/*
- Copyright 2003 Tungsten Graphics, Inc., Cedar Park, Texas.
+ Copyright 2003 VMware, Inc.
Copyright (C) Intel Corp. 2006. All Rights Reserved.
- Intel funded Tungsten Graphics (http://www.tungstengraphics.com) to
+ Intel funded Tungsten Graphics to
develop this 3D driver.
-
+
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
distribute, sublicense, and/or sell copies of the Software, and to
permit persons to whom the Software is furnished to do so, subject to
the following conditions:
-
+
The above copyright notice and this permission notice (including the
next paragraph) shall be included in all copies or substantial
portions of the Software.
-
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
-
+
**********************************************************************/
/*
* Authors:
- * Keith Whitwell <keith@tungstengraphics.com>
+ * Keith Whitwell <keithw@vmware.com>
*/
#include "main/api_exec.h"
#include "main/context.h"
#include "main/fbobject.h"
+#include "main/extensions.h"
#include "main/imports.h"
#include "main/macros.h"
#include "main/points.h"
#include "main/version.h"
#include "main/vtxfmt.h"
+#include "main/texobj.h"
#include "vbo/vbo_context.h"
#include "intel_fbo.h"
#include "intel_mipmap_tree.h"
#include "intel_pixel.h"
-#include "intel_regions.h"
+#include "intel_image.h"
#include "intel_tex.h"
#include "intel_tex_obj.h"
#include "swrast_setup/swrast_setup.h"
#include "tnl/tnl.h"
#include "tnl/t_pipeline.h"
-#include "glsl/ralloc.h"
+#include "util/ralloc.h"
+
+#include "glsl/nir/nir.h"
/***************************************
* Mesa's Driver Functions
(void) target;
switch (brw->gen) {
+ case 9:
+ case 8:
+ samples[0] = 8;
+ samples[1] = 4;
+ samples[2] = 2;
+ return 3;
+
case 7:
samples[0] = 8;
samples[1] = 4;
return 1;
default:
+ assert(brw->gen < 6);
samples[0] = 1;
return 1;
}
}
static const GLubyte *
-intelGetString(struct gl_context * ctx, GLenum name)
+intel_get_string(struct gl_context * ctx, GLenum name)
{
const struct brw_context *const brw = brw_context(ctx);
}
static void
-intel_viewport(struct gl_context *ctx, GLint x, GLint y, GLsizei w, GLsizei h)
+intel_viewport(struct gl_context *ctx)
{
struct brw_context *brw = brw_context(ctx);
__DRIcontext *driContext = brw->driContext;
- (void) x;
- (void) y;
- (void) w;
- (void) h;
-
if (_mesa_is_winsys_fbo(ctx->DrawBuffer)) {
dri2InvalidateDrawable(driContext->driDrawablePriv);
dri2InvalidateDrawable(driContext->driReadablePriv);
}
static void
-intelInvalidateState(struct gl_context * ctx, GLuint new_state)
+intel_update_state(struct gl_context * ctx, GLuint new_state)
{
struct brw_context *brw = brw_context(ctx);
+ struct intel_texture_object *tex_obj;
+ struct intel_renderbuffer *depth_irb;
if (ctx->swrast_context)
_swrast_InvalidateState(ctx, new_state);
_vbo_InvalidateState(ctx, new_state);
brw->NewGLState |= new_state;
+
+ _mesa_unlock_context_textures(ctx);
+
+ /* Resolve the depth buffer's HiZ buffer. */
+ depth_irb = intel_get_renderbuffer(ctx->DrawBuffer, BUFFER_DEPTH);
+ if (depth_irb)
+ intel_renderbuffer_resolve_hiz(brw, depth_irb);
+
+ /* Resolve depth buffer and render cache of each enabled texture. */
+ int maxEnabledUnit = ctx->Texture._MaxEnabledTexImageUnit;
+ for (int i = 0; i <= maxEnabledUnit; i++) {
+ if (!ctx->Texture.Unit[i]._Current)
+ continue;
+ tex_obj = intel_texture_object(ctx->Texture.Unit[i]._Current);
+ if (!tex_obj || !tex_obj->mt)
+ continue;
+ intel_miptree_all_slices_resolve_depth(brw, tex_obj->mt);
+ intel_miptree_resolve_color(brw, tex_obj->mt);
+ brw_render_cache_set_check_flush(brw, tex_obj->mt->bo);
+ }
+
+ _mesa_lock_context_textures(ctx);
}
#define flushFront(screen) ((screen)->image.loader ? (screen)->image.loader->flushFrontBuffer : (screen)->dri2.loader->flushFrontBuffer)
intel_batchbuffer_flush(brw);
intel_flush_front(ctx);
- if (brw->is_front_buffer_rendering)
- brw->need_throttle = true;
+
+ brw->need_flush_throttle = true;
}
-void
-intelFinish(struct gl_context * ctx)
+static void
+intel_finish(struct gl_context * ctx)
{
struct brw_context *brw = brw_context(ctx);
functions->Viewport = intel_viewport;
functions->Flush = intel_glFlush;
- functions->Finish = intelFinish;
- functions->GetString = intelGetString;
- functions->UpdateState = intelInvalidateState;
+ functions->Finish = intel_finish;
+ functions->GetString = intel_get_string;
+ functions->UpdateState = intel_update_state;
intelInitTextureFuncs(functions);
intelInitTextureImageFuncs(functions);
intelInitTextureSubImageFuncs(functions);
intelInitTextureCopyImageFuncs(functions);
+ intelInitCopyImageFuncs(functions);
intelInitClearFuncs(functions);
intelInitBufferFuncs(functions);
intelInitPixelFuncs(functions);
functions->GetSamplePosition = gen6_get_sample_position;
}
-/**
- * Return array of MSAA modes supported by the hardware. The array is
- * terminated by -1 and sorted in decreasing order.
