shader.c

Go to the documentation of this file.

/*
 * Copyright 2002-2003 Jason Edmeades
 * Copyright 2002-2003 Raphael Junqueira
 * Copyright 2004 Christian Costa
 * Copyright 2005 Oliver Stieber
 * Copyright 2006 Ivan Gyurdiev
 * Copyright 2007-2008, 2013 Stefan Dösinger for CodeWeavers
 * Copyright 2009-2011 Henri Verbeet for CodeWeavers
 *
 * This library is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * This library is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with this library; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
 */
 
#include <stdio.h>
#include <string.h>
 
#include "wined3d_private.h"
 
WINE_DEFAULT_DEBUG_CHANNEL(d3d_shader);
 
const struct wined3d_vec4 wined3d_srgb_const[] =
{
    /* pow, mul_high, sub_high, mul_low */
    {0.41666f, 1.055f, 0.055f, 12.92f},
    /* cmp */
    {0.0031308f, 0.0f, 0.0f, 0.0f},
};
 
static const char * const shader_opcode_names[] =
{
    /* WINED3DSIH_ABS                              */ "abs",
    /* WINED3DSIH_ADD                              */ "add",
    /* WINED3DSIH_AND                              */ "and",
    /* WINED3DSIH_ATOMIC_AND                       */ "atomic_and",
    /* WINED3DSIH_ATOMIC_CMP_STORE                 */ "atomic_cmp_store",
    /* WINED3DSIH_ATOMIC_IADD                      */ "atomic_iadd",
    /* WINED3DSIH_ATOMIC_IMAX                      */ "atomic_imax",
    /* WINED3DSIH_ATOMIC_IMIN                      */ "atomic_imin",
    /* WINED3DSIH_ATOMIC_OR                        */ "atomic_or",
    /* WINED3DSIH_ATOMIC_UMAX                      */ "atomic_umax",
    /* WINED3DSIH_ATOMIC_UMIN                      */ "atomic_umin",
    /* WINED3DSIH_ATOMIC_XOR                       */ "atomic_xor",
    /* WINED3DSIH_BEM                              */ "bem",
    /* WINED3DSIH_BFI                              */ "bfi",
    /* WINED3DSIH_BFREV                            */ "bfrev",
    /* WINED3DSIH_BREAK                            */ "break",
    /* WINED3DSIH_BREAKC                           */ "breakc",
    /* WINED3DSIH_BREAKP                           */ "breakp",
    /* WINED3DSIH_BUFINFO                          */ "bufinfo",
    /* WINED3DSIH_CALL                             */ "call",
    /* WINED3DSIH_CALLNZ                           */ "callnz",
    /* WINED3DSIH_CASE                             */ "case",
    /* WINED3DSIH_CMP                              */ "cmp",
    /* WINED3DSIH_CND                              */ "cnd",
    /* WINED3DSIH_CONTINUE                         */ "continue",
    /* WINED3DSIH_CONTINUEP                        */ "continuec",
    /* WINED3DSIH_COUNTBITS                        */ "countbits",
    /* WINED3DSIH_CRS                              */ "crs",
    /* WINED3DSIH_CUT                              */ "cut",
    /* WINED3DSIH_CUT_STREAM                       */ "cut_stream",
    /* WINED3DSIH_DCL                              */ "dcl",
    /* WINED3DSIH_DCL_CONSTANT_BUFFER              */ "dcl_constantBuffer",
    /* WINED3DSIH_DCL_FUNCTION_BODY                */ "dcl_function_body",
    /* WINED3DSIH_DCL_FUNCTION_TABLE               */ "dcl_function_table",
    /* WINED3DSIH_DCL_GLOBAL_FLAGS                 */ "dcl_globalFlags",
    /* WINED3DSIH_DCL_GS_INSTANCES                 */ "dcl_gs_instances",
    /* WINED3DSIH_DCL_HS_FORK_PHASE_INSTANCE_COUNT */ "dcl_hs_fork_phase_instance_count",
    /* WINED3DSIH_DCL_HS_JOIN_PHASE_INSTANCE_COUNT */ "dcl_hs_join_phase_instance_count",
    /* WINED3DSIH_DCL_HS_MAX_TESSFACTOR            */ "dcl_hs_max_tessfactor",
    /* WINED3DSIH_DCL_IMMEDIATE_CONSTANT_BUFFER    */ "dcl_immediateConstantBuffer",
    /* WINED3DSIH_DCL_INDEX_RANGE                  */ "dcl_index_range",
    /* WINED3DSIH_DCL_INDEXABLE_TEMP               */ "dcl_indexableTemp",
    /* WINED3DSIH_DCL_INPUT                        */ "dcl_input",
    /* WINED3DSIH_DCL_INPUT_CONTROL_POINT_COUNT    */ "dcl_input_control_point_count",
    /* WINED3DSIH_DCL_INPUT_PRIMITIVE              */ "dcl_inputPrimitive",
    /* WINED3DSIH_DCL_INPUT_PS                     */ "dcl_input_ps",
    /* WINED3DSIH_DCL_INPUT_PS_SGV                 */ "dcl_input_ps_sgv",
    /* WINED3DSIH_DCL_INPUT_PS_SIV                 */ "dcl_input_ps_siv",
