diff options
Diffstat (limited to 'src/gallium')
30 files changed, 1263 insertions, 708 deletions
diff --git a/src/gallium/auxiliary/draw/draw_context.c b/src/gallium/auxiliary/draw/draw_context.c index d3084fd428..d5ddc4a6a9 100644 --- a/src/gallium/auxiliary/draw/draw_context.c +++ b/src/gallium/auxiliary/draw/draw_context.c @@ -352,7 +352,10 @@ draw_find_shader_output(const struct draw_context *draw, /** - * Return number of the shader outputs. + * Return total number of the shader outputs. This function is similar to + * draw_current_shader_outputs() but this function also counts any extra + * vertex/geometry output attributes that may be filled in by some draw + * stages (such as AA point, AA line). * * If geometry shader is present, its output will be returned, * if not vertex shader is used. @@ -362,8 +365,9 @@ draw_num_shader_outputs(const struct draw_context *draw) { uint count = draw->vs.vertex_shader->info.num_outputs; - /* if geometry shader is present, its outputs go to te - * driver, not the vertex shaders */ + /* If a geometry shader is present, its outputs go to the + * driver, else the vertex shader's outputs. + */ if (draw->gs.geometry_shader) count = draw->gs.geometry_shader->info.num_outputs; @@ -374,7 +378,8 @@ draw_num_shader_outputs(const struct draw_context *draw) /** - * Provide TGSI sampler objects for vertex/geometry shaders that use texture fetches. + * Provide TGSI sampler objects for vertex/geometry shaders that use + * texture fetches. * This might only be used by software drivers for the time being. */ void @@ -454,14 +459,27 @@ void draw_do_flush( struct draw_context *draw, unsigned flags ) } -int draw_current_shader_outputs(struct draw_context *draw) +/** + * Return the number of output attributes produced by the geometry + * shader, if present. If no geometry shader, return the number of + * outputs from the vertex shader. + * \sa draw_num_shader_outputs + */ +uint +draw_current_shader_outputs(const struct draw_context *draw) { if (draw->gs.geometry_shader) return draw->gs.num_gs_outputs; return draw->vs.num_vs_outputs; } -int draw_current_shader_position_output(struct draw_context *draw) + +/** + * Return the index of the shader output which will contain the + * vertex position. + */ +uint +draw_current_shader_position_output(const struct draw_context *draw) { if (draw->gs.geometry_shader) return draw->gs.position_output; diff --git a/src/gallium/auxiliary/draw/draw_pipe_aaline.c b/src/gallium/auxiliary/draw/draw_pipe_aaline.c index e58129b99d..8f6ca15dfa 100644 --- a/src/gallium/auxiliary/draw/draw_pipe_aaline.c +++ b/src/gallium/auxiliary/draw/draw_pipe_aaline.c @@ -49,6 +49,10 @@ #include "draw_pipe.h" +/** Approx number of new tokens for instructions in aa_transform_inst() */ +#define NUM_NEW_TOKENS 50 + + /** * Max texture level for the alpha texture used for antialiasing */ @@ -179,12 +183,7 @@ aa_transform_decl(struct tgsi_transform_context *ctx, static int free_bit(uint bitfield) { - int i; - for (i = 0; i < 32; i++) { - if ((bitfield & (1 << i)) == 0) - return i; - } - return -1; + return ffs(~bitfield) - 1; } @@ -343,11 +342,10 @@ generate_aaline_fs(struct aaline_stage *aaline) const struct pipe_shader_state *orig_fs = &aaline->fs->state; struct pipe_shader_state aaline_fs; struct aa_transform_context transform; - -#define MAX 1000 + const uint newLen = tgsi_num_tokens(orig_fs->tokens) + NUM_NEW_TOKENS; aaline_fs = *orig_fs; /* copy to init */ - aaline_fs.tokens = MALLOC(sizeof(struct tgsi_token) * MAX); + aaline_fs.tokens = tgsi_alloc_tokens(newLen); if (aaline_fs.tokens == NULL) return FALSE; @@ -363,7 +361,7 @@ generate_aaline_fs(struct aaline_stage *aaline) tgsi_transform_shader(orig_fs->tokens, (struct tgsi_token *) aaline_fs.tokens, - MAX, &transform.base); + newLen, &transform.base); #if 0 /* DEBUG */ tgsi_dump(orig_fs->tokens, 0); diff --git a/src/gallium/auxiliary/draw/draw_pipe_aapoint.c b/src/gallium/auxiliary/draw/draw_pipe_aapoint.c index d86717e518..97f3480879 100644 --- a/src/gallium/auxiliary/draw/draw_pipe_aapoint.c +++ b/src/gallium/auxiliary/draw/draw_pipe_aapoint.c @@ -53,6 +53,10 @@ #include "draw_pipe.h" +/** Approx number of new tokens for instructions in aa_transform_inst() */ +#define NUM_NEW_TOKENS 200 + + /* * Enabling NORMALIZE might give _slightly_ better results. * Basically, it controls whether we compute distance as d=sqrt(x*x+y*y) or @@ -81,16 +85,19 @@ struct aapoint_stage { struct draw_stage stage; - int psize_slot; + /** half of pipe_rasterizer_state::point_size */ float radius; + /** vertex attrib slot containing point size */ + int psize_slot; + /** this is the vertex attrib slot for the new texcoords */ uint tex_slot; + + /** vertex attrib slot containing position */ uint pos_slot; - /* - * Currently bound state - */ + /** Currently bound fragment shader */ struct aapoint_fragment_shader *fs; /* @@ -491,11 +498,10 @@ generate_aapoint_fs(struct aapoint_stage *aapoint) const struct pipe_shader_state *orig_fs = &aapoint->fs->state; struct pipe_shader_state aapoint_fs; struct aa_transform_context transform; - -#define MAX 1000 + const uint newLen = tgsi_num_tokens(orig_fs->tokens) + NUM_NEW_TOKENS; aapoint_fs = *orig_fs; /* copy to init */ - aapoint_fs.tokens = MALLOC(sizeof(struct tgsi_token) * MAX); + aapoint_fs.tokens = tgsi_alloc_tokens(newLen); if (aapoint_fs.tokens == NULL) return FALSE; @@ -511,7 +517,7 @@ generate_aapoint_fs(struct aapoint_stage *aapoint) tgsi_transform_shader(orig_fs->tokens, (struct tgsi_token *) aapoint_fs.tokens, - MAX, &transform.base); + newLen, &transform.base); #if 0 /* DEBUG */ printf("draw_aapoint, orig shader:\n"); @@ -575,8 +581,8 @@ aapoint_point(struct draw_stage *stage, struct prim_header *header) const struct aapoint_stage *aapoint = aapoint_stage(stage); struct prim_header tri; struct vertex_header *v[4]; - uint texPos = aapoint->tex_slot; - uint pos_slot = aapoint->pos_slot; + const uint tex_slot = aapoint->tex_slot; + const uint pos_slot = aapoint->pos_slot; float radius, *pos, *tex; uint i; float k; @@ -643,16 +649,16 @@ aapoint_point(struct draw_stage *stage, struct prim_header *header) pos[1] += radius; /* new texcoords */ - tex = v[0]->data[texPos]; + tex = v[0]->data[tex_slot]; ASSIGN_4V(tex, -1, -1, k, 1); - tex = v[1]->data[texPos]; + tex = v[1]->data[tex_slot]; ASSIGN_4V(tex, 1, -1, k, 1); - tex = v[2]->data[texPos]; + tex = v[2]->data[tex_slot]; ASSIGN_4V(tex, 1, 1, k, 1); - tex = v[3]->data[texPos]; + tex = v[3]->data[tex_slot]; ASSIGN_4V(tex, -1, 1, k, 1); /* emit 2 tris for the quad strip */ diff --git a/src/gallium/auxiliary/draw/draw_pipe_pstipple.c b/src/gallium/auxiliary/draw/draw_pipe_pstipple.c index f047d8359c..d0d99aa331 100644 --- a/src/gallium/auxiliary/draw/draw_pipe_pstipple.c +++ b/src/gallium/auxiliary/draw/draw_pipe_pstipple.c @@ -50,6 +50,9 @@ #include "draw_pipe.h" +/** Approx number of new tokens for instructions in pstip_transform_inst() */ +#define NUM_NEW_TOKENS 50 + /** * Subclass of pipe_shader_state to carry extra fragment shader info. @@ -172,12 +175,7 @@ pstip_transform_immed(struct tgsi_transform_context *ctx, static int free_bit(uint bitfield) { - int i; - for (i = 0; i < 32; i++) { - if ((bitfield & (1 << i)) == 0) - return i; - } - return -1; + return ffs(~bitfield) - 1; } @@ -333,11 +331,10 @@ generate_pstip_fs(struct pstip_stage *pstip) /*struct draw_context *draw = pstip->stage.draw;*/ struct pipe_shader_state pstip_fs; struct pstip_transform_context transform; - -#define MAX 1000 + const uint newLen = tgsi_num_tokens(orig_fs->tokens) + NUM_NEW_TOKENS; pstip_fs = *orig_fs; /* copy to init */ - pstip_fs.tokens = MALLOC(sizeof(struct tgsi_token) * MAX); + pstip_fs.tokens = tgsi_alloc_tokens(newLen); if (pstip_fs.tokens == NULL) return FALSE; @@ -352,7 +349,7 @@ generate_pstip_fs(struct pstip_stage *pstip) tgsi_transform_shader(orig_fs->tokens, (struct tgsi_token *) pstip_fs.tokens, - MAX, &transform.base); + newLen, &transform.base); #if 0 /* DEBUG */ tgsi_dump(orig_fs->tokens, 0); diff --git a/src/gallium/auxiliary/draw/draw_private.h b/src/gallium/auxiliary/draw/draw_private.h index 6a7190e975..69466d8749 100644 --- a/src/gallium/auxiliary/draw/draw_private.h +++ b/src/gallium/auxiliary/draw/draw_private.h @@ -280,8 +280,8 @@ void draw_gs_destroy( struct draw_context *draw ); /******************************************************************************* * Common shading code: */ -int draw_current_shader_outputs(struct draw_context *draw); -int draw_current_shader_position_output(struct draw_context *draw); +uint draw_current_shader_outputs(const struct draw_context *draw); +uint draw_current_shader_position_output(const struct draw_context *draw); /******************************************************************************* * Vertex processing (was passthrough) code: diff --git a/src/gallium/auxiliary/tgsi/tgsi_exec.c b/src/gallium/auxiliary/tgsi/tgsi_exec.c index fbb9aa0e63..f7a1bb74a9 100644 --- a/src/gallium/auxiliary/tgsi/tgsi_exec.c +++ b/src/gallium/auxiliary/tgsi/tgsi_exec.c @@ -264,6 +264,12 @@ static void micro_rcp(union tgsi_exec_channel *dst, const union tgsi_exec_channel *src) { +#if 0 /* for debugging */ + assert(src->f[0] != 0.0f); + assert(src->f[1] != 0.0f); + assert(src->f[2] != 0.0f); + assert(src->f[3] != 0.0f); +#endif dst->f[0] = 1.0f / src->f[0]; dst->f[1] = 1.0f / src->f[1]; dst->f[2] = 1.0f / src->f[2]; @@ -284,6 +290,12 @@ static void micro_rsq(union tgsi_exec_channel *dst, const union tgsi_exec_channel *src) { +#if 0 /* for debugging */ + assert(src->f[0] != 0.0f); + assert(src->f[1] != 0.0f); + assert(src->f[2] != 0.0f); + assert(src->f[3] != 0.0f); +#endif dst->f[0] = 1.0f / sqrtf(fabsf(src->f[0])); dst->f[1] = 1.0f / sqrtf(fabsf(src->f[1])); dst->f[2] = 1.0f / sqrtf(fabsf(src->f[2])); @@ -450,12 +462,20 @@ static const union tgsi_exec_channel ZeroVec = { { 0.0, 0.0, 0.0, 0.0 } }; -#define CHECK_INF_OR_NAN(chan) do {\ - assert(!util_is_inf_or_nan((chan)->f[0]));\ - assert(!util_is_inf_or_nan((chan)->f[1]));\ - assert(!util_is_inf_or_nan((chan)->f[2]));\ - assert(!util_is_inf_or_nan((chan)->f[3]));\ - } while (0) +/** + * Assert that none of the float values in 'chan' are infinite or NaN. + * NaN and Inf may occur normally during program execution and should + * not lead to crashes, etc. But when debugging, it's helpful to catch + * them. + */ +static INLINE void +check_inf_or_nan(const union tgsi_exec_channel *chan) +{ + assert(!util_is_inf_or_nan((chan)->f[0])); + assert(!util_is_inf_or_nan((chan)->f[1])); + assert(!util_is_inf_or_nan((chan)->f[2])); + assert(!util_is_inf_or_nan((chan)->f[3])); +} #ifdef DEBUG @@ -1219,8 +1239,9 @@ store_dest(struct tgsi_exec_machine *mach, int offset = 0; /* indirection offset */ int index; - if (dst_datatype == TGSI_EXEC_DATA_FLOAT) { - CHECK_INF_OR_NAN(chan); + /* for debugging */ + if (0 && dst_datatype == TGSI_EXEC_DATA_FLOAT) { + check_inf_or_nan(chan); } /* There is an extra source register that indirectly subscripts @@ -1478,7 +1499,7 @@ emit_primitive(struct tgsi_exec_machine *mach) } /* - * Fetch a four texture samples using STR texture coordinates. + * Fetch four texture samples using STR texture coordinates. */ static void fetch_texel( struct tgsi_sampler *sampler, diff --git a/src/gallium/auxiliary/tgsi/tgsi_parse.c b/src/gallium/auxiliary/tgsi/tgsi_parse.c index fd37fc3079..7e19e1fe36 100644 --- a/src/gallium/auxiliary/tgsi/tgsi_parse.c +++ b/src/gallium/auxiliary/tgsi/tgsi_parse.c @@ -284,3 +284,14 @@ tgsi_dup_tokens(const struct tgsi_token *tokens) memcpy(new_tokens, tokens, bytes); return