- */
-static const int*
-brw_supported_msaa_modes(const struct brw_context *brw)
-{
- static const int gen7_samples[] = {8, 4, 0, -1};
- static const int gen6_samples[] = {4, 0, -1};
- static const int gen4_samples[] = {0, -1};
-
- if (brw->gen >= 7) {
- return gen7_samples;
- } else if (brw->gen == 6) {
- return gen6_samples;
- } else {
- return gen4_samples;
- }
-}
-
-/**
- * Override GL_MAX_SAMPLES and related constants according to value of driconf
- * option 'clamp_max_samples'.
- */
-static void
-brw_override_max_samples(struct brw_context *brw)
-{
- const int clamp_max_samples = driQueryOptioni(&brw->optionCache,
- "clamp_max_samples");
- if (clamp_max_samples < 0)
- return;
-
- const int *supported_msaa_modes = brw_supported_msaa_modes(brw);
- int max_samples = 0;
-
- /* Select the largest supported MSAA mode that does not exceed
- * clamp_max_samples.
- */
- for (int i = 0; supported_msaa_modes[i] != -1; ++i) {
- if (supported_msaa_modes[i] <= clamp_max_samples) {
- max_samples = supported_msaa_modes[i];
- break;
- }
- }
-
- brw->ctx.Const.MaxSamples = max_samples;
- brw->ctx.Const.MaxColorTextureSamples = max_samples;
- brw->ctx.Const.MaxDepthTextureSamples = max_samples;
- brw->ctx.Const.MaxIntegerSamples = max_samples;
-}
-
static void
brw_initialize_context_constants(struct brw_context *brw)
{
struct gl_context *ctx = &brw->ctx;
+ unsigned max_samplers =
+ brw->gen >= 8 || brw->is_haswell ? BRW_MAX_TEX_UNIT : 16;
+
ctx->Const.QueryCounterBits.Timestamp = 36;
ctx->Const.StripTextureBorder = true;
ctx->Const.MaxDualSourceDrawBuffers = 1;
ctx->Const.MaxDrawBuffers = BRW_MAX_DRAW_BUFFERS;
- ctx->Const.FragmentProgram.MaxTextureImageUnits = BRW_MAX_TEX_UNIT;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxTextureImageUnits = max_samplers;
ctx->Const.MaxTextureCoordUnits = 8; /* Mesa limit */
ctx->Const.MaxTextureUnits =
MIN2(ctx->Const.MaxTextureCoordUnits,
- ctx->Const.FragmentProgram.MaxTextureImageUnits);
- ctx->Const.VertexProgram.MaxTextureImageUnits = BRW_MAX_TEX_UNIT;
- if (brw->gen >= 7)
- ctx->Const.GeometryProgram.MaxTextureImageUnits = BRW_MAX_TEX_UNIT;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxTextureImageUnits);
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxTextureImageUnits = max_samplers;
+ if (brw->gen >= 6)
+ ctx->Const.Program[MESA_SHADER_GEOMETRY].MaxTextureImageUnits = max_samplers;
else
- ctx->Const.GeometryProgram.MaxTextureImageUnits = 0;
+ ctx->Const.Program[MESA_SHADER_GEOMETRY].MaxTextureImageUnits = 0;
+ if (_mesa_extension_override_enables.ARB_compute_shader) {
+ ctx->Const.Program[MESA_SHADER_COMPUTE].MaxTextureImageUnits = BRW_MAX_TEX_UNIT;
+ ctx->Const.MaxUniformBufferBindings += 12;
+ } else {
+ ctx->Const.Program[MESA_SHADER_COMPUTE].MaxTextureImageUnits = 0;
+ }
ctx->Const.MaxCombinedTextureImageUnits =
- ctx->Const.VertexProgram.MaxTextureImageUnits +
- ctx->Const.FragmentProgram.MaxTextureImageUnits +
- ctx->Const.GeometryProgram.MaxTextureImageUnits;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxTextureImageUnits +
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxTextureImageUnits +
+ ctx->Const.Program[MESA_SHADER_GEOMETRY].MaxTextureImageUnits +
+ ctx->Const.Program[MESA_SHADER_COMPUTE].MaxTextureImageUnits;
ctx->Const.MaxTextureLevels = 14; /* 8192 */
if (ctx->Const.MaxTextureLevels > MAX_TEXTURE_LEVELS)
ctx->Const.MaxTextureLevels = MAX_TEXTURE_LEVELS;
- ctx->Const.Max3DTextureLevels = 9;
- ctx->Const.MaxCubeTextureLevels = 12;
+ ctx->Const.Max3DTextureLevels = 12; /* 2048 */
+ ctx->Const.MaxCubeTextureLevels = 14; /* 8192 */
+ ctx->Const.MaxTextureMbytes = 1536;
if (brw->gen >= 7)
ctx->Const.MaxArrayTextureLayers = 2048;
ctx->Const.MaxArrayTextureLayers = 512;
ctx->Const.MaxTextureRectSize = 1 << 12;
-
+
ctx->Const.MaxTextureMaxAnisotropy = 16.0;
ctx->Const.MaxRenderbufferSize = 8192;
ctx->Const.AlwaysUseGetTransformFeedbackVertexCount = true;
- const int max_samples = brw_supported_msaa_modes(brw)[0];
+ int max_samples;
+ const int *msaa_modes = intel_supported_msaa_modes(brw->intelScreen);
+ const int clamp_max_samples =
+ driQueryOptioni(&brw->optionCache, "clamp_max_samples");
+
+ if (clamp_max_samples < 0) {
+ max_samples = msaa_modes[0];
+ } else {
+ /* Select the largest supported MSAA mode that does not exceed
+ * clamp_max_samples.