    /* WINED3DSIH_DCL_INPUT_SGV                    */ "dcl_input_sgv",
    /* WINED3DSIH_DCL_INPUT_SIV                    */ "dcl_input_siv",
    /* WINED3DSIH_DCL_INTERFACE                    */ "dcl_interface",
    /* WINED3DSIH_DCL_OUTPUT                       */ "dcl_output",
    /* WINED3DSIH_DCL_OUTPUT_CONTROL_POINT_COUNT   */ "dcl_output_control_point_count",
    /* WINED3DSIH_DCL_OUTPUT_SIV                   */ "dcl_output_siv",
    /* WINED3DSIH_DCL_OUTPUT_TOPOLOGY              */ "dcl_outputTopology",
    /* WINED3DSIH_DCL_RESOURCE_RAW                 */ "dcl_resource_raw",
    /* WINED3DSIH_DCL_RESOURCE_STRUCTURED          */ "dcl_resource_structured",
    /* WINED3DSIH_DCL_SAMPLER                      */ "dcl_sampler",
    /* WINED3DSIH_DCL_STREAM                       */ "dcl_stream",
    /* WINED3DSIH_DCL_TEMPS                        */ "dcl_temps",
    /* WINED3DSIH_DCL_TESSELLATOR_DOMAIN           */ "dcl_tessellator_domain",
    /* WINED3DSIH_DCL_TESSELLATOR_OUTPUT_PRIMITIVE */ "dcl_tessellator_output_primitive",
    /* WINED3DSIH_DCL_TESSELLATOR_PARTITIONING     */ "dcl_tessellator_partitioning",
    /* WINED3DSIH_DCL_TGSM_RAW                     */ "dcl_tgsm_raw",
    /* WINED3DSIH_DCL_TGSM_STRUCTURED              */ "dcl_tgsm_structured",
    /* WINED3DSIH_DCL_THREAD_GROUP                 */ "dcl_thread_group",
    /* WINED3DSIH_DCL_UAV_RAW                      */ "dcl_uav_raw",
    /* WINED3DSIH_DCL_UAV_STRUCTURED               */ "dcl_uav_structured",
    /* WINED3DSIH_DCL_UAV_TYPED                    */ "dcl_uav_typed",
    /* WINED3DSIH_DCL_VERTICES_OUT                 */ "dcl_maxOutputVertexCount",
    /* WINED3DSIH_DEF                              */ "def",
    /* WINED3DSIH_DEFAULT                          */ "default",
    /* WINED3DSIH_DEFB                             */ "defb",
    /* WINED3DSIH_DEFI                             */ "defi",
    /* WINED3DSIH_DIV                              */ "div",
    /* WINED3DSIH_DP2                              */ "dp2",
    /* WINED3DSIH_DP2ADD                           */ "dp2add",
    /* WINED3DSIH_DP3                              */ "dp3",
    /* WINED3DSIH_DP4                              */ "dp4",
    /* WINED3DSIH_DST                              */ "dst",
    /* WINED3DSIH_DSX                              */ "dsx",
    /* WINED3DSIH_DSX_COARSE                       */ "deriv_rtx_coarse",
    /* WINED3DSIH_DSX_FINE                         */ "deriv_rtx_fine",
    /* WINED3DSIH_DSY                              */ "dsy",
    /* WINED3DSIH_DSY_COARSE                       */ "deriv_rty_coarse",
    /* WINED3DSIH_DSY_FINE                         */ "deriv_rty_fine",
    /* WINED3DSIH_ELSE                             */ "else",
    /* WINED3DSIH_EMIT                             */ "emit",
    /* WINED3DSIH_EMIT_STREAM                      */ "emit_stream",
    /* WINED3DSIH_ENDIF                            */ "endif",
    /* WINED3DSIH_ENDLOOP                          */ "endloop",
    /* WINED3DSIH_ENDREP                           */ "endrep",
    /* WINED3DSIH_ENDSWITCH                        */ "endswitch",
    /* WINED3DSIH_EQ                               */ "eq",
    /* WINED3DSIH_EVAL_CENTROID                    */ "eval_centroid",
    /* WINED3DSIH_EVAL_SAMPLE_INDEX                */ "eval_sample_index",
    /* WINED3DSIH_EXP                              */ "exp",
    /* WINED3DSIH_EXPP                             */ "expp",
    /* WINED3DSIH_F16TOF32                         */ "f16tof32",
    /* WINED3DSIH_F32TOF16                         */ "f32tof16",
    /* WINED3DSIH_FCALL                            */ "fcall",
    /* WINED3DSIH_FIRSTBIT_HI                      */ "firstbit_hi",
    /* WINED3DSIH_FIRSTBIT_LO                      */ "firstbit_lo",
    /* WINED3DSIH_FIRSTBIT_SHI                     */ "firstbit_shi",
    /* WINED3DSIH_FRC                              */ "frc",
    /* WINED3DSIH_FTOI                             */ "ftoi",
    /* WINED3DSIH_FTOU                             */ "ftou",
    /* WINED3DSIH_GATHER4                          */ "gather4",
    /* WINED3DSIH_GATHER4_C                        */ "gather4_c",