new_tokens; } + + +/** + * Allocate memory for num_tokens tokens. + */ +struct tgsi_token * +tgsi_alloc_tokens(unsigned num_tokens) +{ + unsigned bytes = num_tokens * sizeof(struct tgsi_token); + return (struct tgsi_token *) MALLOC(bytes); +} diff --git a/src/gallium/auxiliary/tgsi/tgsi_parse.h b/src/gallium/auxiliary/tgsi/tgsi_parse.h index 8150e3cd29..b45ccee2f6 100644 --- a/src/gallium/auxiliary/tgsi/tgsi_parse.h +++ b/src/gallium/auxiliary/tgsi/tgsi_parse.h @@ -130,6 +130,10 @@ tgsi_num_tokens(const struct tgsi_token *tokens); struct tgsi_token * tgsi_dup_tokens(const struct tgsi_token *tokens); +struct tgsi_token * +tgsi_alloc_tokens(unsigned num_tokens); + + #if defined __cplusplus } #endif diff --git a/src/gallium/auxiliary/tgsi/tgsi_ureg.c b/src/gallium/auxiliary/tgsi/tgsi_ureg.c index 60a1cb1af4..27960bac22 100644 --- a/src/gallium/auxiliary/tgsi/tgsi_ureg.c +++ b/src/gallium/auxiliary/tgsi/tgsi_ureg.c @@ -104,6 +104,8 @@ struct ureg_program struct { unsigned index; + unsigned semantic_name; + unsigned semantic_index; } gs_input[UREG_MAX_INPUT]; unsigned nr_gs_inputs; @@ -326,10 +328,14 @@ ureg_DECL_vs_input( struct ureg_program *ureg, struct ureg_src ureg_DECL_gs_input(struct ureg_program *ureg, - unsigned index) + unsigned index, + unsigned semantic_name, + unsigned semantic_index) { if (ureg->nr_gs_inputs < UREG_MAX_INPUT) { ureg->gs_input[ureg->nr_gs_inputs].index = index; + ureg->gs_input[ureg->nr_gs_inputs].semantic_name = semantic_name; + ureg->gs_input[ureg->nr_gs_inputs].semantic_index = semantic_index; ureg->nr_gs_inputs++; } else { set_bad(ureg); @@ -1252,10 +1258,12 @@ static void emit_decls( struct ureg_program *ureg ) } } else { for (i = 0; i < ureg->nr_gs_inputs; i++) { - emit_decl_range(ureg, - TGSI_FILE_INPUT, - ureg->gs_input[i].index, - 1); + emit_decl(ureg, + TGSI_FILE_INPUT, + ureg->gs_input[i].index, + ureg->gs_input[i].semantic_name, + ureg->gs_input[i].semantic_index, + TGSI_INTERPOLATE_CONSTANT); } } diff --git a/src/gallium/auxiliary/tgsi/tgsi_ureg.h b/src/gallium/auxiliary/tgsi/tgsi_ureg.h index 6198ca3464..6be66d0694 100644 --- a/src/gallium/auxiliary/tgsi/tgsi_ureg.h +++ b/src/gallium/auxiliary/tgsi/tgsi_ureg.h @@ -161,7 +161,9 @@ ureg_DECL_vs_input( struct ureg_program *, struct ureg_src ureg_DECL_gs_input(struct ureg_program *, - unsigned index); + unsigned index, + unsigned semantic_name, + unsigned semantic_index); struct ureg_src ureg_DECL_system_value(struct ureg_program *, diff --git a/src/gallium/docs/source/conf.py b/src/gallium/docs/source/conf.py index 9b0c86babd..59c19ed98d 100644 --- a/src/gallium/docs/source/conf.py +++ b/src/gallium/docs/source/conf.py @@ -16,13 +16,13 @@ import sys, os # If extensions (or modules to document with autodoc) are in another directory, # add these directories to sys.path here. If the directory is relative to the # documentation root, use os.path.abspath to make it absolute, like shown here. -#sys.path.append(os.path.abspath('.')) +sys.path.append(os.path.abspath('exts')) # -- General configuration ----------------------------------------------------- # Add any Sphinx extension module names here, as strings. They can be extensions # coming with Sphinx (named 'sphinx.ext.*') or your custom ones. -extensions = ['sphinx.ext.pngmath'] +extensions = ['sphinx.ext.pngmath', 'tgsi'] # Add any paths that contain templates here, relative to this directory. templates_path = ['_templates'] diff --git a/src/gallium/docs/source/exts/tgsi.py b/src/gallium/docs/source/exts/tgsi.py new file mode 100644 index 0000000000..e92cd5c4d1 --- /dev/null +++ b/src/gallium/docs/source/exts/tgsi.py @@ -0,0 +1,17 @@ +# tgsi.py +# Sphinx extension providing formatting for TGSI opcodes +# (c) Corbin Simpson 2010 + +import docutils.nodes +import sphinx.addnodes + +def parse_opcode(env, sig, signode): + opcode, desc = sig.split("-", 1) + opcode = opcode.strip().upper() + desc = " (%s)" % desc.strip() + signode += sphinx.addnodes.desc_name(opcode, opcode) + signode += sphinx.addnodes.desc_annotation(desc, desc) + return opcode + +def setup(app): + app.add_description_unit("opcode", "opcode", "%s (TGSI opcode)", parse_opcode) diff --git a/src/gallium/docs/source/screen.rst b/src/gallium/docs/source/screen.rst index 3e57a282fd..55a4c6990d 100644 --- a/src/gallium/docs/source/screen.rst +++ b/src/gallium/docs/source/screen.rst @@ -227,7 +227,7 @@ buffer_map Map a buffer into memory. -**usage** is a bitmask of :ref:`PIPE_TEXTURE_USAGE` flags. +**usage** is a bitmask of :ref:`PIPE_BUFFER_USAGE` flags. Returns a pointer to the map, or NULL if the mapping failed. @@ -248,7 +248,7 @@ Flush a range of mapped memory into a buffer. The buffer must have been mapped with ``PIPE_BUFFER_USAGE_FLUSH_EXPLICIT``. -**usage** is a bitmask of :ref:`PIPE_TEXTURE_USAGE` flags. +**usage** is a bitmask of :ref:`PIPE_BUFFER_USAGE` flags. buffer_unmap ^^^^^^^^^^^^ diff --git a/src/gallium/docs/source/tgsi.rst b/src/gallium/docs/source/tgsi.rst index 3e702ceeda..5478d86667 100644 --- a/src/gallium/docs/source/tgsi.rst +++ b/src/gallium/docs/source/tgsi.rst @@ -6,6 +6,23 @@ for describing shaders. Since Gallium is inherently shaderful, shaders are an important part of the API. TGSI is the only intermediate representation used by all drivers. +Basics +------ + +All TGSI instructions, known as *opcodes*, operate on arbitrary-precision +floating-point four-component vectors. An opcode may have up to one +destination register, known as *dst*, and between zero and three source +registers, called *src0* through *src2*, or simply *src* if there is only +one. + +Some instructions, like :opcode:`I2F`, permit re-interpretation of vector +components as integers. Other instructions permit using registers as +two-component vectors with double precision; see :ref:`Double Opcodes`. + +When an instruction has a scalar result, the result is usually copied into +each of the components of *dst*. When this happens, the result is said to be +*replicated* to *dst*. :opcode:`RCP` is one such instruction. + Instruction Set --------------- @@ -13,7 +30,7 @@ From GL_NV_vertex_program ^^^^^^^^^^^^^^^^^^^^^^^^^ -ARL - Address Register Load +.. opcode:: ARL - Address Register Load .. math:: @@ -26,7 +43,7 @@ ARL - Address Register Load dst.w = \lfloor src.w\rfloor -MOV - Move +.. opcode:: MOV - Move .. math:: @@ -39,7 +56,7 @@ MOV - Move dst.w = src.w -LIT - Light Coefficients +.. opcode:: LIT - Light Coefficients .. math:: @@ -52,33 +69,25 @@ LIT - Light Coefficients dst.w = 1 -RCP - Reciprocal - -.. math:: +.. opcode:: RCP - Reciprocal - dst.x = \frac{1}{src.x} +This instruction replicates its result. - dst.y = \frac{1}{src.x} +.. math:: - dst.z = \frac{1}{src.x} + dst = \frac{1}{src.x} - dst.w = \frac{1}{src.x} +.. opcode:: RSQ - Reciprocal Square Root -RSQ - Reciprocal Square Root +This instruction replicates its result. .. math:: - dst.x = \frac{1}{\sqrt{|src.x|}} - - dst.y = \frac{1}{\sqrt{|src.x|}} - - dst.z = \frac{1}{\sqrt{|src.x|}} + dst = \frac{1}{\sqrt{|src.x|}} - dst.w = \frac{1}{\sqrt{|src.x|}} - -EXP - Approximate Exponential Base 2 +.. opcode:: EXP - Approximate Exponential Base 2 .. math:: @@ -91,7 +100,7 @@ EXP - Approximate Exponential Base 2 dst.w = 1 -LOG - Approximate Logarithm Base 2 +.. opcode:: LOG - Approximate Logarithm Base 2 .. math:: @@ -104,7 +113,7 @@ LOG - Approximate Logarithm Base 2 dst.w = 1 -MUL - Multiply +.. opcode:: MUL - Multiply .. math:: @@ -117,7 +126,7 @@ MUL - Multiply dst.w = src0.w \times src1.w -ADD - Add +.. opcode:: ADD - Add .. math:: @@ -130,33 +139,25 @@ ADD - Add dst.w = src0.w + src1.w -DP3 - 3-component Dot Product - -.. math:: +.. opcode:: DP3 - 3-component Dot Product - dst.x = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z +This instruction replicates its result. - dst.y = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z +.. math:: - dst.z = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + dst = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z - dst.w = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z +.. opcode:: DP4 - 4-component Dot Product -DP4 - 4-component Dot Product +This instruction replicates its result. .. math:: - dst.x = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src0.w \times src1.w - - dst.y = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src0.w \times src1.w + dst = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src0.w \times src1.w - dst.z = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src0.w \times src1.w - dst.w = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src0.w \times src1.w - - -DST - Distance Vector +.. opcode:: DST - Distance Vector .. math:: @@ -169,7 +170,7 @@ DST - Distance Vector dst.w = src1.w -MIN - Minimum +.. opcode:: MIN - Minimum .. math:: @@ -182,7 +183,7 @@ MIN - Minimum dst.w = min(src0.w, src1.w) -MAX - Maximum +.. opcode:: MAX - Maximum .. math:: @@ -195,7 +196,7 @@ MAX - Maximum dst.w = max(src0.w, src1.w) -SLT - Set On Less Than +.. opcode:: SLT - Set On Less Than .. math:: @@ -208,7 +209,7 @@ SLT - Set On Less Than dst.w = (src0.w < src1.w) ? 1 : 0 -SGE - Set On Greater Equal Than +.. opcode:: SGE - Set On Greater Equal Than .. math:: @@ -221,7 +222,7 @@ SGE - Set On Greater Equal Than dst.w = (src0.w >= src1.w) ? 1 : 0 -MAD - Multiply And Add +.. opcode:: MAD - Multiply And Add .. math:: @@ -234,7 +235,7 @@ MAD - Multiply And Add dst.w = src0.w \times src1.w + src2.w -SUB - Subtract +.. opcode:: SUB - Subtract .. math:: @@ -247,7 +248,7 @@ SUB - Subtract dst.w = src0.w - src1.w -LRP - Linear Interpolate +.. opcode:: LRP - Linear Interpolate .. math:: @@ -260,7 +261,7 @@ LRP - Linear Interpolate dst.w = src0.w \times src1.w + (1 - src0.w) \times src2.w -CND - Condition +.. opcode:: CND - Condition .. math:: @@ -273,7 +274,7 @@ CND - Condition dst.w = (src2.w > 0.5) ? src0.w : src1.w -DP2A - 2-component Dot Product And Add +.. opcode:: DP2A - 2-component Dot Product And Add .. math:: @@ -286,7 +287,7 @@ DP2A - 2-component Dot Product And Add dst.w = src0.x \times src1.x + src0.y \times src1.y + src2.x -FRAC - Fraction +.. opcode:: FRAC - Fraction .. math:: @@ -299,7 +300,7 @@ FRAC - Fraction dst.w = src.w - \lfloor src.w\rfloor -CLAMP - Clamp +.. opcode:: CLAMP - Clamp .. math:: @@ -312,9 +313,9 @@ CLAMP - Clamp dst.w = clamp(src0.w, src1.w, src2.w) -FLR - Floor +.. opcode:: FLR - Floor -This is identical to ARL. +This is identical to :opcode:`ARL`. .. math:: @@ -327,7 +328,7 @@ This is identical to ARL. dst.w = \lfloor src.w\rfloor -ROUND - Round +.. opcode:: ROUND - Round .. math:: @@ -340,45 +341,33 @@ ROUND - Round dst.w = round(src.w) -EX2 - Exponential Base 2 +.. opcode:: EX2 - Exponential Base 2 -.. math:: +This instruction replicates its result. - dst.x = 2^{src.x} - - dst.y = 2^{src.x} +.. math:: - dst.z = 2^{src.x} + dst = 2^{src.x} - dst.w = 2^{src.x} +.. opcode:: LG2 - Logarithm Base 2 -LG2 - Logarithm Base 2 +This instruction replicates its result. .. math:: - dst.x = \log_2{src.x} - - dst.y = \log_2{src.x} - - dst.z = \log_2{src.x} + dst = \log_2{src.x} - dst.w = \log_2{src.x} +.. opcode:: POW - Power -POW - Power +This instruction replicates its result. .. math:: - dst.x = src0.x^{src1.x} + dst = src0.x^{src1.x} - dst.y = src0.x^{src1.x} - - dst.z = src0.x^{src1.x} - - dst.w = src0.x^{src1.x} - -XPD - Cross Product +.. opcode:: XPD - Cross Product .. math:: @@ -391,7 +380,7 @@ XPD - Cross Product dst.w = 1 -ABS - Absolute +.. opcode:: ABS - Absolute .. math:: @@ -404,48 +393,36 @@ ABS - Absolute dst.w = |src.w| -RCC - Reciprocal Clamped +.. opcode:: RCC - Reciprocal Clamped + +This instruction replicates its result. XXX cleanup on aisle three .. math:: - dst.x = (1 / src.x) > 0 ? clamp(1 / src.x, 5.42101e-020, 1.884467e+019) : clamp(1 / src.x, -1.884467e+019, -5.42101e-020) - - dst.y = (1 / src.x) > 0 ? clamp(1 / src.x, 5.42101e-020, 1.884467e+019) : clamp(1 / src.x, -1.884467e+019, -5.42101e-020) - - dst.z = (1 / src.x) > 0 ? clamp(1 / src.x, 5.42101e-020, 1.884467e+019) : clamp(1 / src.x, -1.884467e+019, -5.42101e-020) + dst = (1 / src.x) > 0 ? clamp(1 / src.x, 5.42101e-020, 1.884467e+019) : clamp(1 / src.x, -1.884467e+019, -5.42101e-020) - dst.w = (1 / src.x) > 0 ? clamp(1 / src.x, 5.42101e-020, 1.884467e+019) : clamp(1 / src.x, -1.884467e+019, -5.42101e-020) +.. opcode:: DPH - Homogeneous Dot Product -DPH - Homogeneous Dot Product +This instruction replicates its result. .. math:: - dst.x = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src1.w + dst = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src1.w - dst.y = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src1.w - dst.z = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src1.w +.. opcode:: COS - Cosine - dst.w = src0.x \times src1.x + src0.y \times src1.y + src0.z \times src1.z + src1.w - - -COS - Cosine +This instruction replicates its result. .. math:: - dst.x = \cos{src.x} - - dst.y = \cos{src.x} - - dst.z = \cos{src.x} + dst = \cos{src.x} - dst.w = \cos{src.x} - -DDX - Derivative Relative To X +.. opcode:: DDX - Derivative Relative To X .. math:: @@ -458,7 +435,7 @@ DDX - Derivative Relative To X dst.w = partialx(src.w) -DDY - Derivative Relative To Y +.. opcode:: DDY - Derivative Relative To Y .. math:: @@ -471,32 +448,32 @@ DDY - Derivative Relative To Y dst.w = partialy(src.w) -KILP - Predicated Discard +.. opcode:: KILP - Predicated Discard discard -PK2H - Pack Two 16-bit Floats +.. opcode:: PK2H - Pack Two 16-bit Floats TBD -PK2US - Pack Two Unsigned 16-bit Scalars +.. opcode:: PK2US - Pack Two Unsigned 16-bit Scalars TBD -PK4B - Pack Four Signed 8-bit Scalars +.. opcode:: PK4B - Pack Four Signed 8-bit Scalars TBD -PK4UB - Pack Four Unsigned 8-bit Scalars +.. opcode:: PK4UB - Pack Four Unsigned 8-bit Scalars TBD -RFL - Reflection Vector +.. opcode:: RFL - Reflection Vector .. math:: @@ -508,10 +485,12 @@ RFL - Reflection Vector dst.w = 1 -Considered for removal. +.. note:: + + Considered for removal. -SEQ - Set On Equal +.. opcode:: SEQ - Set On Equal .. math:: @@ -524,21 +503,20 @@ SEQ - Set On Equal dst.w = (src0.w == src1.w) ? 1 : 0 -SFL - Set On False +.. opcode:: SFL - Set On False -.. math:: +This instruction replicates its result. - dst.x = 0 +.. math:: - dst.y = 0 + dst = 0 - dst.z = 0 +.. note:: - dst.w = 0 + Considered for removal. -Considered for removal. -SGT - Set On Greater Than +.. opcode:: SGT - Set On Greater Than .. math:: @@ -551,20 +529,16 @@ SGT - Set On Greater Than dst.w = (src0.w > src1.w) ? 1 : 0 -SIN - Sine +.. opcode:: SIN - Sine -.. math:: +This instruction replicates its result. - dst.x = \sin{src.x} - - dst.y = \sin{src.x} - - dst.z = \sin{src.x} +.. math:: - dst.w = \sin{src.x} + dst = \sin{src.x} -SLE - Set On Less Equal Than +.. opcode:: SLE - Set On Less Equal Than .. math:: @@ -577,7 +551,7 @@ SLE - Set On Less Equal Than dst.w = (src0.w <= src1.w) ? 1 : 0 -SNE - Set On Not Equal +.. opcode:: SNE - Set On Not Equal .. math:: @@ -590,59 +564,63 @@ SNE - Set On Not Equal dst.w = (src0.w != src1.w) ? 1 : 0 -STR - Set On True +.. opcode:: STR - Set On True -.. math:: +This instruction replicates its result. - dst.x = 1 - - dst.y = 1 - - dst.z = 1 +.. math:: - dst.w = 1 + dst = 1 -TEX - Texture Lookup +.. opcode:: TEX - Texture Lookup TBD -TXD - Texture Lookup with Derivatives +.. opcode:: TXD - Texture Lookup with Derivatives TBD -TXP - Projective Texture Lookup +.. opcode:: TXP - Projective Texture Lookup TBD -UP2H - Unpack Two 16-Bit Floats +.. opcode:: UP2H - Unpack Two 16-Bit Floats TBD - Considered for removal. +.. note:: -UP2US - Unpack Two Unsigned 16-Bit Scalars + Considered for removal. + +.. opcode:: UP2US - Unpack Two Unsigned 16-Bit Scalars TBD - Considered for removal. +.. note:: + + Considered for removal. -UP4B - Unpack Four Signed 8-Bit Values +.. opcode:: UP4B - Unpack Four Signed 8-Bit Values TBD - Considered for removal. +.. note:: + + Considered for removal. -UP4UB - Unpack Four Unsigned 8-Bit Scalars +.. opcode:: UP4UB - Unpack Four Unsigned 8-Bit Scalars TBD - Considered for removal. +.. note:: -X2D - 2D Coordinate Transformation + Considered for removal. + +.. opcode:: X2D - 2D Coordinate Transformation .. math:: @@ -654,20 +632,24 @@ X2D - 2D Coordinate Transformation dst.w = src0.y + src1.x \times src2.z + src1.y \times src2.w -Considered for removal. +.. note:: + + Considered for removal. From GL_NV_vertex_program2 ^^^^^^^^^^^^^^^^^^^^^^^^^^ -ARA - Address Register Add +.. opcode:: ARA - Address Register Add TBD - Considered for removal. +.. note:: -ARR - Address Register Load With Round + Considered for removal. + +.. opcode:: ARR - Address Register Load With Round .. math:: @@ -680,26 +662,28 @@ ARR - Address Register Load With Round dst.w = round(src.w) -BRA - Branch +.. opcode:: BRA - Branch pc = target - Considered for removal. +.. note:: + + Considered for removal. -CAL - Subroutine Call +.. opcode:: CAL - Subroutine Call push(pc) pc = target -RET - Subroutine Call Return +.. opcode:: RET - Subroutine Call Return pc = pop() Potential restrictions: * Only occurs at end of function. -SSG - Set Sign +.. opcode:: SSG - Set Sign .. math:: @@ -712,7 +696,7 @@ SSG - Set Sign dst.w = (src.w > 0) ? 1 : (src.w < 0) ? -1 : 0 -CMP - Compare +.. opcode:: CMP - Compare .. math:: @@ -725,7 +709,7 @@ CMP - Compare dst.w = (src0.w < 0) ? src1.w : src2.w -KIL - Conditional Discard +.. opcode:: KIL - Conditional Discard .. math:: @@ -734,7 +718,7 @@ KIL - Conditional Discard endif -SCS - Sine Cosine +.. opcode:: SCS - Sine Cosine .. math:: @@ -747,12 +731,12 @@ SCS - Sine Cosine dst.y = 1 -TXB - Texture Lookup With Bias +.. opcode:: TXB - Texture Lookup With Bias TBD -NRM - 3-component Vector Normalise +.. opcode:: NRM - 3-component Vector Normalise .. math:: @@ -765,7 +749,7 @@ NRM - 3-component Vector Normalise dst.w = 1 -DIV - Divide +.. opcode:: DIV - Divide .. math:: @@ -778,35 +762,31 @@ DIV - Divide dst.w = \frac{src0.w}{src1.w} -DP2 - 2-component Dot Product - -.. math:: - - dst.x = src0.x \times src1.x + src0.y \times src1.y +.. opcode:: DP2 - 2-component Dot Product - dst.y = src0.x \times src1.x + src0.y \times src1.y +This instruction replicates its result. - dst.z = src0.x \times src1.x + src0.y \times src1.y +.. math:: - dst.w = src0.x \times src1.x + src0.y \times src1.y + dst = src0.x \times src1.x + src0.y \times src1.y -TXL - Texture Lookup With LOD +.. opcode:: TXL - Texture Lookup With LOD TBD -BRK - Break +.. opcode:: BRK - Break TBD -IF - If +.. opcode:: IF - If TBD -BGNFOR - Begin a For-Loop +.. opcode:: BGNFOR - Begin a For-Loop dst.x = floor(src.x) dst.y = floor(src.y) @@ -819,25 +799,31 @@ BGNFOR - Begin a For-Loop Note: The destination must be a loop register. The source must be a constant register. - Considered for cleanup / removal. +.. note:: + Considered for cleanup. -REP - Repeat +.. note:: + + Considered for removal. + + +.. opcode:: REP - Repeat TBD -ELSE - Else +.. opcode:: ELSE - Else TBD -ENDIF - End If +.. opcode:: ENDIF - End If TBD -ENDFOR - End a For-Loop +.. opcode:: ENDFOR - End a For-Loop dst.x = dst.x + dst.z dst.y = dst.y - 1.0 @@ -848,30 +834,48 @@ ENDFOR - End a For-Loop Note: The destination must be a loop register. - Considered for cleanup / removal. +.. note:: + + Considered for cleanup. + +.. note:: -ENDREP - End Repeat + Considered for removal. + +.. opcode:: ENDREP - End Repeat TBD -PUSHA - Push Address Register On Stack +.. opcode:: PUSHA - Push Address Register On Stack push(src.x) push(src.y) push(src.z) push(src.w) - Considered for cleanup / removal. +.. note:: + + Considered for cleanup. + +.. note:: -POPA - Pop Address Register From Stack + Considered for removal. + +.. opcode:: POPA - Pop Address Register From Stack dst.w = pop() dst.z = pop() dst.y = pop() dst.x = pop() - Considered for cleanup / removal. +.. note:: + + Considered for cleanup. + +.. note:: + + Considered for removal. From GL_NV_gpu_program4 @@ -879,7 +883,7 @@ From GL_NV_gpu_program4 Support for these opcodes indicated by a special pipe capability bit (TBD). -CEIL - Ceiling +.. opcode:: CEIL - Ceiling .. math:: @@ -892,7 +896,7 @@ CEIL - Ceiling dst.w = \lceil src.w\rceil -I2F - Integer To Float +.. opcode:: I2F - Integer To Float .. math:: @@ -905,7 +909,7 @@ I2F - Integer To Float dst.w = (float) src.w -NOT - Bitwise Not +.. opcode:: NOT - Bitwise Not .. math:: @@ -918,7 +922,7 @@ NOT - Bitwise Not dst.w = ~src.w -TRUNC - Truncate +.. opcode:: TRUNC - Truncate .. math:: @@ -931,7 +935,7 @@ TRUNC - Truncate dst.w = trunc(src.w) -SHL - Shift Left +.. opcode:: SHL - Shift Left .. math:: @@ -944,7 +948,7 @@ SHL - Shift Left dst.w = src0.w << src1.x -SHR - Shift Right +.. opcode:: SHR - Shift Right .. math:: @@ -957,7 +961,7 @@ SHR - Shift Right dst.w = src0.w >> src1.x -AND - Bitwise And +.. opcode:: AND - Bitwise And .. math:: @@ -970,7 +974,7 @@ AND - Bitwise And dst.w = src0.w & src1.w -OR - Bitwise Or +.. opcode:: OR - Bitwise Or .. math:: @@ -983,7 +987,7 @@ OR - Bitwise Or dst.w = src0.w | src1.w -MOD - Modulus +.. opcode:: MOD - Modulus .. math:: @@ -996,7 +1000,7 @@ MOD - Modulus dst.w = src0.w \bmod src1.w -XOR - Bitwise Xor +.. opcode:: XOR - Bitwise Xor .. math:: @@ -1009,7 +1013,7 @@ XOR - Bitwise Xor dst.w = src0.w \oplus src1.w -SAD - Sum Of Absolute Differences +.. opcode:: SAD - Sum Of Absolute Differences .. math:: @@ -1022,17 +1026,17 @@ SAD - Sum Of Absolute Differences dst.w = |src0.w - src1.w| + src2.w -TXF - Texel Fetch +.. opcode:: TXF - Texel Fetch TBD -TXQ - Texture Size Query +.. opcode:: TXQ - Texture Size Query TBD -CONT - Continue +.. opcode:: CONT - Continue TBD @@ -1041,12 +1045,12 @@ From GL_NV_geometry_program4 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ -EMIT - Emit +.. opcode:: EMIT - Emit TBD -ENDPRIM - End Primitive +.. opcode:: ENDPRIM - End Primitive TBD @@ -1055,66 +1059,64 @@ From GLSL ^^^^^^^^^^ -BGNLOOP - Begin a Loop +.. opcode:: BGNLOOP - Begin a Loop TBD -BGNSUB - Begin Subroutine +.. opcode:: BGNSUB - Begin Subroutine TBD -ENDLOOP - End a Loop +.. opcode:: ENDLOOP - End a Loop TBD -ENDSUB - End Subroutine +.. opcode:: ENDSUB - End Subroutine TBD -NOP - No Operation +.. opcode:: NOP - No Operation Do nothing. -NRM4 - 4-component Vector Normalise - -.. math:: +.. opcode:: NRM4 - 4-component Vector Normalise - dst.x = \frac{src.x}{src.x \times src.x + src.y \times src.y + src.z \times src.z + src.w \times src.w} +This instruction replicates its result. - dst.y = \frac{src.y}{src.x \times src.x + src.y \times src.y + src.z \times src.z + src.w \times src.w} - - dst.z = \frac{src.z}{src.x \times src.x + src.y \times src.y + src.z \times src.z + src.w \times src.w} +.. math:: - dst.w = \frac{src.w}{src.x \times src.x + src.y \times src.y + src.z \times src.z + src.w \times src.w} + dst = \frac{src.x}{src.x \times src.x + src.y \times src.y + src.z \times src.z + src.w \times src.w} ps_2_x ^^^^^^^^^^^^ -CALLNZ - Subroutine Call If Not Zero +.. opcode:: CALLNZ - Subroutine Call If Not Zero TBD -IFC - If +.. opcode:: IFC - If TBD -BREAKC - Break Conditional +.. opcode:: BREAKC - Break Conditional TBD +.. _doubleopcodes: + Double Opcodes ^^^^^^^^^^^^^^^ -DADD - Add Double +.. opcode:: DADD - Add Double .. math:: @@ -1123,7 +1125,7 @@ DADD - Add Double dst.zw = src0.zw + src1.zw -DDIV - Divide Double +.. opcode:: DDIV - Divide Double .. math:: @@ -1131,7 +1133,7 @@ DDIV - Divide Double dst.zw = src0.zw / src1.zw -DSEQ - Set Double on Equal +.. opcode:: DSEQ - Set Double on Equal .. math:: @@ -1139,7 +1141,7 @@ DSEQ - Set Double on Equal dst.zw = src0.zw == src1.zw ? 