+ */
+ max_samples = 0;
+ for (int i = 0; msaa_modes[i] != 0; ++i) {
+ if (msaa_modes[i] <= clamp_max_samples) {
+ max_samples = msaa_modes[i];
+ break;
+ }
+ }
+ }
+
ctx->Const.MaxSamples = max_samples;
ctx->Const.MaxColorTextureSamples = max_samples;
ctx->Const.MaxDepthTextureSamples = max_samples;
ctx->Const.MaxIntegerSamples = max_samples;
+ /* gen6_set_sample_maps() sets SampleMap{2,4,8}x variables which are used
+ * to map indices of rectangular grid to sample numbers within a pixel.
+ * These variables are used by GL_EXT_framebuffer_multisample_blit_scaled
+ * extension implementation. For more details see the comment above
+ * gen6_set_sample_maps() definition.
+ */
+ gen6_set_sample_maps(ctx);
+
if (brw->gen >= 7)
ctx->Const.MaxProgramTextureGatherComponents = 4;
+ else if (brw->gen == 6)
+ ctx->Const.MaxProgramTextureGatherComponents = 1;
ctx->Const.MinLineWidth = 1.0;
ctx->Const.MinLineWidthAA = 1.0;
- ctx->Const.MaxLineWidth = 5.0;
- ctx->Const.MaxLineWidthAA = 5.0;
- ctx->Const.LineWidthGranularity = 0.5;
+ if (brw->gen >= 9 || brw->is_cherryview) {
+ ctx->Const.MaxLineWidth = 40.0;
+ ctx->Const.MaxLineWidthAA = 40.0;
+ ctx->Const.LineWidthGranularity = 0.125;
+ } else if (brw->gen >= 6) {
+ ctx->Const.MaxLineWidth = 7.375;
+ ctx->Const.MaxLineWidthAA = 7.375;
+ ctx->Const.LineWidthGranularity = 0.125;
+ } else {
+ ctx->Const.MaxLineWidth = 7.0;
+ ctx->Const.MaxLineWidthAA = 7.0;
+ ctx->Const.LineWidthGranularity = 0.5;
+ }
ctx->Const.MinPointSize = 1.0;
ctx->Const.MinPointSizeAA = 1.0;
if (brw->gen >= 5 || brw->is_g4x)
ctx->Const.MaxClipPlanes = 8;
- ctx->Const.VertexProgram.MaxNativeInstructions = 16 * 1024;
- ctx->Const.VertexProgram.MaxAluInstructions = 0;
- ctx->Const.VertexProgram.MaxTexInstructions = 0;
- ctx->Const.VertexProgram.MaxTexIndirections = 0;
- ctx->Const.VertexProgram.MaxNativeAluInstructions = 0;
- ctx->Const.VertexProgram.MaxNativeTexInstructions = 0;
- ctx->Const.VertexProgram.MaxNativeTexIndirections = 0;
- ctx->Const.VertexProgram.MaxNativeAttribs = 16;
- ctx->Const.VertexProgram.MaxNativeTemps = 256;
- ctx->Const.VertexProgram.MaxNativeAddressRegs = 1;
- ctx->Const.VertexProgram.MaxNativeParameters = 1024;
- ctx->Const.VertexProgram.MaxEnvParams =
- MIN2(ctx->Const.VertexProgram.MaxNativeParameters,
- ctx->Const.VertexProgram.MaxEnvParams);
-
- ctx->Const.FragmentProgram.MaxNativeInstructions = 1024;
- ctx->Const.FragmentProgram.MaxNativeAluInstructions = 1024;
- ctx->Const.FragmentProgram.MaxNativeTexInstructions = 1024;
- ctx->Const.FragmentProgram.MaxNativeTexIndirections = 1024;
- ctx->Const.FragmentProgram.MaxNativeAttribs = 12;
- ctx->Const.FragmentProgram.MaxNativeTemps = 256;
- ctx->Const.FragmentProgram.MaxNativeAddressRegs = 0;
- ctx->Const.FragmentProgram.MaxNativeParameters = 1024;
- ctx->Const.FragmentProgram.MaxEnvParams =
- MIN2(ctx->Const.FragmentProgram.MaxNativeParameters,
- ctx->Const.FragmentProgram.MaxEnvParams);
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxNativeInstructions = 16 * 1024;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxAluInstructions = 0;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxTexInstructions = 0;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxTexIndirections = 0;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxNativeAluInstructions = 0;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxNativeTexInstructions = 0;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxNativeTexIndirections = 0;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxNativeAttribs = 16;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxNativeTemps = 256;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxNativeAddressRegs = 1;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxNativeParameters = 1024;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxEnvParams =
+ MIN2(ctx->Const.Program[MESA_SHADER_VERTEX].MaxNativeParameters,
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxEnvParams);
+
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxNativeInstructions = 1024;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxNativeAluInstructions = 1024;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxNativeTexInstructions = 1024;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxNativeTexIndirections = 1024;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxNativeAttribs = 12;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxNativeTemps = 256;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxNativeAddressRegs = 0;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxNativeParameters = 1024;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxEnvParams =
+ MIN2(ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxNativeParameters,
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxEnvParams);
/* Fragment shaders use real, 32-bit twos-complement integers for all
* integer types.