    /* WINED3DSIH_GATHER4_PO                       */ "gather4_po",
    /* WINED3DSIH_GATHER4_PO_C                     */ "gather4_po_c",
    /* WINED3DSIH_GE                               */ "ge",
    /* WINED3DSIH_HS_CONTROL_POINT_PHASE           */ "hs_control_point_phase",
    /* WINED3DSIH_HS_DECLS                         */ "hs_decls",
    /* WINED3DSIH_HS_FORK_PHASE                    */ "hs_fork_phase",
    /* WINED3DSIH_HS_JOIN_PHASE                    */ "hs_join_phase",
    /* WINED3DSIH_IADD                             */ "iadd",
    /* WINED3DSIH_IBFE                             */ "ibfe",
    /* WINED3DSIH_IEQ                              */ "ieq",
    /* WINED3DSIH_IF                               */ "if",
    /* WINED3DSIH_IFC                              */ "ifc",
    /* WINED3DSIH_IGE                              */ "ige",
    /* WINED3DSIH_ILT                              */ "ilt",
    /* WINED3DSIH_IMAD                             */ "imad",
    /* WINED3DSIH_IMAX                             */ "imax",
    /* WINED3DSIH_IMIN                             */ "imin",
    /* WINED3DSIH_IMM_ATOMIC_ALLOC                 */ "imm_atomic_alloc",
    /* WINED3DSIH_IMM_ATOMIC_AND                   */ "imm_atomic_and",
    /* WINED3DSIH_IMM_ATOMIC_CMP_EXCH              */ "imm_atomic_cmp_exch",
    /* WINED3DSIH_IMM_ATOMIC_CONSUME               */ "imm_atomic_consume",
    /* WINED3DSIH_IMM_ATOMIC_EXCH                  */ "imm_atomic_exch",
    /* WINED3DSIH_IMM_ATOMIC_IADD                  */ "imm_atomic_iadd",
    /* WINED3DSIH_IMM_ATOMIC_IMAX                  */ "imm_atomic_imax",
    /* WINED3DSIH_IMM_ATOMIC_IMIN                  */ "imm_atomic_imin",
    /* WINED3DSIH_IMM_ATOMIC_OR                    */ "imm_atomic_or",
    /* WINED3DSIH_IMM_ATOMIC_UMAX                  */ "imm_atomic_umax",
    /* WINED3DSIH_IMM_ATOMIC_UMIN                  */ "imm_atomic_umin",
    /* WINED3DSIH_IMM_ATOMIC_XOR                   */ "imm_atomic_xor",
    /* WINED3DSIH_IMUL                             */ "imul",
    /* WINED3DSIH_INE                              */ "ine",
    /* WINED3DSIH_INEG                             */ "ineg",
    /* WINED3DSIH_ISHL                             */ "ishl",
    /* WINED3DSIH_ISHR                             */ "ishr",
    /* WINED3DSIH_ITOF                             */ "itof",
    /* WINED3DSIH_LABEL                            */ "label",
    /* WINED3DSIH_LD                               */ "ld",
    /* WINED3DSIH_LD2DMS                           */ "ld2dms",
    /* WINED3DSIH_LD_RAW                           */ "ld_raw",
    /* WINED3DSIH_LD_STRUCTURED                    */ "ld_structured",
    /* WINED3DSIH_LD_UAV_TYPED                     */ "ld_uav_typed",
    /* WINED3DSIH_LIT                              */ "lit",
    /* WINED3DSIH_LOD                              */ "lod",
    /* WINED3DSIH_LOG                              */ "log",
    /* WINED3DSIH_LOGP                             */ "logp",
    /* WINED3DSIH_LOOP                             */ "loop",
    /* WINED3DSIH_LRP                              */ "lrp",
    /* WINED3DSIH_LT                               */ "lt",
    /* WINED3DSIH_M3x2                             */ "m3x2",
    /* WINED3DSIH_M3x3                             */ "m3x3",
    /* WINED3DSIH_M3x4                             */ "m3x4",
    /* WINED3DSIH_M4x3                             */ "m4x3",
    /* WINED3DSIH_M4x4                             */ "m4x4",
    /* WINED3DSIH_MAD                              */ "mad",
    /* WINED3DSIH_MAX                              */ "max",
    /* WINED3DSIH_MIN                              */ "min",
    /* WINED3DSIH_MOV                              */ "mov",
    /* WINED3DSIH_MOVA                             */ "mova",
    /* WINED3DSIH_MOVC                             */ "movc",
    /* WINED3DSIH_MUL                              */ "mul",
    /* WINED3DSIH_NE                               */ "ne",
    /* WINED3DSIH_NOP                              */ "nop",
    /* WINED3DSIH_NOT                              */ "not",
    /* WINED3DSIH_NRM                              */ "nrm",
    /* WINED3DSIH_OR                               */ "or",