1.0F : 0.0F -DSLT - Set Double on Less than +.. opcode:: DSLT - Set Double on Less than .. math:: @@ -1147,7 +1149,7 @@ DSLT - Set Double on Less than dst.zw = src0.zw < src1.zw ? 1.0F : 0.0F -DFRAC - Double Fraction +.. opcode:: DFRAC - Double Fraction .. math:: @@ -1156,7 +1158,7 @@ DFRAC - Double Fraction dst.zw = src.zw - \lfloor src.zw\rfloor -DFRACEXP - Convert Double Number to Fractional and Integral Components +.. opcode:: DFRACEXP - Convert Double Number to Fractional and Integral Components .. math:: @@ -1164,7 +1166,7 @@ DFRACEXP - Convert Double Number to Fractional and Integral Components dst0.zw = frexp(src.zw, dst1.zw) -DLDEXP - Multiple Double Number by Integral Power of 2 +.. opcode:: DLDEXP - Multiple Double Number by Integral Power of 2 .. math:: @@ -1172,7 +1174,7 @@ DLDEXP - Multiple Double Number by Integral Power of 2 dst.zw = ldexp(src0.zw, src1.zw) -DMIN - Minimum Double +.. opcode:: DMIN - Minimum Double .. math:: @@ -1180,7 +1182,7 @@ DMIN - Minimum Double dst.zw = min(src0.zw, src1.zw) -DMAX - Maximum Double +.. opcode:: DMAX - Maximum Double .. math:: @@ -1188,7 +1190,7 @@ DMAX - Maximum Double dst.zw = max(src0.zw, src1.zw) -DMUL - Multiply Double +.. opcode:: DMUL - Multiply Double .. math:: @@ -1197,7 +1199,7 @@ DMUL - Multiply Double dst.zw = src0.zw \times src1.zw -DMAD - Multiply And Add Doubles +.. opcode:: DMAD - Multiply And Add Doubles .. math:: @@ -1206,7 +1208,7 @@ DMAD - Multiply And Add Doubles dst.zw = src0.zw \times src1.zw + src2.zw -DRCP - Reciprocal Double +.. opcode:: DRCP - Reciprocal Double .. math:: @@ -1214,7 +1216,7 @@ DRCP - Reciprocal Double dst.zw = \frac{1}{src.zw} -DSQRT - Square root double +.. opcode:: DSQRT - Square root double .. math:: @@ -1269,20 +1271,8 @@ Keywords discard Discard fragment. - dst First destination register. - - dst0 First destination register. - pc Program counter. - src First source register. - - src0 First source register. - - src1 Second source register. - - src2 Third source register. - target Label of target instruction. @@ -1441,3 +1431,43 @@ GL_ARB_fragment_coord_conventions extension. DirectX 9 uses INTEGER. DirectX 10 uses HALF_INTEGER. + + + +Texture Sampling and Texture Formats +------------------------------------ + +This table shows how texture image components are returned as (x,y,z,w) tuples +by TGSI texture instructions, such as :opcode:`TEX`, :opcode:`TXD`, and +:opcode:`TXP`. For reference, OpenGL and Direct3D conventions are shown as +well. + ++--------------------+--------------+--------------------+--------------+ +| Texture Components | Gallium | OpenGL | Direct3D 9 | ++====================+==============+====================+==============+ +| R | XXX TBD | (r, 0, 0, 1) | (r, 1, 1, 1) | ++--------------------+--------------+--------------------+--------------+ +| RG | XXX TBD | (r, g, 0, 1) | (r, g, 1, 1) | ++--------------------+--------------+--------------------+--------------+ +| RGB | (r, g, b, 1) | (r, g, b, 1) | (r, g, b, 1) | ++--------------------+--------------+--------------------+--------------+ +| RGBA | (r, g, b, a) | (r, g, b, a) | (r, g, b, a) | ++--------------------+--------------+--------------------+--------------+ +| A | (0, 0, 0, a) | (0, 0, 0, a) | (0, 0, 0, a) | ++--------------------+--------------+--------------------+--------------+ +| L | (l, l, l, 1) | (l, l, l, 1) | (l, l, l, 1) | ++--------------------+--------------+--------------------+--------------+ +| LA | (l, l, l, a) | (l, l, l, a) | (l, l, l, a) | ++--------------------+--------------+--------------------+--------------+ +| I | (i, i, i, i) | (i, i, i, i) | N/A | ++--------------------+--------------+--------------------+--------------+ +| UV | XXX TBD | (0, 0, 0, 1) | (u, v, 1, 1) | +| | | [#envmap-bumpmap]_ | | ++--------------------+--------------+--------------------+--------------+ +| Z | XXX TBD | (z, z, z, 1) | (0, z, 0, 1) | +| | | [#depth-tex-mode]_ | | ++--------------------+--------------+--------------------+--------------+ + +.. [#envmap-bumpmap] http://www.opengl.org/registry/specs/ATI/envmap_bumpmap.txt +.. [#depth-tex-mode] the default is (z, z, z, 1) but may also be (0, 0, 0, z) + or (z, z, z, z) depending on the value of GL_DEPTH_TEXTURE_MODE. diff --git a/src/gallium/drivers/llvmpipe/lp_state_derived.c b/src/gallium/drivers/llvmpipe/lp_state_derived.c index 0155b9be50..353ae176fd 100644 --- a/src/gallium/drivers/llvmpipe/lp_state_derived.c +++ b/src/gallium/drivers/llvmpipe/lp_state_derived.c @@ -65,7 +65,7 @@ llvmpipe_get_vertex_info(struct llvmpipe_context *llvmpipe) /* compute vertex layout now */ const struct lp_fragment_shader *lpfs = llvmpipe->fs; struct vertex_info *vinfo_vbuf = &llvmpipe->vertex_info_vbuf; - const uint num = draw_current_shader_outputs(llvmpipe->draw); + const uint num = draw_num_shader_outputs(llvmpipe->draw); uint i; /* Tell draw_vbuf to simply emit the whole post-xform vertex diff --git a/src/gallium/drivers/r300/r300_fs.c b/src/gallium/drivers/r300/r300_fs.c index 60ea9c171d..39bcdc8fe6 100644 --- a/src/gallium/drivers/r300/r300_fs.c +++ b/src/gallium/drivers/r300/r300_fs.c @@ -77,17 +77,21 @@ void r300_shader_read_fs_inputs(struct tgsi_shader_info* info, static void find_output_registers(struct r300_fragment_program_compiler * compiler, struct r300_fragment_shader * fs) { - unsigned i; + unsigned i, colorbuf_count = 0; /* Mark the outputs as not present initially */ - compiler->OutputColor = fs->info.num_outputs; + compiler->OutputColor[0] = fs->info.num_outputs; + compiler->OutputColor[1] = fs->info.num_outputs; + compiler->OutputColor[2] = fs->info.num_outputs; + compiler->OutputColor[3] = fs->info.num_outputs; compiler->OutputDepth = fs->info.num_outputs; /* Now see where they really are. */ for(i = 0; i < fs->info.num_outputs; ++i) { switch(fs->info.output_semantic_name[i]) { case TGSI_SEMANTIC_COLOR: - compiler->OutputColor = i; + compiler->OutputColor[colorbuf_count] = i; + colorbuf_count++; break; case TGSI_SEMANTIC_POSITION: compiler->OutputDepth = i; diff --git a/src/gallium/drivers/r300/r300_texture.h b/src/gallium/drivers/r300/r300_texture.h index 453fb1accc..b37be26133 100644 --- a/src/gallium/drivers/r300/r300_texture.h +++ b/src/gallium/drivers/r300/r300_texture.h @@ -37,24 +37,31 @@ unsigned r300_texture_get_stride(struct r300_screen* screen, unsigned r300_texture_get_offset(struct r300_texture* tex, unsigned level, unsigned zslice, unsigned face); -/* Note the signature of R300_EASY_TX_FORMAT(A, R, G, B, FORMAT)... */ +/* Translate a pipe_format into a useful texture format for sampling. + * + * R300_EASY_TX_FORMAT swizzles the texture. + * Note the signature of R300_EASY_TX_FORMAT: + * R300_EASY_TX_FORMAT(B, G, R, A, FORMAT); + * + * The FORMAT specifies how the texture sampler will treat the texture, and + * makes available X, Y, Z, W, ZERO, and ONE for swizzling. */ static INLINE uint32_t r300_translate_texformat(enum pipe_format format) { switch (format) { /* X8 */ case PIPE_FORMAT_A8_UNORM: + return R300_EASY_TX_FORMAT(ZERO, ZERO, ZERO, X, X8); case PIPE_FORMAT_I8_UNORM: return R300_EASY_TX_FORMAT(X, X, X, X, X8); case PIPE_FORMAT_L8_UNORM: return R300_EASY_TX_FORMAT(X, X, X, ONE, X8); /* X16 */ case PIPE_FORMAT_R16_UNORM: + case PIPE_FORMAT_Z16_UNORM: return R300_EASY_TX_FORMAT(X, X, X, X, X16); case PIPE_FORMAT_R16_SNORM: return R300_EASY_TX_FORMAT(X, X, X, X, X16) | R300_TX_FORMAT_SIGNED; - case PIPE_FORMAT_Z16_UNORM: - return R300_EASY_TX_FORMAT(X, X, X, X, X16); /* Y8X8 */ case PIPE_FORMAT_A8L8_UNORM: return R300_EASY_TX_FORMAT(X, X, X, Y, Y8X8); diff --git a/src/gallium/drivers/r300/r300_tgsi_to_rc.c b/src/gallium/drivers/r300/r300_tgsi_to_rc.c index a792c2cf98..941ec17016 100644 --- a/src/gallium/drivers/r300/r300_tgsi_to_rc.c +++ b/src/gallium/drivers/r300/r300_tgsi_to_rc.c @@ -201,6 +201,8 @@ static void transform_srcreg( struct rc_src_register * dst, struct tgsi_full_src_register * src) { + unsigned i, j; + dst->File = translate_register_file(src->Register.File); dst->Index = translate_register_index(ttr, src->Register.File, src->Register.Index); dst->RelAddr = src->Register.Indirect; @@ -210,6 +212,21 @@ static void transform_srcreg( dst->Swizzle |= tgsi_util_get_full_src_register_swizzle(src, 3) << 9; dst->Abs = src->Register.Absolute; dst->Negate = src->Register.Negate ? RC_MASK_XYZW : 0; + + if (src->Register.File == TGSI_FILE_IMMEDIATE) { + for (i = 0; i < ttr->imms_to_swizzle_count; i++) { + if (ttr->imms_to_swizzle[i].index == src->Register.Index) { + dst->File = RC_FILE_TEMPORARY; + dst->Index = 0; + dst->Swizzle = 0; + for (j = 0; j < 4; j++) { + dst->Swizzle |= GET_SWZ(ttr->imms_to_swizzle[i].swizzle, + tgsi_util_get_full_src_register_swizzle(src, j)) << (j * 3); + } + break; + } + } + } } static void transform_texture(struct rc_instruction * dst, struct tgsi_instruction_texture src, @@ -277,21 +294,45 @@ static void transform_instruction(struct tgsi_to_rc * ttr, struct tgsi_full_inst &ttr->compiler->Program.ShadowSamplers); } -static void handle_immediate(struct tgsi_to_rc * ttr, struct tgsi_full_immediate * imm) +static void handle_immediate(struct tgsi_to_rc * ttr, + struct tgsi_full_immediate * imm, + unsigned index) { struct rc_constant constant; - int i; + unsigned swizzle = 0; + boolean can_swizzle = TRUE; + unsigned i; - constant.Type = RC_CONSTANT_IMMEDIATE; - constant.Size = 4; - for(i = 0; i < 4; ++i) - constant.u.Immediate[i] = imm->u[i].Float; - rc_constants_add(&ttr->compiler->Program.Constants, &constant); + for (i = 0; i < 4; i++) { + if (imm->u[i].Float == 0.0f) { + swizzle |= RC_SWIZZLE_ZERO << (i * 3); + } else if (imm->u[i].Float == 0.5f) { + swizzle |= RC_SWIZZLE_HALF << (i * 3); + } else if (imm->u[i].Float == 1.0f) { + swizzle |= RC_SWIZZLE_ONE << (i * 3); + } else { + can_swizzle = FALSE; + break; + } + } + + if (can_swizzle) { + ttr->imms_to_swizzle[ttr->imms_to_swizzle_count].index = index; + ttr->imms_to_swizzle[ttr->imms_to_swizzle_count].swizzle = swizzle; + ttr->imms_to_swizzle_count++; + } else { + constant.Type = RC_CONSTANT_IMMEDIATE; + constant.Size = 4; + for(i = 0; i < 4; ++i) + constant.u.Immediate[i] = imm->u[i].Float; + rc_constants_add(&ttr->compiler->Program.Constants, &constant); + } } void r300_tgsi_to_rc(struct tgsi_to_rc * ttr, const struct tgsi_token * tokens) { struct tgsi_parse_context parser; + unsigned imm_index = 0; int i; /* Allocate constants placeholders. @@ -308,6 +349,9 @@ void r300_tgsi_to_rc(struct tgsi_to_rc * ttr, const struct tgsi_token * tokens) ttr->immediate_offset = ttr->compiler->Program.Constants.Count; + ttr->imms_to_swizzle = malloc(ttr->info->immediate_count * sizeof(struct swizzled_imms)); + ttr->imms_to_swizzle_count = 0; + tgsi_parse_init(&parser, tokens); while (!tgsi_parse_end_of_tokens(&parser)) { @@ -317,7 +361,8 @@ void r300_tgsi_to_rc(struct tgsi_to_rc * ttr, const struct tgsi_token * tokens) case TGSI_TOKEN_TYPE_DECLARATION: break; case TGSI_TOKEN_TYPE_IMMEDIATE: - handle_immediate(ttr, &parser.FullToken.FullImmediate); + handle_immediate(ttr, &parser.FullToken.FullImmediate, imm_index); + imm_index++; break; case TGSI_TOKEN_TYPE_INSTRUCTION: transform_instruction(ttr, &parser.FullToken.FullInstruction); @@ -327,6 +372,8 @@ void r300_tgsi_to_rc(struct tgsi_to_rc * ttr, const struct tgsi_token * tokens) tgsi_parse_free(&parser); + free(ttr->imms_to_swizzle); + rc_calculate_inputs_outputs(ttr->compiler); } diff --git a/src/gallium/drivers/r300/r300_tgsi_to_rc.h b/src/gallium/drivers/r300/r300_tgsi_to_rc.h index 93e90ec6d2..39b473c7bf 100644 --- a/src/gallium/drivers/r300/r300_tgsi_to_rc.h +++ b/src/gallium/drivers/r300/r300_tgsi_to_rc.h @@ -29,11 +29,18 @@ struct tgsi_full_declaration; struct tgsi_shader_info; struct tgsi_token; +struct swizzled_imms { + unsigned index; + unsigned swizzle; +}; + struct tgsi_to_rc { struct radeon_compiler * compiler; const struct tgsi_shader_info * info; int immediate_offset; + struct swizzled_imms * imms_to_swizzle; + unsigned imms_to_swizzle_count; }; void r300_tgsi_to_rc(struct tgsi_to_rc * ttr, const struct tgsi_token * tokens); diff --git a/src/gallium/drivers/softpipe/sp_state_derived.c b/src/gallium/drivers/softpipe/sp_state_derived.c index f6856a5f69..d2eda7324c 100644 --- a/src/gallium/drivers/softpipe/sp_state_derived.c +++ b/src/gallium/drivers/softpipe/sp_state_derived.c @@ -30,7 +30,6 @@ #include "pipe/p_shader_tokens.h" #include "draw/draw_context.h" #include "draw/draw_vertex.h" -#include "draw/draw_private.h" #include "sp_context.h" #include "sp_screen.h" #include "sp_state.h" @@ -67,7 +66,7 @@ softpipe_get_vertex_info(struct softpipe_context *softpipe) /* compute vertex layout now */ const struct sp_fragment_shader *spfs = softpipe->fs; struct vertex_info *vinfo_vbuf = &softpipe->vertex_info_vbuf; - const uint num = draw_current_shader_outputs(softpipe->draw); + const uint num = draw_num_shader_outputs(softpipe->draw); uint i; /* Tell draw_vbuf to simply emit the whole post-xform vertex diff --git a/src/gallium/drivers/svga/svga_context.c b/src/gallium/drivers/svga/svga_context.c index e88ef8d8fe..c4181c3f5b 100644 --- a/src/gallium/drivers/svga/svga_context.c +++ b/src/gallium/drivers/svga/svga_context.c @@ -215,7 +215,6 @@ struct pipe_context *svga_context_create( struct pipe_screen *screen ) svga->state.hw_draw.num_views = 0; svga->dirty = ~0; - svga->state.white_fs_id = SVGA3D_INVALID_ID; LIST_INITHEAD(&svga->dirty_buffers); diff --git a/src/gallium/drivers/svga/svga_context.h b/src/gallium/drivers/svga/svga_context.h index 66259fd010..ba86256eb2 100644 --- a/src/gallium/drivers/svga/svga_context.h +++ b/src/gallium/drivers/svga/svga_context.h @@ -327,10 +327,6 @@ struct svga_context unsigned texture_timestamp; - /* Internally generated shaders: - */ - unsigned white_fs_id; - /* */ struct svga_sw_state sw; diff --git a/src/gallium/drivers/svga/svga_state_fs.c b/src/gallium/drivers/svga/svga_state_fs.c index adc7120217..2973444d0a 100644 --- a/src/gallium/drivers/svga/svga_state_fs.c +++ b/src/gallium/drivers/svga/svga_state_fs.c @@ -108,70 +108,6 @@ fail: return ret; } -/* The blend workaround for simulating logicop xor behaviour requires - * that the incoming fragment color be white. This change achieves - * that by hooking up a hard-wired fragment shader that just emits - * color 1,1,1,1 - * - * This is a slightly incomplete solution as it assumes that the - * actual bound shader has no other effects beyond generating a - * fragment color. In particular shaders containing TEXKIL and/or - * depth-write will not have the correct behaviour, nor will those - * expecting to use alphatest. - * - * These are avoidable issues, but they are not much worse than the - * unavoidable ones associated with this technique, so it's not clear - * how much effort should be expended trying to resolve them - the - * ultimate result will still not be correct in most cases. - * - * Shader below was generated with: - * SVGA_DEBUG=tgsi ./mesa/progs/fp/fp-tri white.txt - */ -static int emit_white_fs( struct svga_context *svga ) -{ - int ret = PIPE_ERROR; - - /* ps_3_0 - * def c0, 1.000000, 0.000000, 0.000000, 1.000000 - * mov oC0, c0.x - * end - */ - static const unsigned white_tokens[] = { - 0xffff0300, - 0x05000051, - 0xa00f0000, - 0x3f800000, - 0x00000000, - 0x00000000, - 0x3f800000, - 0x02000001, - 0x800f0800, - 0xa0000000, - 0x0000ffff, - }; - - assert(SVGA3D_INVALID_ID == UTIL_BITMASK_INVALID_INDEX); - svga->state.white_fs_id = util_bitmask_add(svga->fs_bm); - if(svga->state.white_fs_id == SVGA3D_INVALID_ID) - goto no_fs_id; - - ret = SVGA3D_DefineShader(svga->swc, - svga->state.white_fs_id, - SVGA3D_SHADERTYPE_PS, - white_tokens, - sizeof(white_tokens)); - if (ret) - goto no_definition; - - return 0; - -no_definition: - util_bitmask_clear(svga->fs_bm, svga->state.white_fs_id); - svga->state.white_fs_id = SVGA3D_INVALID_ID; -no_fs_id: - return ret; -} - /* SVGA_NEW_TEXTURE_BINDING * SVGA_NEW_RAST @@ -199,6 +135,23 @@ static int make_fs_key( const struct svga_context *svga, PIPE_WINDING_CW); } + /* The blend workaround for simulating logicop xor behaviour + * requires that the incoming fragment color be white. This change + * achieves that by creating a varient of the current fragment + * shader that overrides all output colors with 1,1,1,1 + * + * This will work for most shaders, including those containing + * TEXKIL and/or depth-write. However, it will break on the + * combination of xor-logicop plus alphatest. + * + * Ultimately, we could implement alphatest in the shader using + * texkil prior to overriding the outgoing fragment color. + * + * SVGA_NEW_BLEND + */ + if (svga->curr.blend->need_white_fragments) { + key->white_fragments = 1; + } /* XXX: want to limit this to the textures that the shader actually * refers to. @@ -238,40 +191,29 @@ static int emit_hw_fs( struct svga_context *svga, unsigned id = SVGA3D_INVALID_ID; int ret = 0; + struct svga_fragment_shader *fs = svga->curr.fs; + struct svga_fs_compile_key key; + /* SVGA_NEW_BLEND + * SVGA_NEW_TEXTURE_BINDING + * SVGA_NEW_RAST + * SVGA_NEW_NEED_SWTNL + * SVGA_NEW_SAMPLER */ - if (svga->curr.blend->need_white_fragments) { - if (svga->state.white_fs_id == SVGA3D_INVALID_ID) { - ret = emit_white_fs( svga ); - if (ret) - return ret; - } - id = svga->state.white_fs_id; - } - else { - struct svga_fragment_shader *fs = svga->curr.fs; - struct svga_fs_compile_key key; - - /* SVGA_NEW_TEXTURE_BINDING - * SVGA_NEW_RAST - * SVGA_NEW_NEED_SWTNL - * SVGA_NEW_SAMPLER - */ - ret = make_fs_key( svga, &key ); + ret = make_fs_key( svga, &key ); + if (ret) + return ret; + + result = search_fs_key( fs, &key ); + if (!result) { + ret = compile_fs( svga, fs, &key, &result ); if (ret) return ret; - - result = search_fs_key( fs, &key ); - if (!result) { - ret = compile_fs( svga, fs, &key, &result ); - if (ret) - return ret; - } - - assert (result); - id = result->id; } + assert (result); + id = result->id; + assert(id != SVGA3D_INVALID_ID); if (result != svga->state.hw_draw.fs) { diff --git a/src/gallium/drivers/svga/svga_tgsi.h b/src/gallium/drivers/svga/svga_tgsi.h index 737a2213af..063c9cf422 100644 --- a/src/gallium/drivers/svga/svga_tgsi.h +++ b/src/gallium/drivers/svga/svga_tgsi.h @@ -49,6 +49,7 @@ struct svga_fs_compile_key { unsigned light_twoside:1; unsigned front_cw:1; + unsigned white_fragments:1; unsigned num_textures:8; unsigned num_unnormalized_coords:8; struct { diff --git a/src/gallium/drivers/svga/svga_tgsi_decl_sm30.c b/src/gallium/drivers/svga/svga_tgsi_decl_sm30.c index 43fc0d3235..73102a72a8 100644 --- a/src/gallium/drivers/svga/svga_tgsi_decl_sm30.c +++ b/src/gallium/drivers/svga/svga_tgsi_decl_sm30.c @@ -194,8 +194,19 @@ static boolean ps30_output( struct svga_shader_emitter *emit, switch (semantic.Name) { case TGSI_SEMANTIC_COLOR: - emit->output_map[idx] = dst_register( SVGA3DREG_COLOROUT, - semantic.Index ); + if (emit->unit == PIPE_SHADER_FRAGMENT && + emit->key.fkey.white_fragments) { + + emit->output_map[idx] = dst_register( SVGA3DREG_TEMP, + emit->nr_hw_temp++ ); + emit->temp_col[idx] = emit->output_map[idx]; + emit->true_col[idx] = dst_register( SVGA3DREG_COLOROUT, + semantic.Index ); + } + else { + emit->output_map[idx] = dst_register( SVGA3DREG_COLOROUT, + semantic.Index ); + } break; case TGSI_SEMANTIC_POSITION: emit->output_map[idx] = dst_register( SVGA3DREG_TEMP, diff --git a/src/gallium/drivers/svga/svga_tgsi_emit.h b/src/gallium/drivers/svga/svga_tgsi_emit.h index 2557824293..e8f75485d5 100644 --- a/src/gallium/drivers/svga/svga_tgsi_emit.h +++ b/src/gallium/drivers/svga/svga_tgsi_emit.h @@ -79,6 +79,8 @@ struct svga_shader_emitter int ps30_input_count; + int dynamic_branching_level; + boolean in_main_func; boolean created_zero_immediate; @@ -199,6 +201,23 @@ static INLINE boolean emit_op3( struct svga_shader_emitter *emit, } +static INLINE boolean emit_op4( struct svga_shader_emitter *emit, + SVGA3dShaderInstToken inst, + SVGA3dShaderDestToken dest, + struct src_register src0, + struct src_register src1, + struct src_register src2, + struct src_register src3) +{ + return (emit_instruction( emit, inst ) && + emit_dst( emit, dest ) && + emit_src( emit, src0 ) && + emit_src( emit, src1 ) && + emit_src( emit, src2 ) && + emit_src( emit, src3 )); +} + + #define TRANSLATE_SWIZZLE(x,y,z,w) ((x) | ((y) << 2) | ((z) << 4) | ((w) << 6)) #define SWIZZLE_XYZW \ TRANSLATE_SWIZZLE(TGSI_SWIZZLE_X,TGSI_SWIZZLE_Y,TGSI_SWIZZLE_Z,TGSI_SWIZZLE_W) diff --git a/src/gallium/drivers/svga/svga_tgsi_insn.c b/src/gallium/drivers/svga/svga_tgsi_insn.c index dc5eb8fc60..be821e9821 100644 --- a/src/gallium/drivers/svga/svga_tgsi_insn.c +++ b/src/gallium/drivers/svga/svga_tgsi_insn.c @@ -46,8 +46,6 @@ translate_opcode( case TGSI_OPCODE_ABS: return SVGA3DOP_ABS; case TGSI_OPCODE_ADD: return SVGA3DOP_ADD; case TGSI_OPCODE_BREAKC: return SVGA3DOP_BREAKC; - case TGSI_OPCODE_DDX: return SVGA3DOP_DSX; - case TGSI_OPCODE_DDY: return SVGA3DOP_DSY; case TGSI_OPCODE_DP2A: return SVGA3DOP_DP2ADD; case TGSI_OPCODE_DP3: return SVGA3DOP_DP3; case TGSI_OPCODE_DP4: return SVGA3DOP_DP4; @@ -415,6 +413,88 @@ static boolean submit_op3( struct svga_shader_emitter *emit, } + + +/* SVGA shaders may not refer to >1 constant register in a single + * instruction. This function checks for that usage and inserts a + * move to temporary if detected. + */ +static boolean submit_op4( struct svga_shader_emitter *emit, + SVGA3dShaderInstToken inst, + SVGA3dShaderDestToken dest, + struct src_register src0, + struct src_register src1, + struct src_register src2, + struct src_register src3) +{ + SVGA3dShaderDestToken temp0; + SVGA3dShaderDestToken temp3; + boolean need_temp0 = FALSE; + boolean need_temp3 = FALSE; + SVGA3dShaderRegType type0, type1, type2, type3; + + temp0.value = 0; + temp3.value = 0; + type0 = SVGA3dShaderGetRegType( src0.base.value ); + type1 = SVGA3dShaderGetRegType( src1.base.value ); + type2 = SVGA3dShaderGetRegType( src2.base.value ); + type3 = SVGA3dShaderGetRegType( src2.base.value ); + + /* Make life a little easier - this is only used by the TXD + * instruction which is guaranteed not to have a constant/input reg + * in one slot at least: + */ + assert(type1 == SVGA3DREG_SAMPLER); + + if (type0 == SVGA3DREG_CONST && + ((type3 == SVGA3DREG_CONST && src0.base.num != src3.base.num) || + (type2 == SVGA3DREG_CONST && src0.base.num != src2.base.num))) + need_temp0 = TRUE; + + if (type3 == SVGA3DREG_CONST && + (type2 == SVGA3DREG_CONST && src3.base.num != src2.base.num)) + need_temp3 = TRUE; + + if (type0 == SVGA3DREG_INPUT && + ((type3 == SVGA3DREG_INPUT && src0.base.num != src3.base.num) || + (type2 == SVGA3DREG_INPUT && src0.base.num != src2.base.num))) + need_temp0 = TRUE; + + if (type3 == SVGA3DREG_INPUT && + (type2 == SVGA3DREG_INPUT && src3.base.num != src2.base.num)) + need_temp3 = TRUE; + + if (need_temp0) + { + temp0 = get_temp( emit ); + + if (!emit_op1( emit, inst_token( SVGA3DOP_MOV ), temp0, src0 )) + return FALSE; + + src0 = src( temp0 ); + } + + if (need_temp3) + { + temp3 = get_temp( emit ); + + if (!emit_op1( emit, inst_token( SVGA3DOP_MOV ), temp3, src3 )) + return FALSE; + + src3 = src( temp3 ); + } + + if (!emit_op4( emit, inst, dest, src0, src1, src2, src3 )) + return FALSE; + + if (need_temp3) + release_temp( emit, temp3 ); + if (need_temp0) + release_temp( emit, temp0 ); + return TRUE; +} + + static boolean emit_def_const( struct svga_shader_emitter *emit, SVGA3dShaderConstType type, unsigned idx, @@ -660,6 +740,8 @@ static boolean emit_if(struct svga_shader_emitter *emit, if_token.control = SVGA3DOPCOMPC_NE; zero = scalar(zero, TGSI_SWIZZLE_X); + emit->dynamic_branching_level++; + return (emit_instruction( emit, if_token ) && emit_src( emit, src ) && emit_src( emit, zero ) ); @@ -668,6 +750,8 @@ static boolean emit_if(struct svga_shader_emitter *emit, static boolean emit_endif(struct svga_shader_emitter *emit, const struct tgsi_full_instruction *insn) { + emit->dynamic_branching_level--; + return (emit_instruction( emit, inst_token( SVGA3DOP_ENDIF ))); } @@ -1011,10 +1095,10 @@ static boolean emit_kilp(struct svga_shader_emitter *emit, { SVGA3dShaderInstToken inst; SVGA3dShaderDestToken temp; - struct src_register one = get_zero_immediate( emit ); + struct src_register one = scalar( get_zero_immediate( emit ), + TGSI_SWIZZLE_W ); inst = inst_token( SVGA3DOP_TEXKILL ); - one = scalar( one, TGSI_SWIZZLE_W ); /* texkill doesn't allow negation on the operand so lets move * negation of {1} to a temp register */ @@ -1169,41 +1253,79 @@ static boolean emit_tex2(struct svga_shader_emitter *emit, SVGA3dShaderDestToken dst ) { SVGA3dShaderInstToken inst; - struct src_register src0; - struct src_register src1; - + struct src_register texcoord; + struct src_register sampler; + SVGA3dShaderDestToken tmp; + inst.value = 0; - inst.op = SVGA3DOP_TEX; switch (insn->Instruction.Opcode) { case TGSI_OPCODE_TEX: + inst.op = SVGA3DOP_TEX; break; case TGSI_OPCODE_TXP: + inst.op = SVGA3DOP_TEX; inst.control = SVGA3DOPCONT_PROJECT; break; case TGSI_OPCODE_TXB: + inst.op = SVGA3DOP_TEX; inst.control = SVGA3DOPCONT_BIAS; break; + case TGSI_OPCODE_TXL: + inst.op = SVGA3DOP_TEXLDL; + break; default: assert(0); return FALSE; } - src0 = translate_src_register( emit, &insn->Src[0] ); - src1 = translate_src_register( emit, &insn->Src[1] ); + texcoord = translate_src_register( emit, &insn->Src[0] ); + sampler = translate_src_register( emit, &insn->Src[1] ); - if (emit->key.fkey.tex[src1.base.num].unnormalized) { - struct src_register wh = get_tex_dimensions( emit, src1.base.num ); - SVGA3dShaderDestToken tmp = get_temp( emit ); + if (emit->key.fkey.tex[sampler.base.num].unnormalized || + emit->dynamic_branching_level > 0) + tmp = get_temp( emit ); + + /* Can't do mipmapping inside dynamic branch constructs. Force LOD + * zero in that case. + */ + if (emit->dynamic_branching_level > 0 && + inst.op == SVGA3DOP_TEX && + SVGA3dShaderGetRegType(texcoord.base.value) == SVGA3DREG_TEMP) { + struct src_register zero = get_zero_immediate( emit ); + + /* MOV tmp, texcoord */ + if (!submit_op1( emit, + inst_token( SVGA3DOP_MOV ), + tmp, + texcoord )) + return FALSE; + + /* MOV tmp.w, zero */ + if (!submit_op1( emit, + inst_token( SVGA3DOP_MOV ), + writemask( tmp, TGSI_WRITEMASK_W ), + scalar( zero, TGSI_SWIZZLE_X ))) + return FALSE; + + texcoord = src( tmp ); + inst.op = SVGA3DOP_TEXLDL; + } + + /* Explicit normalization of texcoords: + */ + if (emit->key.fkey.tex[sampler.base.num].unnormalized) { + struct src_register wh = get_tex_dimensions( emit, sampler.base.num ); /* MUL tmp, SRC0, WH */ if (!submit_op2( emit, inst_token( SVGA3DOP_MUL ), - tmp, src0, wh )) + tmp, texcoord, wh )) return FALSE; - src0 = src( tmp ); + + texcoord = src( tmp ); } - return submit_op2( emit, inst, dst, src0, src1 ); + return submit_op2( emit, inst, dst, texcoord, sampler ); } @@ -1211,31 +1333,33 @@ static boolean emit_tex2(struct svga_shader_emitter *emit, /* Translate texture instructions to SVGA3D representation. */ -static boolean emit_tex3(struct svga_shader_emitter *emit, +static boolean emit_tex4(struct svga_shader_emitter *emit, const struct tgsi_full_instruction *insn, SVGA3dShaderDestToken dst ) { SVGA3dShaderInstToken inst; - struct src_register src0; - struct src_register src1; - struct src_register src2; + struct src_register texcoord; + struct src_register ddx; + struct src_register ddy; + struct src_register sampler; + + texcoord = translate_src_register( emit, &insn->Src[0] ); + ddx = translate_src_register( emit, &insn->Src[1] ); + ddy = translate_src_register( emit, &insn->Src[2] ); + sampler = translate_src_register( emit, &insn->Src[3] ); inst.value = 0; switch (insn->Instruction.Opcode) { case TGSI_OPCODE_TXD: - inst.op = SVGA3DOP_TEXLDD; - break; - case TGSI_OPCODE_TXL: - inst.op = SVGA3DOP_TEXLDL; + inst.op = SVGA3DOP_TEXLDD; /* 4 args! */ break; + default: + assert(0); + return FALSE; } - src0 = translate_src_register( emit, &insn->Src[0] ); - src1 = translate_src_register( emit, &insn->Src[1] ); - src2 = translate_src_register( emit, &insn->Src[2] ); - - return submit_op3( emit, inst, dst, src0, src1, src2 ); + return submit_op4( emit, inst, dst, texcoord, sampler, ddx, ddy ); } @@ -1271,12 +1395,12 @@ static boolean emit_tex(struct svga_shader_emitter *emit, case TGSI_OPCODE_TEX: case TGSI_OPCODE_TXB: case TGSI_OPCODE_TXP: + case TGSI_OPCODE_TXL: if (!emit_tex2( emit, insn, tex_result )) return FALSE; break; - case TGSI_OPCODE_TXL: case TGSI_OPCODE_TXD: - if (!emit_tex3( emit, insn, tex_result )) + if (!emit_tex4( emit, insn, tex_result )) return FALSE; break; default: @@ -1330,6 +1454,8 @@ static boolean emit_bgnloop2( struct svga_shader_emitter *emit, struct src_register loop_reg = src_register( SVGA3DREG_LOOP, 0 ); struct src_register const_int = get_loop_const( emit ); + emit->dynamic_branching_level++; + return (emit_instruction( emit, inst ) && emit_src( emit, loop_reg ) && emit_src( emit, const_int ) ); @@ -1339,6 +1465,9 @@ static boolean emit_endloop2( struct svga_shader_emitter *emit, const struct tgsi_full_instruction *insn ) { SVGA3dShaderInstToken inst = inst_token( SVGA3DOP_ENDLOOP ); + + emit->dynamic_branching_level--; + return emit_instruction( emit, inst ); } @@ -1398,6 +1527,46 @@ static boolean emit_simple_instruction(struct svga_shader_emitter *emit, } } + +static boolean emit_deriv(struct svga_shader_emitter *emit, + const struct tgsi_full_instruction *insn ) +{ + if (emit->dynamic_branching_level > 0 && + insn->Src[0].Register.File == TGSI_FILE_TEMPORARY) + { + struct src_register zero = get_zero_immediate( emit ); + SVGA3dShaderDestToken dst = + translate_dst_register( emit, insn, 0 ); + + /* Deriv opcodes not valid inside dynamic branching, workaround + * by zeroing out the destination. + */ + if (!submit_op1(emit, + inst_token( SVGA3DOP_MOV ), + dst, + scalar(zero, TGSI_SWIZZLE_X))) + return FALSE; + + return TRUE; + } + else { + unsigned opcode; + + switch (insn->Instruction.Opcode) { + case TGSI_OPCODE_DDX: + opcode = SVGA3DOP_DSX; + break; + case TGSI_OPCODE_DDY: + opcode = SVGA3DOP_DSY; + break; + default: + return FALSE; + } + + return emit_simple_instruction( emit, opcode, insn ); + } +} + static boolean emit_arl(struct svga_shader_emitter *emit, const struct tgsi_full_instruction *insn) { @@ -2002,6 +2171,10 @@ static boolean svga_emit_instruction( struct svga_shader_emitter *emit, case TGSI_OPCODE_TXD: return emit_tex( emit, insn ); + case TGSI_OPCODE_DDX: + case TGSI_OPCODE_DDY: + return emit_deriv( emit, insn ); + case TGSI_OPCODE_BGNSUB: return emit_bgnsub( emit, position, insn ); @@ -2254,11 +2427,28 @@ static boolean emit_ps_postamble( struct svga_shader_emitter *emit ) for (i = 0; i < PIPE_MAX_COLOR_BUFS; i++) { if (SVGA3dShaderGetRegType(emit->true_col[i].value) != 0) { - if (!submit_op1( emit, - inst_token(SVGA3DOP_MOV), - emit->true_col[i], - src(emit->temp_col[i]) )) - return FALSE; + /* Potentially override output colors with white for XOR + * logicop workaround. + */ + if (emit->unit == PIPE_SHADER_FRAGMENT && + emit->key.fkey.white_fragments) { + + struct src_register one = scalar( get_zero_immediate( emit ), + TGSI_SWIZZLE_W ); + + if (!submit_op1( emit, + inst_token(SVGA3DOP_MOV), + emit->true_col[i], + one )) + return FALSE; + } + else { + if (!submit_op1( emit, + inst_token(SVGA3DOP_MOV), + emit->true_col[i], + src(emit->temp_col[i]) )) + return FALSE; + } } } @@ -2467,6 +2657,9 @@ needs_to_create_zero( struct svga_shader_emitter *emit ) if (emit->key.fkey.light_twoside) return TRUE; + if (emit->key.fkey.white_fragments) + return TRUE; + if (emit->emit_frontface) return TRUE; @@ -2476,6 +2669,10 @@ needs_to_create_zero( struct svga_shader_emitter *emit ) } if (emit->info.opcode_count[TGSI_OPCODE_IF] >= 1 || + emit->info.opcode_count[TGSI_OPCODE_BGNLOOP] >= 1 || + emit->info.opcode_count[TGSI_OPCODE_BGNFOR] >= 1 || + emit->info.opcode_count[TGSI_OPCODE_DDX] >= 1 || + emit->info.opcode_count[TGSI_OPCODE_DDY] >= 1 || emit->info.opcode_count[TGSI_OPCODE_SGE] >= 1 || emit->info.opcode_count[TGSI_OPCODE_SGT] >= 1 || emit->info.opcode_count[TGSI_OPCODE_SLE] >= 1 || @@ -2702,6 +2899,8 @@ boolean svga_shader_emit_instructions( struct svga_shader_emitter *emit, goto done; } + assert(emit->dynamic_branching_level == 0); + /* Need to terminate the whole shader: */ ret = emit_instruction( emit, inst_token( SVGA3DOP_END ) ); diff --git a/src/gallium/state_trackers/egl/common/egl_g3d.c b/src/gallium/state_trackers/egl/common/egl_g3d.c index 8a73e81d4a..30e2c347bd 100644 --- a/src/gallium/state_trackers/egl/common/egl_g3d.c +++ b/src/gallium/state_trackers/egl/common/egl_g3d.c @@ -1087,8 +1087,8 @@ egl_g3d_bind_tex_image(_EGLDriver *drv, _EGLDisplay *dpy, _EGLSurface *surf, EGLint buffer) { struct egl_g3d_surface *gsurf = egl_g3d_surface(surf); - _EGLContext *ctx = _eglGetAPIContext(EGL_OPENGL_ES_API); - struct egl_g3d_context *gctx = egl_g3d_context(ctx); + _EGLContext *es1 = _eglGetAPIContext(EGL_OPENGL_ES_API); + struct egl_g3d_context *gctx; enum pipe_format target_format; int target; @@ -1118,6 +1118,11 @@ egl_g3d_bind_tex_image(_EGLDriver *drv, _EGLDisplay *dpy, return _eglError(EGL_BAD_MATCH, "eglBindTexImage"); } + if (!es1) + return EGL_TRUE; + if (!gsurf->render_surface) + return EGL_FALSE; + /* flush properly if the surface is bound */ if (gsurf->base.CurrentContext) { gctx = egl_g3d_context(gsurf->base.CurrentContext); @@ -1125,14 +1130,11 @@ egl_g3d_bind_tex_image(_EGLDriver *drv, _EGLDisplay *dpy, PIPE_FLUSH_RENDER_CACHE | PIPE_FLUSH_FRAME, NULL); } - if (gctx) { - if (!gsurf->render_surface) - return EGL_FALSE; + gctx = egl_g3d_context(es1); + gctx->stapi->st_bind_texture_surface(gsurf->render_surface, + target, gsurf->base.MipmapLevel, target_format); - gctx->stapi->st_bind_texture_surface(gsurf->render_surface, - target, gsurf->base.MipmapLevel, target_format); - gsurf->base.BoundToTexture = EGL_TRUE; - } + gsurf->base.BoundToTexture = EGL_TRUE; return EGL_TRUE; } diff --git a/src/gallium/state_trackers/vega/asm_fill.h b/src/gallium/state_trackers/vega/asm_fill.h index 2f394ad6c5..27773467fa 100644 --- a/src/gallium/state_trackers/vega/asm_fill.h +++ b/src/gallium/state_trackers/vega/asm_fill.h @@ -27,166 +27,375 @@ #ifndef ASM_FILL_H #define ASM_FILL_H -static const char solid_fill_asm[] = - "MOV %s, CONST[0]\n"; - - -static const