*/
- ctx->Const.FragmentProgram.LowInt.RangeMin = 31;
- ctx->Const.FragmentProgram.LowInt.RangeMax = 30;
- ctx->Const.FragmentProgram.LowInt.Precision = 0;
- ctx->Const.FragmentProgram.HighInt = ctx->Const.FragmentProgram.LowInt;
- ctx->Const.FragmentProgram.MediumInt = ctx->Const.FragmentProgram.LowInt;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].LowInt.RangeMin = 31;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].LowInt.RangeMax = 30;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].LowInt.Precision = 0;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].HighInt = ctx->Const.Program[MESA_SHADER_FRAGMENT].LowInt;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MediumInt = ctx->Const.Program[MESA_SHADER_FRAGMENT].LowInt;
+
+ ctx->Const.Program[MESA_SHADER_VERTEX].LowInt.RangeMin = 31;
+ ctx->Const.Program[MESA_SHADER_VERTEX].LowInt.RangeMax = 30;
+ ctx->Const.Program[MESA_SHADER_VERTEX].LowInt.Precision = 0;
+ ctx->Const.Program[MESA_SHADER_VERTEX].HighInt = ctx->Const.Program[MESA_SHADER_VERTEX].LowInt;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MediumInt = ctx->Const.Program[MESA_SHADER_VERTEX].LowInt;
if (brw->gen >= 7) {
- ctx->Const.FragmentProgram.MaxAtomicCounters = MAX_ATOMIC_COUNTERS;
- ctx->Const.VertexProgram.MaxAtomicCounters = MAX_ATOMIC_COUNTERS;
- ctx->Const.GeometryProgram.MaxAtomicCounters = MAX_ATOMIC_COUNTERS;
- ctx->Const.FragmentProgram.MaxAtomicBuffers = BRW_MAX_ABO;
- ctx->Const.VertexProgram.MaxAtomicBuffers = BRW_MAX_ABO;
- ctx->Const.GeometryProgram.MaxAtomicBuffers = BRW_MAX_ABO;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxAtomicCounters = MAX_ATOMIC_COUNTERS;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxAtomicCounters = MAX_ATOMIC_COUNTERS;
+ ctx->Const.Program[MESA_SHADER_GEOMETRY].MaxAtomicCounters = MAX_ATOMIC_COUNTERS;
+ ctx->Const.Program[MESA_SHADER_COMPUTE].MaxAtomicCounters = MAX_ATOMIC_COUNTERS;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxAtomicBuffers = BRW_MAX_ABO;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxAtomicBuffers = BRW_MAX_ABO;
+ ctx->Const.Program[MESA_SHADER_GEOMETRY].MaxAtomicBuffers = BRW_MAX_ABO;
+ ctx->Const.Program[MESA_SHADER_COMPUTE].MaxAtomicBuffers = BRW_MAX_ABO;
ctx->Const.MaxCombinedAtomicBuffers = 3 * BRW_MAX_ABO;
}
ctx->Const.QuadsFollowProvokingVertexConvention = false;
ctx->Const.NativeIntegers = true;
- ctx->Const.UniformBooleanTrue = 1;
+ ctx->Const.VertexID_is_zero_based = true;
+
+ /* Regarding the CMP instruction, the Ivybridge PRM says:
+ *
+ * "For each enabled channel 0b or 1b is assigned to the appropriate flag
+ * bit and 0/all zeros or all ones (e.g, byte 0xFF, word 0xFFFF, DWord
+ * 0xFFFFFFFF) is assigned to dst."
+ *
+ * but PRMs for earlier generations say
+ *
+ * "In dword format, one GRF may store up to 8 results. When the register
+ * is used later as a vector of Booleans, as only LSB at each channel
+ * contains meaning [sic] data, software should make sure all higher bits
+ * are masked out (e.g. by 'and-ing' an [sic] 0x01 constant)."
+ *
+ * We select the representation of a true boolean uniform to be ~0, and fix
+ * the results of Gen <= 5 CMP instruction's with -(result & 1).
+ */
+ ctx->Const.UniformBooleanTrue = ~0;
/* From the gen4 PRM, volume 4 page 127:
*
if (brw->gen >= 6) {
ctx->Const.MaxVarying = 32;
- ctx->Const.VertexProgram.MaxOutputComponents = 128;
- ctx->Const.GeometryProgram.MaxInputComponents = 64;
- ctx->Const.GeometryProgram.MaxOutputComponents = 128;
- ctx->Const.FragmentProgram.MaxInputComponents = 128;
+ ctx->Const.Program[MESA_SHADER_VERTEX].MaxOutputComponents = 128;
+ ctx->Const.Program[MESA_SHADER_GEOMETRY].MaxInputComponents = 64;
+ ctx->Const.Program[MESA_SHADER_GEOMETRY].MaxOutputComponents = 128;
+ ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxInputComponents = 128;
}
+ static const nir_shader_compiler_options gen4_nir_options = {
+ .native_integers = true,
+ .lower_ffma = true,
+ };
+
+ static const nir_shader_compiler_options gen6_nir_options = {
+ .native_integers = true,
+ };
+
/* We want the GLSL compiler to emit code that uses condition codes */
- for (int i = 0; i < MESA_SHADER_TYPES; i++) {
- ctx->ShaderCompilerOptions[i].MaxIfDepth = brw->gen < 6 ? 16 : UINT_MAX;
- ctx->ShaderCompilerOptions[i].EmitCondCodes = true;
- ctx->ShaderCompilerOptions[i].EmitNoNoise = true;
- ctx->ShaderCompilerOptions[i].EmitNoMainReturn = true;
- ctx->ShaderCompilerOptions[i].EmitNoIndirectInput = true;
- ctx->ShaderCompilerOptions[i].EmitNoIndirectOutput = true;
-
- ctx->ShaderCompilerOptions[i].EmitNoIndirectUniform =
+ for (int i = 0; i < MESA_SHADER_STAGES; i++) {
+ ctx->Const.ShaderCompilerOptions[i].MaxIfDepth = brw->gen < 6 ? 16 : UINT_MAX;
+ ctx->Const.ShaderCompilerOptions[i].EmitCondCodes = true;
+ ctx->Const.ShaderCompilerOptions[i].EmitNoNoise = true;
+ ctx->Const.ShaderCompilerOptions[i].EmitNoMainReturn = true;
+ ctx->Const.ShaderCompilerOptions[i].EmitNoIndirectInput = true;
+ ctx->Const.ShaderCompilerOptions[i].EmitNoIndirectOutput =
(i == MESA_SHADER_FRAGMENT);
- ctx->ShaderCompilerOptions[i].EmitNoIndirectTemp =
+ ctx->Const.ShaderCompilerOptions[i].EmitNoIndirectTemp =
(i == MESA_SHADER_FRAGMENT);
- ctx->ShaderCompilerOptions[i].LowerClipDistance = true;
+ ctx->Const.ShaderCompilerOptions[i].EmitNoIndirectUniform = false;
+ ctx->Const.ShaderCompilerOptions[i].LowerClipDistance = true;
+ if (brw->gen >= 6)
+ ctx->Const.ShaderCompilerOptions[i].NirOptions = &gen6_nir_options;
+ else
+ ctx->Const.ShaderCompilerOptions[i].NirOptions = &gen4_nir_options;
+ }
+
+ ctx->Const.ShaderCompilerOptions[MESA_SHADER_VERTEX].OptimizeForAOS = true;
+ ctx->Const.ShaderCompilerOptions[MESA_SHADER_GEOMETRY].OptimizeForAOS = true;
+
+ if (brw->scalar_vs) {
+ /* If we're using the scalar backend for vertex shaders, we need to
+ * configure these accordingly.