    /* WINED3DSIH_PHASE                            */ "phase",
    /* WINED3DSIH_POW                              */ "pow",
    /* WINED3DSIH_RCP                              */ "rcp",
    /* WINED3DSIH_REP                              */ "rep",
    /* WINED3DSIH_RESINFO                          */ "resinfo",
    /* WINED3DSIH_RET                              */ "ret",
    /* WINED3DSIH_RETP                             */ "retp",
    /* WINED3DSIH_ROUND_NE                         */ "round_ne",
    /* WINED3DSIH_ROUND_NI                         */ "round_ni",
    /* WINED3DSIH_ROUND_PI                         */ "round_pi",
    /* WINED3DSIH_ROUND_Z                          */ "round_z",
    /* WINED3DSIH_RSQ                              */ "rsq",
    /* WINED3DSIH_SAMPLE                           */ "sample",
    /* WINED3DSIH_SAMPLE_B                         */ "sample_b",
    /* WINED3DSIH_SAMPLE_C                         */ "sample_c",
    /* WINED3DSIH_SAMPLE_C_LZ                      */ "sample_c_lz",
    /* WINED3DSIH_SAMPLE_GRAD                      */ "sample_d",
    /* WINED3DSIH_SAMPLE_INFO                      */ "sample_info",
    /* WINED3DSIH_SAMPLE_LOD                       */ "sample_l",
    /* WINED3DSIH_SAMPLE_POS                       */ "sample_pos",
    /* WINED3DSIH_SETP                             */ "setp",
    /* WINED3DSIH_SGE                              */ "sge",
    /* WINED3DSIH_SGN                              */ "sgn",
    /* WINED3DSIH_SINCOS                           */ "sincos",
    /* WINED3DSIH_SLT                              */ "slt",
    /* WINED3DSIH_SQRT                             */ "sqrt",
    /* WINED3DSIH_STORE_RAW                        */ "store_raw",
    /* WINED3DSIH_STORE_STRUCTURED                 */ "store_structured",
    /* WINED3DSIH_STORE_UAV_TYPED                  */ "store_uav_typed",
    /* WINED3DSIH_SUB                              */ "sub",
    /* WINED3DSIH_SWAPC                            */ "swapc",
    /* WINED3DSIH_SWITCH                           */ "switch",
    /* WINED3DSIH_SYNC                             */ "sync",
    /* WINED3DSIH_TEX                              */ "texld",
    /* WINED3DSIH_TEXBEM                           */ "texbem",
    /* WINED3DSIH_TEXBEML                          */ "texbeml",
    /* WINED3DSIH_TEXCOORD                         */ "texcrd",
    /* WINED3DSIH_TEXDEPTH                         */ "texdepth",
    /* WINED3DSIH_TEXDP3                           */ "texdp3",
    /* WINED3DSIH_TEXDP3TEX                        */ "texdp3tex",
    /* WINED3DSIH_TEXKILL                          */ "texkill",
    /* WINED3DSIH_TEXLDD                           */ "texldd",
    /* WINED3DSIH_TEXLDL                           */ "texldl",
    /* WINED3DSIH_TEXM3x2DEPTH                     */ "texm3x2depth",
    /* WINED3DSIH_TEXM3x2PAD                       */ "texm3x2pad",
    /* WINED3DSIH_TEXM3x2TEX                       */ "texm3x2tex",
    /* WINED3DSIH_TEXM3x3                          */ "texm3x3",
    /* WINED3DSIH_TEXM3x3DIFF                      */ "texm3x3diff",
    /* WINED3DSIH_TEXM3x3PAD                       */ "texm3x3pad",
    /* WINED3DSIH_TEXM3x3SPEC                      */ "texm3x3spec",
    /* WINED3DSIH_TEXM3x3TEX                       */ "texm3x3tex",
    /* WINED3DSIH_TEXM3x3VSPEC                     */ "texm3x3vspec",
    /* WINED3DSIH_TEXREG2AR                        */ "texreg2ar",
    /* WINED3DSIH_TEXREG2GB                        */ "texreg2gb",
    /* WINED3DSIH_TEXREG2RGB                       */ "texreg2rgb",
    /* WINED3DSIH_UBFE                             */ "ubfe",
    /* WINED3DSIH_UDIV                             */ "udiv",
    /* WINED3DSIH_UGE                              */ "uge",
    /* WINED3DSIH_ULT                              */ "ult",
    /* WINED3DSIH_UMAX                             */ "umax",
    /* WINED3DSIH_UMIN                             */ "umin",
    /* WINED3DSIH_UMUL                             */ "umul",
    /* WINED3DSIH_USHR                             */ "ushr",
    /* WINED3DSIH_UTOF                             */ "utof",
    /* WINED3DSIH_XOR                              */ "xor",
};
 
