i]->Height2 != height) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "2D Image[0][i] bad height");
+ return;
+ }
+ if (width==1 && height==1) {
+ return; /* found smallest needed mipmap, all done! */
+ }
+ }
+ }
+ }
+ else if (t->Target == GL_TEXTURE_3D) {
+ /* Test 3-D mipmaps */
+ GLuint width = t->Image[0][baseLevel]->Width2;
+ GLuint height = t->Image[0][baseLevel]->Height2;
+ GLuint depth = t->Image[0][baseLevel]->Depth2;
+ for (i = baseLevel + 1; i < maxLevels; i++) {
+ if (width > 1) {
+ width /= 2;
+ }
+ if (height > 1) {
+ height /= 2;
+ }
+ if (depth > 1) {
+ depth /= 2;
+ }
+ if (i >= minLevel && i <= maxLevel) {
+ if (!t->Image[0][i]) {
+ incomplete(t, "3D Image[0][i] == NULL");
+ t->Complete = GL_FALSE;
+ return;
+ }
+ if (t->Image[0][i]->_BaseFormat == GL_DEPTH_COMPONENT) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "GL_DEPTH_COMPONENT only works with 1/2D tex");
+ return;
+ }
+ if (t->Image[0][i]->Width2 != width) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "3D Image[0][i] bad width");
+ return;
+ }
+ if (t->Image[0][i]->Height2 != height) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "3D Image[0][i] bad height");
+ return;
+ }
+ if (t->Image[0][i]->Depth2 != depth) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "3D Image[0][i] bad depth");
+ return;
+ }
+ }
+ if (width == 1 && height == 1 && depth == 1) {
+ return; /* found smallest needed mipmap, all done! */
+ }
+ }
+ }
+ else if (t->Target == GL_TEXTURE_CUBE_MAP_ARB) {
+ /* make sure 6 cube faces are consistant */
+ GLuint width = t->Image[0][baseLevel]->Width2;
+ GLuint height = t->Image[0][baseLevel]->Height2;
+ for (i = baseLevel + 1; i < maxLevels; i++) {
+ if (width > 1) {
+ width /= 2;
+ }
+ if (height > 1) {
+ height /= 2;
+ }
+ if (i >= minLevel && i <= maxLevel) {
+ GLuint face;
+ for (face = 0; face < 6; face++) {
+ /* check that we have images defined */
+ if (!t->Image[face][i]) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "CubeMap Image[n][i] == NULL");
+ return;
+ }
+ /* Don't support GL_DEPTH_COMPONENT for cube maps */
+ if (t->Image[face][i]->_BaseFormat == GL_DEPTH_COMPONENT) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "GL_DEPTH_COMPONENT only works with 1/2D tex");
+ return;
+ }
+ /* check that all six images have same size */
+ if (t->Image[face][i]->Width2!=width ||
+ t->Image[face][i]->Height2!=height) {
+ t->Complete = GL_FALSE;
+ incomplete(t, "CubeMap Image[n][i] bad size");
+ return;
+ }
+ }
+ }
+ if (width == 1 && height == 1) {
+ return; /* found smallest needed mipmap, all done! */
+ }
+ }
+ }
+ else if (t->Target == GL_TEXTURE_RECTANGLE_NV) {
+ /* XXX special checking? */
+ }
+ else {
+ /* Target = ??? */
+ _mesa_problem(ctx, "Bug in gl_test_texture_object_completeness\n");
+ }
+ }
+}
+
+/*@}*/
+
+
+/***********************************************************************/
+/** \name API functions */
+/*@{*/
+
+/**
+ * Texture name generation lock.
+ *
+ * Used by _mesa_GenTextures() to guarantee that the generation and allocation
+ * of texture IDs is atomic.
+ */
+_glthread_DECLARE_STATIC_MUTEX(GenTexturesLock);
+
+/**
+ * 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().