+ */
+ ctx->Const.ShaderCompilerOptions[MESA_SHADER_VERTEX].EmitNoIndirectOutput = true;
+ ctx->Const.ShaderCompilerOptions[MESA_SHADER_VERTEX].EmitNoIndirectTemp = true;
+ ctx->Const.ShaderCompilerOptions[MESA_SHADER_VERTEX].OptimizeForAOS = false;
}
- ctx->ShaderCompilerOptions[MESA_SHADER_VERTEX].PreferDP4 = true;
+ /* ARB_viewport_array */
+ if (brw->gen >= 7 && ctx->API == API_OPENGL_CORE) {
+ ctx->Const.MaxViewports = GEN7_NUM_VIEWPORTS;
+ ctx->Const.ViewportSubpixelBits = 0;
+
+ /* Cast to float before negating becuase MaxViewportWidth is unsigned.
+ */
+ ctx->Const.ViewportBounds.Min = -(float)ctx->Const.MaxViewportWidth;
+ ctx->Const.ViewportBounds.Max = ctx->Const.MaxViewportWidth;
+ }
+
+ /* ARB_gpu_shader5 */
+ if (brw->gen >= 7)
+ ctx->Const.MaxVertexStreams = MIN2(4, MAX_VERTEX_STREAMS);
}
/**
brw->disable_throttling = true;
}
- brw->disable_derivative_optimization =
- driQueryOptionb(&brw->optionCache, "disable_derivative_optimization");
-
brw->precompile = driQueryOptionb(&brw->optionCache, "shader_precompile");
ctx->Const.ForceGLSLExtensionsWarn =
ctx->Const.DisableGLSLLineContinuations =
driQueryOptionb(options, "disable_glsl_line_continuations");
+
+ ctx->Const.AllowGLSLExtensionDirectiveMidShader =
+ driQueryOptionb(options, "allow_glsl_extension_directive_midshader");
+}
+
+/* drop when libdrm 2.4.61 is released */
+#ifndef I915_PARAM_REVISION
+#define I915_PARAM_REVISION 32
+#endif
+
+static int
+brw_get_revision(int fd)
+{
+ struct drm_i915_getparam gp;
+ int revision;
+ int ret;
+
+ memset(&gp, 0, sizeof(gp));
+ gp.param = I915_PARAM_REVISION;
+ gp.value = &revision;
+
+ ret = drmCommandWriteRead(fd, DRM_I915_GETPARAM, &gp, sizeof(gp));
+ if (ret)
+ revision = -1;
+
+ return revision;
}
GLboolean
struct intel_screen *screen = sPriv->driverPrivate;
const struct brw_device_info *devinfo = screen->devinfo;
struct dd_function_table functions;
- struct gl_config visual;
- if (flags & ~(__DRI_CTX_FLAG_DEBUG
- | __DRI_CTX_FLAG_FORWARD_COMPATIBLE
- | __DRI_CTX_FLAG_ROBUST_BUFFER_ACCESS)) {
+ /* Only allow the __DRI_CTX_FLAG_ROBUST_BUFFER_ACCESS flag if the kernel
+ * provides us with context reset notifications.