char linear_grad_asm[] = - "MOV TEMP[0].xy, IN[0]\n" - "MOV TEMP[0].z, CONST[1].yyyy\n" - "DP3 TEMP[1], CONST[2], TEMP[0]\n" - "DP3 TEMP[2], CONST[3], TEMP[0]\n" - "DP3 TEMP[3], CONST[4], TEMP[0]\n" - "RCP TEMP[3], TEMP[3]\n" - "MUL TEMP[1], TEMP[1], TEMP[3]\n" - "MUL TEMP[2], TEMP[2], TEMP[3]\n" - "MOV TEMP[4].x, TEMP[1]\n" - "MOV TEMP[4].y, TEMP[2]\n" - "MUL TEMP[0], CONST[0].yyyy, TEMP[4].yyyy\n" - "MAD TEMP[1], CONST[0].xxxx, TEMP[4].xxxx, TEMP[0]\n" - "MUL TEMP[2], TEMP[1], CONST[0].zzzz\n" - "TEX %s, TEMP[2], SAMP[0], 1D\n"; - -static const char radial_grad_asm[] = - "MOV TEMP[0].xy, IN[0]\n" - "MOV TEMP[0].z, CONST[1].yyyy\n" - "DP3 TEMP[1], CONST[2], TEMP[0]\n" - "DP3 TEMP[2], CONST[3], TEMP[0]\n" - "DP3 TEMP[3], CONST[4], TEMP[0]\n" - "RCP TEMP[3], TEMP[3]\n" - "MUL TEMP[1], TEMP[1], TEMP[3]\n" - "MUL TEMP[2], TEMP[2], TEMP[3]\n" - "MOV TEMP[5].x, TEMP[1]\n" - "MOV TEMP[5].y, TEMP[2]\n" - "MUL TEMP[0], CONST[0].yyyy, TEMP[5].yyyy\n" - "MAD TEMP[1], CONST[0].xxxx, TEMP[5].xxxx, TEMP[0]\n" - "ADD TEMP[1], TEMP[1], TEMP[1]\n" - "MUL TEMP[3], TEMP[5].yyyy, TEMP[5].yyyy\n" - "MAD TEMP[4], TEMP[5].xxxx, TEMP[5].xxxx, TEMP[3]\n" - "MOV TEMP[4], -TEMP[4]\n" - "MUL TEMP[2], CONST[0].zzzz, TEMP[4]\n" - "MUL TEMP[0], CONST[1].wwww, TEMP[2]\n" - "MUL TEMP[3], TEMP[1], TEMP[1]\n" - "SUB TEMP[2], TEMP[3], TEMP[0]\n" - "RSQ TEMP[2], |TEMP[2]|\n" - "RCP TEMP[2], TEMP[2]\n" - "SUB TEMP[1], TEMP[2], TEMP[1]\n" - "ADD TEMP[0], CONST[0].zzzz, CONST[0].zzzz\n" - "RCP TEMP[0], TEMP[0]\n" - "MUL TEMP[2], TEMP[1], TEMP[0]\n" - "TEX %s, TEMP[2], SAMP[0], 1D\n"; - -static const char pattern_asm[] = - "MOV TEMP[0].xy, IN[0]\n" - "MOV TEMP[0].z, CONST[1].yyyy\n" - "DP3 TEMP[1], CONST[2], TEMP[0]\n" - "DP3 TEMP[2], CONST[3], TEMP[0]\n" - "DP3 TEMP[3], CONST[4], TEMP[0]\n" - "RCP TEMP[3], TEMP[3]\n" - "MUL TEMP[1], TEMP[1], TEMP[3]\n" - "MUL TEMP[2], TEMP[2], TEMP[3]\n" - "MOV TEMP[4].x, TEMP[1]\n" - "MOV TEMP[4].y, TEMP[2]\n" - "RCP TEMP[0], CONST[1].zwzw\n" - "MOV TEMP[1], TEMP[4]\n" - "MUL TEMP[1].x, TEMP[1], TEMP[0]\n" - "MUL TEMP[1].y, TEMP[1], TEMP[0]\n" - "TEX %s, TEMP[1], SAMP[0], 2D\n"; - - -static const char mask_asm[] = - "TEX TEMP[1], IN[0], SAMP[1], 2D\n" - "MUL TEMP[0].w, TEMP[0].wwww, TEMP[1].wwww\n" - "MOV %s, TEMP[0]\n"; - - -static const char image_normal_asm[] = - "TEX %s, IN[1], SAMP[3], 2D\n"; - -static const char image_multiply_asm[] = - "TEX TEMP[1], IN[1], SAMP[3], 2D\n" - "MUL %s, TEMP[0], TEMP[1]\n"; - -static const char image_stencil_asm[] = - "TEX TEMP[1], IN[1], SAMP[3], 2D\n" - "MUL %s, TEMP[0], TEMP[1]\n"; - - -#define EXTENDED_BLEND_OVER \ - "SUB TEMP[3], CONST[1].yyyy, TEMP[1].wwww\n" \ - "SUB TEMP[4], CONST[1].yyyy, TEMP[0].wwww\n" \ - "MUL TEMP[3], TEMP[0], TEMP[3]\n" \ - "MUL TEMP[4], TEMP[1], TEMP[4]\n" \ - "ADD TEMP[3], TEMP[3], TEMP[4]\n" - -static const char blend_multiply_asm[] = - "TEX TEMP[1], IN[0], SAMP[2], 2D\n" - EXTENDED_BLEND_OVER - "MUL TEMP[4], TEMP[0], TEMP[1]\n" - "ADD TEMP[1], TEMP[4], TEMP[3]\n"/*result.rgb*/ - "MUL TEMP[2], TEMP[0].wwww, TEMP[1].wwww\n" - "ADD TEMP[3], TEMP[0].wwww, TEMP[1].wwww\n" - "SUB TEMP[1].w, TEMP[3], TEMP[2]\n" - "MOV %s, TEMP[1]\n"; -#if 1 -static const char blend_screen_asm[] = - "TEX TEMP[1], IN[0], SAMP[2], 2D\n" - "ADD TEMP[3], TEMP[0], TEMP[1]\n" - "MUL TEMP[2], TEMP[0], TEMP[1]\n" - "SUB %s, TEMP[3], TEMP[2]\n"; -#else -static const char blend_screen_asm[] = - "TEX TEMP[1], IN[0], SAMP[2], 2D\n" - "MOV %s, TEMP[1]\n"; -#endif - -static const char blend_darken_asm[] = - "TEX TEMP[1], IN[0], SAMP[2], 2D\n" - EXTENDED_BLEND_OVER - "MUL TEMP[4], TEMP[0], TEMP[1].wwww\n" - "MUL TEMP[5], TEMP[1], TEMP[0].wwww\n" - "MIN TEMP[4], TEMP[4], TEMP[5]\n" - "ADD TEMP[1], TEMP[3], TEMP[4]\n" - "MUL TEMP[2], TEMP[0].wwww, TEMP[1].wwww\n" - "ADD TEMP[3], TEMP[0].wwww, TEMP[1].wwww\n" - "SUB TEMP[1].w, TEMP[3], TEMP[2]\n" - "MOV %s, TEMP[1]\n"; - -static const char blend_lighten_asm[] = - "TEX TEMP[1], IN[0], SAMP[2], 2D\n" - EXTENDED_BLEND_OVER - "MUL TEMP[4], TEMP[0], TEMP[1].wwww\n" - "MUL TEMP[5], TEMP[1], TEMP[0].wwww\n" - "MAX TEMP[4], TEMP[4], TEMP[5]\n" - "ADD TEMP[1], TEMP[3], TEMP[4]\n" - "MUL TEMP[2], TEMP[0].wwww, TEMP[1].wwww\n" - "ADD TEMP[3], TEMP[0].wwww, TEMP[1].wwww\n" - "SUB TEMP[1].w, TEMP[3], TEMP[2]\n" - "MOV %s, TEMP[1]\n"; - - -static const char premultiply_asm[] = - "MUL TEMP[0].xyz, TEMP[0], TEMP[0].wwww\n"; - -static const char unpremultiply_asm[] = - "TEX TEMP[0], IN[0], SAMP[1], 2D\n"; - - -static const char color_bw_asm[] = - "ADD TEMP[1], CONST[1].yyyy, CONST[1].yyyy\n" - "RCP TEMP[2], TEMP[1]\n" - "ADD TEMP[1], CONST[1].yyyy, TEMP[2]\n" - "ADD TEMP[2].x, TEMP[0].xxxx, TEMP[0].yyyy\n" - "ADD TEMP[2].x, TEMP[0].zzzz, TEMP[0].xxxx\n" - "SGE TEMP[0].xyz, TEMP[2].xxxx, TEMP[1]\n" - "SGE TEMP[0].w, TEMP[0].wwww, TEMP[2].yyyy\n" - "MOV %s, TEMP[0]\n"; +#include "tgsi/tgsi_ureg.h" + +typedef void (* ureg_func)( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant); + +static INLINE void +solid_fill( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_MOV(ureg, *out, constant[0]); +} + +static INLINE void +linear_grad( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + + ureg_MOV(ureg, + ureg_writemask(temp[0], TGSI_WRITEMASK_XY), + in[0]); + ureg_MOV(ureg, + ureg_writemask(temp[0], TGSI_WRITEMASK_Z), + ureg_scalar(constant[1], TGSI_SWIZZLE_Y)); + ureg_DP3(ureg, temp[1], constant[2], ureg_src(temp[0])); + ureg_DP3(ureg, temp[2], constant[3], ureg_src(temp[0])); + ureg_DP3(ureg, temp[3], constant[4], ureg_src(temp[0])); + ureg_RCP(ureg, temp[3], ureg_src(temp[3])); + ureg_MUL(ureg, temp[1], ureg_src(temp[1]), ureg_src(temp[3])); + ureg_MUL(ureg, temp[2], ureg_src(temp[2]), ureg_src(temp[3])); + ureg_MOV(ureg, ureg_writemask(temp[4], TGSI_WRITEMASK_X), ureg_src(temp[1])); + ureg_MOV(ureg, ureg_writemask(temp[4], TGSI_WRITEMASK_Y), ureg_src(temp[2])); + ureg_MUL(ureg, temp[0], + ureg_scalar(constant[0], TGSI_SWIZZLE_Y), + ureg_scalar(ureg_src(temp[4]), TGSI_SWIZZLE_Y)); + ureg_MAD(ureg, temp[1], + ureg_scalar(constant[0], TGSI_SWIZZLE_X), + ureg_scalar(ureg_src(temp[4]), TGSI_SWIZZLE_X), + ureg_src(temp[0])); + ureg_MUL(ureg, temp[2], ureg_src(temp[1]), + ureg_scalar(constant[0], TGSI_SWIZZLE_Z)); + ureg_TEX(ureg, *out, TGSI_TEXTURE_1D, ureg_src(temp[2]), sampler[0]); +} + +static INLINE void +radial_grad( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + + ureg_MOV(ureg, ureg_writemask(temp[0], TGSI_WRITEMASK_XY), in[0]); + ureg_MOV(ureg, + ureg_writemask(temp[0], TGSI_WRITEMASK_Z), + ureg_scalar(constant[1], TGSI_SWIZZLE_Y)); + ureg_DP3(ureg, temp[1], constant[2], ureg_src(temp[0])); + ureg_DP3(ureg, temp[2], constant[3], ureg_src(temp[0])); + ureg_DP3(ureg, temp[3], constant[4], ureg_src(temp[0])); + ureg_RCP(ureg, temp[3], ureg_src(temp[3])); + ureg_MUL(ureg, temp[1], ureg_src(temp[1]), ureg_src(temp[3])); + ureg_MUL(ureg, temp[2], ureg_src(temp[2]), ureg_src(temp[3])); + ureg_MOV(ureg, ureg_writemask(temp[5], TGSI_WRITEMASK_X), ureg_src(temp[1])); + ureg_MOV(ureg, ureg_writemask(temp[5], TGSI_WRITEMASK_Y), ureg_src(temp[2])); + ureg_MUL(ureg, temp[0], ureg_scalar(constant[0], TGSI_SWIZZLE_Y), + ureg_scalar(ureg_src(temp[5]), TGSI_SWIZZLE_Y)); + ureg_MAD(ureg, temp[1], + ureg_scalar(constant[0], TGSI_SWIZZLE_X), + ureg_scalar(ureg_src(temp[5]), TGSI_SWIZZLE_X), ureg_src(temp[0])); + ureg_ADD(ureg, temp[1], ureg_src(temp[1]), ureg_src(temp[1])); + ureg_MUL(ureg, temp[3], + ureg_scalar(ureg_src(temp[5]), TGSI_SWIZZLE_Y), + ureg_scalar(ureg_src(temp[5]), TGSI_SWIZZLE_Y)); + ureg_MAD(ureg, temp[4], + ureg_scalar(ureg_src(temp[5]), TGSI_SWIZZLE_X), + ureg_scalar(ureg_src(temp[5]), TGSI_SWIZZLE_X), + ureg_src(temp[3])); + ureg_MOV(ureg, temp[4], ureg_negate(ureg_src(temp[4]))); + ureg_MUL(ureg, temp[2], + ureg_scalar(constant[0], TGSI_SWIZZLE_Z), + ureg_src(temp[4])); + ureg_MUL(ureg, temp[0], + ureg_scalar(constant[1], TGSI_SWIZZLE_W), + ureg_src(temp[2])); + ureg_MUL(ureg, temp[3], ureg_src(temp[1]), ureg_src(temp[1])); + + ureg_SUB(ureg, temp[2], ureg_src(temp[3]), ureg_src(temp[0])); + ureg_RSQ(ureg, temp[2], ureg_abs(ureg_src(temp[2]))); + ureg_RCP(ureg, temp[2], ureg_src(temp[2])); + ureg_SUB(ureg, temp[1], ureg_src(temp[2]), ureg_src(temp[1])); + ureg_ADD(ureg, temp[0], + ureg_scalar(constant[0], TGSI_SWIZZLE_Z), + ureg_scalar(constant[0], TGSI_SWIZZLE_Z)); + ureg_RCP(ureg, temp[0], ureg_src(temp[0])); + ureg_MUL(ureg, temp[2], ureg_src(temp[1]), ureg_src(temp[0])); + ureg_TEX(ureg, *out, TGSI_TEXTURE_1D, ureg_src(temp[2]), sampler[0]); + +} + + +static INLINE void +pattern( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_MOV(ureg, + ureg_writemask(temp[0], TGSI_WRITEMASK_XY), + in[0]); + ureg_MOV(ureg, + ureg_writemask(temp[0], TGSI_WRITEMASK_Z), + ureg_scalar(constant[1], TGSI_SWIZZLE_Y)); + ureg_DP3(ureg, temp[1], constant[2], ureg_src(temp[0])); + ureg_DP3(ureg, temp[2], constant[3], ureg_src(temp[0])); + ureg_DP3(ureg, temp[3], constant[4], ureg_src(temp[0])); + ureg_RCP(ureg, temp[3], ureg_src(temp[3])); + ureg_MUL(ureg, temp[1], ureg_src(temp[1]), ureg_src(temp[3])); + ureg_MUL(ureg, temp[2], ureg_src(temp[2]), ureg_src(temp[3])); + ureg_MOV(ureg, ureg_writemask(temp[4], TGSI_WRITEMASK_X), ureg_src(temp[1])); + ureg_MOV(ureg, ureg_writemask(temp[4], TGSI_WRITEMASK_Y), ureg_src(temp[2])); + ureg_RCP(ureg, temp[0], + ureg_swizzle(constant[1], + TGSI_SWIZZLE_Z, + TGSI_SWIZZLE_W, + TGSI_SWIZZLE_Z, + TGSI_SWIZZLE_W)); + ureg_MOV(ureg, temp[1], ureg_src(temp[4])); + ureg_MUL(ureg, + ureg_writemask(temp[1], TGSI_WRITEMASK_X), + ureg_src(temp[1]), + ureg_src(temp[0])); + ureg_MUL(ureg, + ureg_writemask(temp[1], TGSI_WRITEMASK_Y), + ureg_src(temp[1]), + ureg_src(temp[0])); + ureg_TEX(ureg, *out, TGSI_TEXTURE_2D, ureg_src(temp[1]), sampler[0]); +} + +static INLINE void +mask( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_TEX(ureg, temp[1], TGSI_TEXTURE_2D, in[0], sampler[1]); + ureg_MUL(ureg, ureg_writemask(temp[0], TGSI_WRITEMASK_W), + ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W), + ureg_scalar(ureg_src(temp[1]), TGSI_SWIZZLE_W)); + ureg_MOV(ureg, *out, ureg_src(temp[0])); +} + +static INLINE void +image_normal( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_TEX(ureg, *out, TGSI_TEXTURE_2D, in[1], sampler[3]); +} + + +static INLINE void +image_multiply( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_TEX(ureg, temp[1], TGSI_TEXTURE_2D, in[1], sampler[3]); + ureg_MUL(ureg, *out, ureg_src(temp[0]), ureg_src(temp[1])); +} + + +static INLINE void +image_stencil( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_TEX(ureg, temp[1], TGSI_TEXTURE_2D, in[1], sampler[3]); + ureg_MUL(ureg, *out, ureg_src(temp[0]), ureg_src(temp[1])); +} + +#define EXTENDED_BLENDER_OVER_FUNC \ + ureg_SUB(ureg, temp[3], \ + ureg_scalar(constant[1], TGSI_SWIZZLE_Y), \ + ureg_scalar(ureg_src(temp[1]), TGSI_SWIZZLE_W)); \ + ureg_SUB(ureg, temp[3], \ + ureg_scalar(constant[1], TGSI_SWIZZLE_Y), \ + ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W)); \ + ureg_MUL(ureg, temp[3], ureg_src(temp[0]), ureg_src(temp[3])); \ + ureg_MUL(ureg, temp[4], ureg_src(temp[1]), ureg_src(temp[4])); \ + ureg_ADD(ureg, temp[3], ureg_src(temp[3]), ureg_src(temp[4])); + + +static INLINE void +blend_multiply( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_TEX(ureg, temp[1], TGSI_TEXTURE_2D, in[0], sampler[2]); + EXTENDED_BLENDER_OVER_FUNC + ureg_MUL(ureg, temp[4], ureg_src(temp[0]), ureg_src(temp[1])); + ureg_ADD(ureg, temp[1], ureg_src(temp[4]), ureg_src(temp[3])); + + ureg_MUL(ureg, temp[2], ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W), + ureg_scalar(ureg_src(temp[1]), TGSI_SWIZZLE_W)); + ureg_ADD(ureg, temp[3], ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W), + ureg_scalar(ureg_src(temp[1]), TGSI_SWIZZLE_W)); + ureg_SUB(ureg, ureg_writemask(temp[1], TGSI_WRITEMASK_W), + ureg_src(temp[3]), ureg_src(temp[2])); + + ureg_MOV(ureg, *out, ureg_src(temp[1])); +} + +static INLINE void +blend_screen( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_TEX(ureg, temp[1], TGSI_TEXTURE_2D, in[0], sampler[2]); + ureg_ADD(ureg, temp[3], ureg_src(temp[0]), ureg_src(temp[1])); + ureg_MUL(ureg, temp[2], ureg_src(temp[0]), ureg_src(temp[1])); + ureg_SUB(ureg, *out, ureg_src(temp[3]), ureg_src(temp[2])); +} + +static INLINE void +blend_darken( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_TEX(ureg, temp[1], TGSI_TEXTURE_2D, in[0], sampler[2]); + EXTENDED_BLENDER_OVER_FUNC + ureg_MUL(ureg, temp[4], ureg_src(temp[0]), + ureg_scalar(ureg_src(temp[1]), TGSI_SWIZZLE_W)); + ureg_MUL(ureg, temp[5], ureg_src(temp[1]), + ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W)); + ureg_MIN(ureg, temp[4], ureg_src(temp[4]), ureg_src(temp[5])); + ureg_ADD(ureg, temp[1], ureg_src(temp[3]), ureg_src(temp[4])); + + ureg_MUL(ureg, temp[2], ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W), + ureg_scalar(ureg_src(temp[1]), TGSI_SWIZZLE_W)); + ureg_ADD(ureg, temp[3], ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W), + ureg_scalar(ureg_src(temp[1]), TGSI_SWIZZLE_W)); + ureg_SUB(ureg, ureg_writemask(temp[1], TGSI_WRITEMASK_W), + ureg_src(temp[3]), ureg_src(temp[2])); + + ureg_MOV(ureg, *out, ureg_src(temp[1])); +} + +static INLINE void +blend_lighten( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_TEX(ureg, temp[1], TGSI_TEXTURE_2D, in[0], sampler[2]); + EXTENDED_BLENDER_OVER_FUNC + ureg_MUL(ureg, temp[4], ureg_src(temp[0]), + ureg_scalar(ureg_src(temp[1]), TGSI_SWIZZLE_W)); + ureg_MUL(ureg, temp[5], ureg_src(temp[1]), + ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W)); + ureg_MAX(ureg, temp[4], ureg_src(temp[4]), ureg_src(temp[5])); + ureg_ADD(ureg, temp[1], ureg_src(temp[3]), ureg_src(temp[4])); + + ureg_MUL(ureg, temp[2], ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W), + ureg_scalar(ureg_src(temp[1]), TGSI_SWIZZLE_W)); + ureg_ADD(ureg, temp[3], ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W), + ureg_scalar(ureg_src(temp[1]), TGSI_SWIZZLE_W)); + ureg_SUB(ureg, ureg_writemask(temp[1], TGSI_WRITEMASK_W), + ureg_src(temp[3]), ureg_src(temp[2])); + + ureg_MOV(ureg, *out, ureg_src(temp[1])); +} + +static INLINE void +premultiply( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_MUL(ureg, + ureg_writemask(temp[0], TGSI_WRITEMASK_XYZ), + ureg_src(temp[0]), + ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W)); +} + +static INLINE void +unpremultiply( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_TEX(ureg, temp[0], TGSI_TEXTURE_2D, in[0], sampler[1]); +} + + +static INLINE void +color_bw( struct ureg_program *ureg, + struct ureg_dst *out, + struct ureg_src *in, + struct ureg_src *sampler, + struct ureg_dst *temp, + struct ureg_src *constant) +{ + ureg_ADD(ureg, temp[1], + ureg_scalar(constant[1], TGSI_SWIZZLE_Y), + ureg_scalar(constant[1], TGSI_SWIZZLE_Y)); + ureg_RCP(ureg, temp[2], ureg_src(temp[1])); + ureg_ADD(ureg, temp[1], + ureg_scalar(constant[1], TGSI_SWIZZLE_Y), + ureg_src(temp[2])); + ureg_ADD(ureg, ureg_writemask(temp[2], TGSI_WRITEMASK_X), + ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_X), + ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_Y)); + ureg_ADD(ureg, ureg_writemask(temp[2], TGSI_WRITEMASK_X), + ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_Z), + ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_X)); + ureg_SGE(ureg, + ureg_writemask(temp[0], TGSI_WRITEMASK_XYZ), + ureg_scalar(ureg_src(temp[2]), TGSI_SWIZZLE_X), + ureg_src(temp[1])); + ureg_SGE(ureg, + ureg_writemask(temp[0], TGSI_WRITEMASK_W), + ureg_scalar(ureg_src(temp[0]), TGSI_SWIZZLE_W), + ureg_scalar(ureg_src(temp[2]), TGSI_SWIZZLE_Y)); + ureg_MOV(ureg, *out, ureg_src(temp[0])); +} struct shader_asm_info { VGint id; - VGint num_tokens; - const char * txt; + ureg_func func; VGboolean needs_position; @@ -203,44 +412,44 @@ struct shader_asm_info { static const struct shader_asm_info shaders_asm[] = { /* fills */ - {VEGA_SOLID_FILL_SHADER, 40, solid_fill_asm, + {VEGA_SOLID_FILL_SHADER, solid_fill, VG_FALSE, 0, 1, 0, 0, 0, 0}, - {VEGA_LINEAR_GRADIENT_SHADER, 200, linear_grad_asm, + {VEGA_LINEAR_GRADIENT_SHADER, linear_grad, VG_TRUE, 0, 5, 0, 1, 0, 5}, - {VEGA_RADIAL_GRADIENT_SHADER, 200, radial_grad_asm, + {VEGA_RADIAL_GRADIENT_SHADER, radial_grad, VG_TRUE, 0, 5, 0, 1, 0, 6}, - {VEGA_PATTERN_SHADER, 100, pattern_asm, + {VEGA_PATTERN_SHADER, pattern, VG_TRUE, 1, 4, 0, 1, 0, 5}, /* image draw modes */ - {VEGA_IMAGE_NORMAL_SHADER, 200, image_normal_asm, + {VEGA_IMAGE_NORMAL_SHADER, image_normal, VG_TRUE, 0, 0, 3, 1, 0, 0}, - {VEGA_IMAGE_MULTIPLY_SHADER, 200, image_multiply_asm, + {VEGA_IMAGE_MULTIPLY_SHADER, image_multiply, VG_TRUE, 0, 0, 3, 1, 0, 2}, - {VEGA_IMAGE_STENCIL_SHADER, 200, image_stencil_asm, + {VEGA_IMAGE_STENCIL_SHADER, image_stencil, VG_TRUE, 0, 0, 3, 1, 0, 2}, - {VEGA_MASK_SHADER, 100, mask_asm, + {VEGA_MASK_SHADER, mask, VG_TRUE, 0, 0, 1, 1, 0, 2}, /* extra blend modes */ - {VEGA_BLEND_MULTIPLY_SHADER, 200, blend_multiply_asm, + {VEGA_BLEND_MULTIPLY_SHADER, blend_multiply, VG_TRUE, 1, 1, 2, 1, 0, 5}, - {VEGA_BLEND_SCREEN_SHADER, 200, blend_screen_asm, + {VEGA_BLEND_SCREEN_SHADER, blend_screen, VG_TRUE, 0, 0, 2, 1, 0, 4}, - {VEGA_BLEND_DARKEN_SHADER, 200, blend_darken_asm, + {VEGA_BLEND_DARKEN_SHADER, blend_darken, VG_TRUE, 1, 1, 2, 1, 0, 6}, - {VEGA_BLEND_LIGHTEN_SHADER, 200, blend_lighten_asm, + {VEGA_BLEND_LIGHTEN_SHADER, blend_lighten, VG_TRUE, 1, 1, 2, 1, 0, 6}, /* premultiply */ - {VEGA_PREMULTIPLY_SHADER, 100, premultiply_asm, + {VEGA_PREMULTIPLY_SHADER, premultiply, VG_FALSE, 0, 0, 0, 0, 0, 1}, - {VEGA_UNPREMULTIPLY_SHADER, 100, unpremultiply_asm, + {VEGA_UNPREMULTIPLY_SHADER, unpremultiply, VG_FALSE, 0, 0, 0, 0, 0, 1}, /* color transform to black and white */ - {VEGA_BW_SHADER, 150, color_bw_asm, + {VEGA_BW_SHADER, color_bw, VG_FALSE, 1, 1, 0, 0, 0, 3}, }; #endif diff --git a/src/gallium/state_trackers/vega/shaders_cache.c b/src/gallium/state_trackers/vega/shaders_cache.c index 593e60fc63..f43fe6ee4c 100644 --- a/src/gallium/state_trackers/vega/shaders_cache.c +++ b/src/gallium/state_trackers/vega/shaders_cache.c @@ -123,17 +123,23 @@ static INLINE VGint range_max(VGint max, VGint current) return MAX2(max, current); } -static void -create_preamble(char *txt, - const struct shader_asm_info *shaders[SHADER_STAGES], - int num_shaders) +static void * +combine_shaders(const struct shader_asm_info *shaders[SHADER_STAGES], int num_shaders, + struct pipe_context *pipe, + struct pipe_shader_state *shader) { VGboolean declare_input = VG_FALSE; VGint start_const = -1, end_const = 0; VGint start_temp = -1, end_temp = 0; VGint start_sampler = -1, end_sampler = 0; - VGint i; + VGint i, current_shader = 0; VGint num_consts, num_temps, num_samplers; + struct ureg_program *ureg; + struct ureg_src in[2]; + struct ureg_src *sampler = NULL; + struct ureg_src *constant = NULL; + struct ureg_dst out, *temp = NULL; + void *p = NULL; for (i = 0; i < num_shaders; ++i) { if (shaders[i]->num_consts) @@ -158,99 +164,94 @@ create_preamble(char *txt, if (start_temp < 0) start_temp = 0; if (start_sampler < 0) - start_sampler = 0; + start_sampler = 0; num_consts = end_const - start_const; num_temps = end_temp - start_temp; num_samplers = end_sampler - start_sampler; - /* end exclusive */ - --end_const; - --end_temp; - --end_sampler; - sprintf(txt, "FRAG\n"); + ureg = ureg_create(TGSI_PROCESSOR_FRAGMENT); + if (!ureg) + return NULL; if (declare_input) { - sprintf(txt + strlen(txt), "DCL IN[0], POSITION, LINEAR\n"); - sprintf(txt + strlen(txt), "DCL IN[1], GENERIC[0], PERSPECTIVE\n"); + in[0] = ureg_DECL_fs_input(ureg, + TGSI_SEMANTIC_POSITION, + 0, + TGSI_INTERPOLATE_LINEAR); + in[1] = ureg_DECL_fs_input(ureg, + TGSI_SEMANTIC_GENERIC, + 0, + TGSI_INTERPOLATE_PERSPECTIVE); } /* we always have a color output */ - sprintf(txt + strlen(txt), "DCL OUT[0], COLOR, CONSTANT\n"); - - if (num_consts > 1) - sprintf(txt + strlen(txt), "DCL CONST[%d..%d], CONSTANT\n", start_const, end_const); - else if (num_consts == 1) - sprintf(txt + strlen(txt), "DCL CONST[%d], CONSTANT\n", start_const); - - if (num_temps > 1) - sprintf(txt + strlen(txt), "DCL TEMP[%d..%d], CONSTANT\n", start_temp, end_temp); - else if (num_temps > 1) - sprintf(txt + strlen(txt), "DCL TEMP[%d], CONSTANT\n", start_temp); - - if (num_samplers > 1) - sprintf(txt + strlen(txt), "DCL SAMP[%d..%d], CONSTANT\n", start_sampler, end_sampler); - else if (num_samplers == 1) - sprintf(txt + strlen(txt), "DCL SAMP[%d], CONSTANT\n", start_sampler); -} + out = ureg_DECL_output(ureg, TGSI_SEMANTIC_COLOR, 0); -static void * -combine_shaders(const struct shader_asm_info *shaders[SHADER_STAGES], int num_shaders, - struct pipe_context *pipe, - struct pipe_shader_state *shader) -{ - char *combined_txt; - int combined_len = MAX_PREAMBLE; - int combined_tokens = 0; - int i = 0; - int current_shader = 0; - int current_len; + if (num_consts >= 1) { + constant = (struct ureg_src *) malloc(sizeof(struct ureg_src) * end_const); + for (i = start_const; i < end_const; i++) { + constant[i] = ureg_DECL_constant(ureg, i); + } - for (i = 0; i < num_shaders; ++i) { - combined_len += strlen(shaders[i]->txt); - combined_tokens += shaders[i]->num_tokens; } - /* add for the %s->TEMP[0] substitutions */ - combined_len += num_shaders * 7 /*TEMP[0]*/ + 4 /*"END\n"*/; - combined_txt = (char*)malloc(combined_len); - combined_txt[0] = '\0'; + if (num_temps >= 1) { + temp = (struct ureg_dst *) malloc(sizeof(struct ureg_dst) * end_temp); + for (i = start_temp; i < end_temp; i++) { + temp[i] = ureg_DECL_temporary(ureg); + } + } - create_preamble(combined_txt, shaders, num_shaders); + if (num_samplers >= 1) { + sampler = (struct ureg_src *) malloc(sizeof(struct ureg_src) * end_sampler); + for (i = start_sampler; i < end_sampler; i++) { + sampler[i] = ureg_DECL_sampler(ureg, i); + } + } while (current_shader < num_shaders) { - const char temp[] = "TEMP[0]"; - const char out[] = "OUT[0]"; - const char *subst = temp; - - current_len = strlen(combined_txt); - - /* if the last shader then output */ - if (current_shader + 1 == num_shaders) - subst = out; - - snprintf(combined_txt + current_len, - combined_len - current_len, - shaders[current_shader]->txt, - subst); - ++current_shader; + if ((current_shader + 1) == num_shaders) { + shaders[current_shader]->func(ureg, + &out, + in, + sampler, + temp, + constant); + } else { + shaders[current_shader]->func(ureg, + &temp[0], + in, + sampler, + temp, + constant); + } + current_shader++; } + ureg_END(ureg); - current_len = strlen(combined_txt); - snprintf(combined_txt + current_len, - combined_len - current_len, - "END\n"); + shader->tokens = ureg_finalize(ureg); + if(!shader->tokens) + return NULL; - debug_printf("Combined shader is : \n%s\n", - combined_txt); + p = pipe->create_fs_state(pipe, shader); + ureg_destroy(ureg); - shader->tokens = tokens_from_assembly( - combined_txt, combined_tokens); + if (num_temps >= 1) { + for (i = start_temp; i < end_temp; i++) { + ureg_release_temporary(ureg, temp[i]); + } + } - free(combined_txt); + if (temp) + free(temp); + if (constant) + free(constant); + if (sampler) + free(sampler); - return pipe->create_fs_state(pipe, shader); + return p; } static void * |