static const char * const semantic_names[] =
{
    /* WINED3D_DECL_USAGE_POSITION      */ "SV_POSITION",
    /* WINED3D_DECL_USAGE_BLEND_WEIGHT  */ "BLENDWEIGHT",
    /* WINED3D_DECL_USAGE_BLEND_INDICES */ "BLENDINDICES",
    /* WINED3D_DECL_USAGE_NORMAL        */ "NORMAL",
    /* WINED3D_DECL_USAGE_PSIZE         */ "PSIZE",
    /* WINED3D_DECL_USAGE_TEXCOORD      */ "TEXCOORD",
    /* WINED3D_DECL_USAGE_TANGENT       */ "TANGENT",
    /* WINED3D_DECL_USAGE_BINORMAL      */ "BINORMAL",
    /* WINED3D_DECL_USAGE_TESS_FACTOR   */ "TESSFACTOR",
    /* WINED3D_DECL_USAGE_POSITIONT     */ "POSITIONT",
    /* WINED3D_DECL_USAGE_COLOR         */ "COLOR",
    /* WINED3D_DECL_USAGE_FOG           */ "FOG",
    /* WINED3D_DECL_USAGE_DEPTH         */ "DEPTH",
    /* WINED3D_DECL_USAGE_SAMPLE        */ "SAMPLE",
};
 