+ */
+ uint32_t allowed_flags = __DRI_CTX_FLAG_DEBUG
+ | __DRI_CTX_FLAG_FORWARD_COMPATIBLE;
+
+ if (screen->has_context_reset_notification)
+ allowed_flags |= __DRI_CTX_FLAG_ROBUST_BUFFER_ACCESS;
+
+ if (flags & ~allowed_flags) {
*dri_ctx_error = __DRI_CTX_ERROR_UNKNOWN_FLAG;
return false;
}
struct brw_context *brw = rzalloc(NULL, struct brw_context);
if (!brw) {
- printf("%s: failed to alloc context\n", __FUNCTION__);
+ fprintf(stderr, "%s: failed to alloc context\n", __FUNCTION__);
*dri_ctx_error = __DRI_CTX_ERROR_NO_MEMORY;
return false;
}
brw->is_g4x = devinfo->is_g4x;
brw->is_baytrail = devinfo->is_baytrail;
brw->is_haswell = devinfo->is_haswell;
+ brw->is_cherryview = devinfo->is_cherryview;
brw->has_llc = devinfo->has_llc;
- brw->has_hiz = devinfo->has_hiz_and_separate_stencil && brw->gen < 8;
+ brw->has_hiz = devinfo->has_hiz_and_separate_stencil;
brw->has_separate_stencil = devinfo->has_hiz_and_separate_stencil;
brw->has_pln = devinfo->has_pln;
brw->has_compr4 = devinfo->has_compr4;
brw->has_negative_rhw_bug = devinfo->has_negative_rhw_bug;
brw->needs_unlit_centroid_workaround =
devinfo->needs_unlit_centroid_workaround;
+ brw->revision = brw_get_revision(sPriv->fd);
brw->must_use_separate_stencil = screen->hw_must_use_separate_stencil;
brw->has_swizzling = screen->hw_has_swizzling;
- if (brw->gen >= 7) {
+ brw->vs.base.stage = MESA_SHADER_VERTEX;
+ brw->gs.base.stage = MESA_SHADER_GEOMETRY;
+ brw->wm.base.stage = MESA_SHADER_FRAGMENT;
+ if (brw->gen >= 8) {
+ gen8_init_vtable_surface_functions(brw);
+ brw->vtbl.emit_depth_stencil_hiz = gen8_emit_depth_stencil_hiz;
+ } else if (brw->gen >= 7) {
gen7_init_vtable_surface_functions(brw);
- gen7_init_vtable_sampler_functions(brw);
brw->vtbl.emit_depth_stencil_hiz = gen7_emit_depth_stencil_hiz;
+ } else if (brw->gen >= 6) {
+ gen6_init_vtable_surface_functions(brw);
+ brw->vtbl.emit_depth_stencil_hiz = gen6_emit_depth_stencil_hiz;
} else {
gen4_init_vtable_surface_functions(brw);
- gen4_init_vtable_sampler_functions(brw);
brw->vtbl.emit_depth_stencil_hiz = brw_emit_depth_stencil_hiz;
}
struct gl_context *ctx = &brw->ctx;
- if (mesaVis == NULL) {
- memset(&visual, 0, sizeof visual);
- mesaVis = &visual;
- }
-
if (!_mesa_initialize_context(ctx, api, mesaVis, shareCtx, &functions)) {
*dri_ctx_error = __DRI_CTX_ERROR_NO_MEMORY;
- printf("%s: failed to init mesa context\n", __FUNCTION__);
+ fprintf(stderr, "%s: failed to init mesa context\n", __FUNCTION__);
intelDestroyContext(driContextPriv);
return false;
}
+ driContextSetFlags(ctx, flags);
+
/* Initialize the software rasterizer and helper modules.
*
* As of GL 3.1 core, the gen4+ driver doesn't need the swrast context for
brw_process_driconf_options(brw);
brw_process_intel_debug_variable(brw);
+
+ if (brw->gen >= 8 && !(INTEL_DEBUG & DEBUG_VEC4VS))
+ brw->scalar_vs = true;
+
brw_initialize_context_constants(brw);
ctx->Const.ResetStrategy = notify_reset
/* Reinitialize the context point state. It depends on ctx->Const values. */
_mesa_init_point(ctx);
- intel_batchbuffer_init(brw);
-
- brw_init_state(brw);
-
- intelInitExtensions(ctx);
-
intel_fbo_init(brw);
+ intel_batchbuffer_init(brw);
+
if (brw->gen >= 6) {
/* Create a new hardware context. Using a hardware context means that
* our GPU state will be saved/restored on context switch, allowing us
}
}
- /* Notification of GPU resets requires hardware contexts and a kernel new
- * enough to support DRM_IOCTL_I915_GET_RESET_STATS, which isn't upstream
- * yet.
- */
- if (notify_reset) {
- /* This is the wrong error code, but the correct error code (one that
- * will cause EGL to generate EGL_BAD_MATCH) doesn't seem to exist.
- */
- *dri_ctx_error = __DRI_CTX_ERROR_UNKNOWN_ATTRIBUTE;
- intelDestroyContext(driContextPriv);
- return false;
- }
+ brw_init_state(brw);
- brw_init_surface_formats(brw);
+ intelInitExtensions(ctx);
- if (brw->is_g4x || brw->gen >= 5) {
- brw->CMD_VF_STATISTICS = GM45_3DSTATE_VF_STATISTICS;
- brw->CMD_PIPELINE_SELECT = CMD_PIPELINE_SELECT_GM45;
- } else {
- brw->CMD_VF_STATISTICS = GEN4_3DSTATE_VF_STATISTICS;
- brw->CMD_PIPELINE_SELECT = CMD_PIPELINE_SELECT_965;
- }
+ brw_init_surface_formats(brw);
brw->max_vs_threads = devinfo->max_vs_threads;
+ brw->max_hs_threads = devinfo->max_hs_threads;
+ brw->max_ds_threads = devinfo->max_ds_threads;
brw->max_gs_threads = devinfo->max_gs_threads;
brw->max_wm_threads = devinfo->max_wm_threads;
brw->urb.size = devinfo->urb.size;
brw->urb.min_vs_entries = devinfo->urb.min_vs_entries;
brw->urb.max_vs_entries = devinfo->urb.max_vs_entries;
+ brw->urb.max_hs_entries = devinfo->urb.max_hs_entries;
+ brw->urb.max_ds_entries = devinfo->urb.max_ds_entries;
brw->urb.max_gs_entries = devinfo->urb.max_gs_entries;
/* Estimate the size of the mappable aperture into the GTT. There's an
brw->max_gtt_map_object_size = gtt_size / 4;
if (brw->gen == 6)
- brw->urb.gen6_gs_previously_active = false;
+ brw->urb.gs_present = false;
brw->prim_restart.in_progress = false;
brw->prim_restart.enable_cut_index = false;
brw->gs.enabled = false;
-
- if (brw->gen < 6) {
- brw->curbe.last_buf = calloc(1, 4096);
- brw->curbe.next_buf = calloc(1, 4096);
- }
+ brw->sf.viewport_transform_enable = true;
ctx->VertexProgram._MaintainTnlProgram = true;
ctx->FragmentProgram._MaintainTexEnvProgram = true;
if ((flags & __DRI_CTX_FLAG_ROBUST_BUFFER_ACCESS) != 0)
ctx->Const.ContextFlags |= GL_CONTEXT_FLAG_ROBUST_ACCESS_BIT_ARB;
- brw_fs_alloc_reg_sets(brw);
- brw_vec4_alloc_reg_set(brw);
-
if (INTEL_DEBUG & DEBUG_SHADER_TIME)
brw_init_shader_time(brw);
_mesa_compute_version(ctx);
- /* Here we override context constants. We apply the overrides after
- * calculation of the context version because we do not want the overridden
- * constants to change the version.