+ *
+ * While holding the GenTexturesLock lock, 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);
+ GLuint first;
+ GLint i;
+ ASSERT_OUTSIDE_BEGIN_END(ctx);
+
+ if (n < 0) {
+ _mesa_error( ctx, GL_INVALID_VALUE, "glGenTextures" );
+ return;
+ }
+
+ if (!textures)
+ return;
+
+ /*
+ * This must be atomic (generation and allocation of texture IDs)
+ */
+ _glthread_LOCK_MUTEX(GenTexturesLock);
+
+ 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;
+ GLenum target = 0;
+ texObj = (*ctx->Driver.NewTextureObject)( ctx, name, target);
+ if (!texObj) {
+ _glthread_UNLOCK_MUTEX(GenTexturesLock);
+ _mesa_error(ctx, GL_OUT_OF_MEMORY, "glGenTextures");
+ return;
+ }
+
+ /* insert into hash table */
+ _glthread_LOCK_MUTEX(ctx->Shared->Mutex);
+ _mesa_HashInsert(ctx->Shared->TexObjects, texObj->Name, texObj);
+ _glthread_UNLOCK_MUTEX(ctx->Shared->Mutex);
+
+ textures[i] = name;
+ }
+
+ _glthread_UNLOCK_MUTEX(GenTexturesLock);
+}
+
+
+/**
+ * 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(GLcontext *ctx, struct gl_texture_object *texObj)
+{
+ const GLuint n = (ctx->DrawBuffer == ctx->ReadBuffer) ? 1 : 2;
+ GLuint i;
+
+ for (i = 0; i < n; i++) {
+ struct gl_framebuffer *fb = (i == 0) ? ctx->DrawBuffer : ctx->ReadBuffer;
+ if (fb->Name) {
+ GLuint j;
+ for (j = 0; j < BUFFER_COUNT; j++) {
+ if (fb->Attachment[j].Type == GL_TEXTURE &&
+ fb->Attachment[j].Texture == texObj) {
+ _mesa_remove_attachment(ctx, fb->Attachment + j);
+ }
+ }
+ }
+ }
+}
+
+
+/**
+ * Check if the given texture object is bound to any texture image units and
+ * unbind it if so.
+ * XXX all RefCount accesses should be protected by a mutex.
+ */
+static void
+unbind_texobj_from_texunits(GLcontext *ctx, struct gl_texture_object *texObj)
+{
+ GLuint u;
+
+ for (u = 0; u < MAX_TEXTURE_IMAGE_UNITS; u++) {
+ struct gl_texture_unit *unit = &ctx->Texture.Unit[u];
+ if (texObj == unit->Current1D) {
+ unit->Current1D = ctx->Shared->Default1D;
+ ctx->Shared->Default1D->RefCount++;
+ texObj->RefCount--;
+ if (texObj == unit->_Current)
+ unit->_Current = unit->Current1D;
+ }
+ else if (texObj == unit->Current2D) {
+ unit->Current2D = ctx->Shared->Default2D;
+ ctx->Shared->Default2D->RefCount++;
+ texObj->RefCount--;
+ if (texObj == unit->_Current)
+ unit->_Current = unit->Current2D;
+ }
+ else if (texObj == unit->Current3D) {
+ unit->Current3D = ctx->Shared->Default3D;
+ ctx->Shared->Default3D->RefCount++;
+ texObj->RefCount--;
+ if (texObj == unit->_Current)
+ unit->_Current = unit->Current3D;
+ }
+ else if (texObj == unit->CurrentCubeMap) {
+ unit->CurrentCubeMap = ctx->Shared->DefaultCubeMap;
+ ctx->Shared->DefaultCubeMap->RefCount++;
+ texObj->RefCount--;
+ if (texObj == unit->_Current)
+ unit->_Current = unit->CurrentCubeMap;
+ }
+ else if (texObj == unit->CurrentRect) {
+ unit->CurrentRect = ctx->Shared->DefaultRect;
+ ctx->Shared->DefaultRect->RefCount++;
+ texObj->RefCount--;
+ if (texObj == unit->_Current)
+ unit->_Current = unit->CurrentRect;
+ }
+ }
+}
+
+
+/**
+ * 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;
+ ASSERT_OUTSIDE_BEGIN_END_AND_FLUSH(ctx); /* too complex */
+
+ 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) {
+
+ /* 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 and decrement the reference count.
+ */
+ unbind_texobj_from_texunits(ctx, delObj);
+
+ ctx->NewState |= _NEW_TEXTURE;
+
+ /* The texture _name_ is now free for re-use.
+ * Remove it from the hash table now.
+ */
+ _glthread_LOCK_MUTEX(ctx->Shared->Mutex);
+ _mesa_HashRemove(ctx->Shared->TexObjects, delObj->Name);
+ _glthread_UNLOCK_MUTEX(ctx->Shared->Mutex);
+
+ /* The actual texture object will not be freed until it's no
+ * longer bound in any context.
+ * XXX all RefCount accesses should be protected by a mutex.
+ */
+ delObj->RefCount--;
+ if (delObj->RefCount == 0) {
+ ASSERT(delObj->Name != 0); /* Never delete default tex objs */
+ ASSERT(ctx->Driver.DeleteTexture);
+ (*ctx->Driver.DeleteTexture)(ctx, delObj);
+ }
+ }
+ }
+ }
+}
+
+
+/**
+ * Bind a named texture to a texturing target.