- */
- brw_override_max_samples(brw);
-
_mesa_initialize_dispatch_tables(ctx);
_mesa_initialize_vbo_vtxfmt(ctx);
brw_init_performance_monitors(brw);
}
+ vbo_use_buffer_objects(ctx);
+ vbo_always_unmap_buffers(ctx);
+
return true;
}
}
_mesa_meta_free(&brw->ctx);
+ brw_meta_fast_clear_free(brw);
if (INTEL_DEBUG & DEBUG_SHADER_TIME) {
/* Force a report. */
brw_draw_destroy(brw);
drm_intel_bo_unreference(brw->curbe.curbe_bo);
- drm_intel_bo_unreference(brw->vs.base.const_bo);
- drm_intel_bo_unreference(brw->wm.base.const_bo);
-
- free(brw->curbe.last_buf);
- free(brw->curbe.next_buf);
+ if (brw->vs.base.scratch_bo)
+ drm_intel_bo_unreference(brw->vs.base.scratch_bo);
+ if (brw->gs.base.scratch_bo)
+ drm_intel_bo_unreference(brw->gs.base.scratch_bo);
+ if (brw->wm.base.scratch_bo)
+ drm_intel_bo_unreference(brw->wm.base.scratch_bo);
drm_intel_gem_context_destroy(brw->hw_ctx);
intel_batchbuffer_free(brw);
- drm_intel_bo_unreference(brw->first_post_swapbuffers_batch);
- brw->first_post_swapbuffers_batch = NULL;
+ drm_intel_bo_unreference(brw->throttle_batch[1]);
+ drm_intel_bo_unreference(brw->throttle_batch[0]);
+ brw->throttle_batch[1] = NULL;
+ brw->throttle_batch[0] = NULL;
driDestroyOptionCache(&brw->optionCache);
fb->Visual.sRGBCapable = false;
for (int i = 0; i < BUFFER_COUNT; i++) {
if (fb->Attachment[i].Renderbuffer &&
- fb->Attachment[i].Renderbuffer->Format == MESA_FORMAT_SARGB8) {
- fb->Attachment[i].Renderbuffer->Format = MESA_FORMAT_ARGB8888;
+ fb->Attachment[i].Renderbuffer->Format == MESA_FORMAT_B8G8R8A8_SRGB) {
+ fb->Attachment[i].Renderbuffer->Format = MESA_FORMAT_B8G8R8A8_UNORM;
}
}
}
struct gl_context *ctx = &brw->ctx;
struct gl_framebuffer *fb, *readFb;
- if (driDrawPriv == NULL && driReadPriv == NULL) {
+ if (driDrawPriv == NULL) {
fb = _mesa_get_incomplete_framebuffer();
- readFb = _mesa_get_incomplete_framebuffer();
} else {
fb = driDrawPriv->driverPrivate;
- readFb = driReadPriv->driverPrivate;
driContextPriv->dri2.draw_stamp = driDrawPriv->dri2.stamp - 1;
+ }
+
+ if (driReadPriv == NULL) {
+ readFb = _mesa_get_incomplete_framebuffer();
+ } else {
+ readFb = driReadPriv->driverPrivate;
driContextPriv->dri2.read_stamp = driReadPriv->dri2.stamp - 1;
}
intel_gles3_srgb_workaround(brw, fb);
intel_gles3_srgb_workaround(brw, readFb);
- intel_prepare_render(brw);
+ /* If the context viewport hasn't been initialized, force a call out to
+ * the loader to get buffers so we have a drawable size for the initial
+ * viewport. */
+ if (!brw->ctx.ViewportInitialized)
+ intel_prepare_render(brw);
+
_mesa_make_current(ctx, fb, readFb);
} else {
_mesa_make_current(NULL, NULL, NULL);
if (rb->mt->num_samples <= 1)
intel_miptree_resolve_color(brw, rb->mt);
else
- intel_miptree_downsample(brw, rb->mt);
+ intel_renderbuffer_downsample(brw, rb);
}
}
void
intel_prepare_render(struct brw_context *brw)
{
+ struct gl_context *ctx = &brw->ctx;
__DRIcontext *driContext = brw->driContext;
__DRIdrawable *drawable;
* that will happen next will probably dirty the front buffer. So
* mark it as dirty here.
*/
- if (brw->is_front_buffer_rendering)
+ if (brw_is_front_buffer_drawing(ctx->DrawBuffer))
brw->front_buffer_dirty = true;
-
- /* Wait for the swapbuffers before the one we just emitted, so we
- * don't get too many swaps outstanding for apps that are GPU-heavy
- * but not CPU-heavy.
- *
- * We're using intelDRI2Flush (called from the loader before
- * swapbuffer) and glFlush (for front buffer rendering) as the
- * indicator that a frame is done and then throttle when we get
- * here as we prepare to render the next frame. At this point for
- * round trips for swap/copy and getting new buffers are done and
- * we'll spend less time waiting on the GPU.
- *
- * Unfortunately, we don't have a handle to the batch containing
- * the swap, and getting our hands on that doesn't seem worth it,
- * so we just us the first batch we emitted after the last swap.