+ *
+ * \param target texture target.
+ * \param texName texture name.
+ *
+ * \sa glBindTexture().
+ *
+ * Determines the old texture object bound and returns immediately if rebinding
+ * the same texture. Get the current texture which is either a default texture
+ * if name is null, a named texture from the hash, or a new texture if the
+ * given texture name is new. Increments its reference count, binds it, and
+ * calls dd_function_table::BindTexture. Decrements the old texture reference
+ * count and deletes it if it reaches zero.
+ */
+void GLAPIENTRY
+_mesa_BindTexture( GLenum target, GLuint texName )
+{
+ GET_CURRENT_CONTEXT(ctx);
+ GLuint unit = ctx->Texture.CurrentUnit;
+ struct gl_texture_unit *texUnit = &ctx->Texture.Unit[unit];
+ struct gl_texture_object *oldTexObj;
+ struct gl_texture_object *newTexObj = NULL;
+ ASSERT_OUTSIDE_BEGIN_END(ctx);
+
+ if (MESA_VERBOSE & (VERBOSE_API|VERBOSE_TEXTURE))
+ _mesa_debug(ctx, "glBindTexture %s %d\n",
+ _mesa_lookup_enum_by_nr(target), (GLint) texName);
+
+ /*
+ * Get pointer to currently bound texture object (oldTexObj)
+ */
+ switch (target) {
+ case GL_TEXTURE_1D:
+ oldTexObj = texUnit->Current1D;
+ break;
+ case GL_TEXTURE_2D:
+ oldTexObj = texUnit->Current2D;
+ break;
+ case GL_TEXTURE_3D:
+ oldTexObj = texUnit->Current3D;
+ break;
+ case GL_TEXTURE_CUBE_MAP_ARB:
+ if (!ctx->Extensions.ARB_texture_cube_map) {
+ _mesa_error( ctx, GL_INVALID_ENUM, "glBindTexture(target)" );
+ return;
+ }
+ oldTexObj = texUnit->CurrentCubeMap;
+ break;
+ case GL_TEXTURE_RECTANGLE_NV:
+ if (!ctx->Extensions.NV_texture_rectangle) {
+ _mesa_error( ctx, GL_INVALID_ENUM, "glBindTexture(target)" );
+ return;
+ }
+ oldTexObj = texUnit->CurrentRect;
+ break;
+ default:
+ _mesa_error( ctx, GL_INVALID_ENUM, "glBindTexture(target)" );
+ return;
+ }
+
+ if (oldTexObj->Name == texName)
+ /* XXX this might be wrong. If the texobj is in use by another
+ * context and a texobj parameter was changed, this might be our
+ * only chance to update this context's hardware state.
+ */
+ return; /* rebinding the same texture- no change */
+
+ /*
+ * Get pointer to new texture object (newTexObj)
+ */
+ if (texName == 0) {
+ /* newTexObj = a default texture object */
+ switch (target) {
+ case GL_TEXTURE_1D:
+ newTexObj = ctx->Shared->Default1D;
+ break;
+ case GL_TEXTURE_2D:
+ newTexObj = ctx->Shared->Default2D;
+ break;
+ case GL_TEXTURE_3D:
+ newTexObj = ctx->Shared->Default3D;
+ break;
+ case GL_TEXTURE_CUBE_MAP_ARB:
+ newTexObj = ctx->Shared->DefaultCubeMap;
+ break;
+ case GL_TEXTURE_RECTANGLE_NV:
+ newTexObj = ctx->Shared->DefaultRect;
+ break;
+ default:
+ ; /* Bad targets are caught above */
+ }
+ }
+ 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 dimensions don't match the target */
+ _mesa_error( ctx, GL_INVALID_OPERATION,
+ "glBindTexture(wrong dimensionality)" );
+ return;
+ }
+ if (newTexObj->Target == 0 && target == GL_TEXTURE_RECTANGLE_NV) {
+ /* have to init wrap and filter state here - kind of klunky */
+ newTexObj->WrapS = GL_CLAMP_TO_EDGE;
+ newTexObj->WrapT = GL_CLAMP_TO_EDGE;
+ newTexObj->WrapR = GL_CLAMP_TO_EDGE;
+ newTexObj->MinFilter = GL_LINEAR;
+ if (ctx->Driver.TexParameter) {
+ static const GLfloat fparam_wrap[1] = {(GLfloat) GL_CLAMP_TO_EDGE};
+ static const GLfloat fparam_filter[1] = {(GLfloat) GL_LINEAR};
+ (*ctx->Driver.TexParameter)( ctx, target, newTexObj, GL_TEXTURE_WRAP_S, fparam_wrap );
+ (*ctx->Driver.TexParameter)( ctx, target, newTexObj, GL_TEXTURE_WRAP_T, fparam_wrap );
+ (*ctx->Driver.TexParameter)( ctx, target, newTexObj, GL_TEXTURE_WRAP_R, fparam_wrap );
+ (*ctx->Driver.TexParameter)( ctx, target, newTexObj, GL_TEXTURE_MIN_FILTER, fparam_filter );
+ }
+ }
+ }
+ else {
+ /* 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 */
+ _glthread_LOCK_MUTEX(ctx->Shared->Mutex);
+ _mesa_HashInsert(ctx->Shared->TexObjects, texName, newTexObj);
+ _glthread_UNLOCK_MUTEX(ctx->Shared->Mutex);
+ }
+ newTexObj->Target = target;
+ }
+
+ /* XXX all RefCount accesses should be protected by a mutex. */
+ newTexObj->RefCount++;
+
+ /* do the actual binding, but first flush outstanding vertices:
+ */
+ FLUSH_VERTICES(ctx, _NEW_TEXTURE);
+
+ switch (target) {
+ case GL_TEXTURE_1D:
+ texUnit->Current1D = newTexObj;
+ break;
+ case GL_TEXTURE_2D:
+ texUnit->Current2D = newTexObj;
+ break;
+ case GL_TEXTURE_3D:
+ texUnit->Current3D = newTexObj;
+ break;
+ case GL_TEXTURE_CUBE_MAP_ARB:
+ texUnit->CurrentCubeMap = newTexObj;
+ break;
+ case GL_TEXTURE_RECTANGLE_NV:
+ texUnit->CurrentRect = newTexObj;
+ break;
+ default:
+ _mesa_problem(ctx, "bad target in BindTexture");
+ return;
+ }
+
+ /* Pass BindTexture call to device driver */
+ if (ctx->Driver.BindTexture)
+ (*ctx->Driver.BindTexture)( ctx, target, newTexObj );
+
+ /* Decrement the reference count on the old texture and check if it's
+ * time to delete it.
+ */
+ /* XXX all RefCount accesses should be protected by a mutex. */
+ oldTexObj->RefCount--;
+ ASSERT(oldTexObj->RefCount >= 0);
+ if (oldTexObj->RefCount == 0) {
+ ASSERT(oldTexObj->Name != 0);
+ ASSERT(ctx->Driver.DeleteTexture);
+ (*ctx->Driver.DeleteTexture)( ctx, oldTexObj );
+ }
+}
+
+
+/**
+ * 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;
+ ASSERT_OUTSIDE_BEGIN_END_AND_FLUSH(ctx);
+
+ 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 );
+ if (ctx->Driver.PrioritizeTexture)
+ ctx->Driver.PrioritizeTexture( ctx, t, t->Priority );
+ }
+ }
+ }
+
+ 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 \p residences is left unchanged,
+ *
+ * \sa glAreTexturesResident().
+ *
+ * Looks up each texture in the hash and calls
+ * dd_function_table::IsTextureResident.
+ */
+GLboolean GLAPIENTRY
+_mesa_AreTexturesResident(GLsizei n, const GLuint *texName,
+ GLboolean *residences)
+{
+ GET_CURRENT_CONTEXT(ctx);
+ GLboolean allResident = GL_TRUE;
+ GLint i, j;
+ ASSERT_OUTSIDE_BEGIN_END_WITH_RETVAL(ctx, GL_FALSE);
+
+ if (n < 0) {
+ _mesa_error(ctx, GL_INVALID_VALUE, "glAreTexturesResident(n)");
+ return GL_FALSE;
+ }
+
+ if (!texName || !residences)
+ return GL_FALSE;
+
+ 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;
+ }
+ if (!ctx->Driver.IsTextureResident ||
+ ctx->Driver.IsTextureResident(ctx, t)) {
+ /* The texture is resident */
+ if (!allResident)
+ residences[i] = GL_TRUE;
+ }
+ else {
+ /* The texture is not resident */
+ if (allResident) {
+ allResident = GL_FALSE;
+ for (j = 0; j < i; j++)
+ residences[j] = GL_TRUE;
+ }
+ residences[i] = 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 (!texture)
+ return GL_FALSE;
+
+ t = _mesa_lookup_texture(ctx, texture);
+
+ /* IsTexture is true only after object has been bound once. */
+ return t && t->Target;
+}
+
+/*@}*/