- */
- if (brw->need_throttle && brw->first_post_swapbuffers_batch) {
- if (!brw->disable_throttling)
- drm_intel_bo_wait_rendering(brw->first_post_swapbuffers_batch);
- drm_intel_bo_unreference(brw->first_post_swapbuffers_batch);
- brw->first_post_swapbuffers_batch = NULL;
- brw->need_throttle = false;
- }
}
/**
back_rb = intel_get_renderbuffer(fb, BUFFER_BACK_LEFT);
memset(attachments, 0, sizeof(attachments));
- if ((brw->is_front_buffer_rendering ||
- brw->is_front_buffer_reading ||
+ if ((brw_is_front_buffer_drawing(fb) ||
+ brw_is_front_buffer_reading(fb) ||
!back_rb) && front_rb) {
/* If a fake front buffer is in use, then querying for
* __DRI_BUFFER_FRONT_LEFT will cause the server to copy the image from
* DRI2BufferDepthStencil are handled as special cases.
*
* \param buffer_name is a human readable name, such as "dri2 front buffer",
- * that is passed to intel_region_alloc_for_handle().
+ * that is passed to drm_intel_bo_gem_create_from_name().
*
* \see intel_update_renderbuffers()
- * \see intel_region_alloc_for_handle()
*/
static void
intel_process_dri2_buffer(struct brw_context *brw,
struct intel_renderbuffer *rb,
const char *buffer_name)
{
- struct intel_region *region = NULL;
+ struct gl_framebuffer *fb = drawable->driverPrivate;
+ drm_intel_bo *bo;
if (!rb)
return;
* use of a mapping of the buffer involves a bunch of page faulting which is
* moderately expensive.
*/
- if (num_samples == 0) {
- if (rb->mt &&
- rb->mt->region &&
- rb->mt->region->name == buffer->name)
- return;
- } else {
- if (rb->mt &&
- rb->mt->singlesample_mt &&
- rb->mt->singlesample_mt->region &&
- rb->mt->singlesample_mt->region->name == buffer->name)
- return;
+ struct intel_mipmap_tree *last_mt;
+ if (num_samples == 0)
+ last_mt = rb->mt;
+ else
+ last_mt = rb->singlesample_mt;
+
+ uint32_t old_name = 0;
+ if (last_mt) {
+ /* The bo already has a name because the miptree was created by a
+ * previous call to intel_process_dri2_buffer(). If a bo already has a
+ * name, then drm_intel_bo_flink() is a low-cost getter. It does not
+ * create a new name.
+ */
+ drm_intel_bo_flink(last_mt->bo, &old_name);
}
+ if (old_name == buffer->name)
+ return;
+
if (unlikely(INTEL_DEBUG & DEBUG_DRI)) {
fprintf(stderr,
"attaching buffer %d, at %d, cpp %d, pitch %d\n",
}
intel_miptree_release(&rb->mt);
- region = intel_region_alloc_for_handle(brw->intelScreen,
- buffer->cpp,
- drawable->w,
- drawable->h,
- buffer->pitch,
- buffer->name,
- buffer_name);
- if (!region)
+ bo = drm_intel_bo_gem_create_from_name(brw->bufmgr, buffer_name,
+ buffer->name);
+ if (!bo) {
+ fprintf(stderr,
+ "Failed to open BO for returned DRI2 buffer "
+ "(%dx%d, %s, named %d).\n"
+ "This is likely a bug in the X Server that will lead to a "
+ "crash soon.\n",
+ drawable->w, drawable->h, buffer_name, buffer->name);
return;
+ }
+
+ intel_update_winsys_renderbuffer_miptree(brw, rb, bo,
+ drawable->w, drawable->h,
+ buffer->pitch);
+
+ if (brw_is_front_buffer_drawing(fb) &&
+ (buffer->attachment == __DRI_BUFFER_FRONT_LEFT ||
+ buffer->attachment == __DRI_BUFFER_FAKE_FRONT_LEFT) &&
+ rb->Base.Base.NumSamples > 1) {
+ intel_renderbuffer_upsample(brw, rb);
+ }
- rb->mt = intel_miptree_create_for_dri2_buffer(brw,
- buffer->attachment,
- intel_rb_format(rb),
- num_samples,
- region);
- intel_region_release(®ion);
+ assert(rb->mt);
+
+ drm_intel_bo_unreference(bo);
}
/**
__DRIimage *buffer,
enum __DRIimageBufferMask buffer_type)
{
- struct intel_region *region = buffer->region;
+ struct gl_framebuffer *fb = drawable->driverPrivate;
- if (!rb || !region)
+ if (!rb || !buffer->bo)
return;
unsigned num_samples = rb->Base.Base.NumSamples;
- if (rb->mt &&
- rb->mt->region &&
- rb->mt->region == region)
+ /* Check and see if we're already bound to the right
+ * buffer object
+ */
+ struct intel_mipmap_tree *last_mt;
+ if (num_samples == 0)
+ last_mt = rb->mt;
+ else
+ last_mt = rb->singlesample_mt;
+
+ if (last_mt && last_mt->bo == buffer->bo)
return;
- intel_miptree_release(&rb->mt);
- rb->mt = intel_miptree_create_for_image_buffer(intel,
- buffer_type,
- intel_rb_format(rb),
- num_samples,
- region);
+ intel_update_winsys_renderbuffer_miptree(intel, rb, buffer->bo,
+ buffer->width, buffer->height,
+ buffer->pitch);
+
+ if (brw_is_front_buffer_drawing(fb) &&
+ buffer_type == __DRI_IMAGE_BUFFER_FRONT &&
+ rb->Base.Base.NumSamples > 1) {
+ intel_renderbuffer_upsample(intel, rb);
+ }
}
static void
else
return;
- if ((brw->is_front_buffer_rendering || brw->is_front_buffer_reading || !back_rb) && front_rb)
+ if (front_rb && (brw_is_front_buffer_drawing(fb) ||
+ brw_is_front_buffer_reading(fb) || !back_rb)) {
buffer_mask |= __DRI_IMAGE_BUFFER_FRONT;
+ }
if (back_rb)
buffer_mask |= __DRI_IMAGE_BUFFER_BACK;
projects / mesa.git / blobdiff