const char *debug_d3dshaderinstructionhandler(enum WINED3D_SHADER_INSTRUCTION_HANDLER handler_idx)
{
    if (handler_idx >= ARRAY_SIZE(shader_opcode_names))
        return wine_dbg_sprintf("UNRECOGNIZED(%#x)", handler_idx);
 
    return shader_opcode_names[handler_idx];
}
 
enum vkd3d_shader_visibility vkd3d_shader_visibility_from_wined3d(enum wined3d_shader_type shader_type)
{
    switch (shader_type)
    {
        case WINED3D_SHADER_TYPE_VERTEX:
            return VKD3D_SHADER_VISIBILITY_VERTEX;
        case WINED3D_SHADER_TYPE_HULL:
            return VKD3D_SHADER_VISIBILITY_HULL;
        case WINED3D_SHADER_TYPE_DOMAIN:
            return VKD3D_SHADER_VISIBILITY_DOMAIN;
        case WINED3D_SHADER_TYPE_GEOMETRY:
            return VKD3D_SHADER_VISIBILITY_GEOMETRY;
        case WINED3D_SHADER_TYPE_PIXEL:
            return VKD3D_SHADER_VISIBILITY_PIXEL;
        case WINED3D_SHADER_TYPE_COMPUTE:
            return VKD3D_SHADER_VISIBILITY_COMPUTE;
        default:
            ERR("Invalid shader type %s.\n", debug_shader_type(shader_type));
            return VKD3D_SHADER_VISIBILITY_ALL;
    }
}
 
static const char *shader_semantic_name_from_usage(enum wined3d_decl_usage usage)
{
    if (usage >= ARRAY_SIZE(semantic_names))
    {
        FIXME("Unrecognized usage %#x.\n", usage);
        return "UNRECOGNIZED";
    }
 
    return semantic_names[usage];
}
 
static enum wined3d_decl_usage shader_usage_from_semantic_name(const char *name)
{
    unsigned int i;
 
    for (i = 0; i < ARRAY_SIZE(semantic_names); ++i)
    {
        if (!strcmp(name, semantic_names[i]))
            return i;
    }
 
    return ~0U;
}
 
static enum wined3d_sysval_semantic shader_sysval_semantic_from_usage(enum wined3d_decl_usage usage)
{
    switch (usage)
    {
        case WINED3D_DECL_USAGE_POSITION:
            return WINED3D_SV_POSITION;
        default:
            return 0;
    }
}
 
BOOL shader_match_semantic(const char *semantic_name, enum wined3d_decl_usage usage)
{
    return !strcmp(semantic_name, shader_semantic_name_from_usage(usage));
}
 
static void shader_signature_from_semantic(struct wined3d_shader_signature_element *e,
        const struct wined3d_shader_semantic *s)
{
    e->semantic_name = shader_semantic_name_from_usage(s->usage);
    e->semantic_idx = s->usage_idx;
    e->stream_idx = 0;
    e->sysval_semantic = shader_sysval_semantic_from_usage(s->usage);
    e->component_type = WINED3D_TYPE_FLOAT;
    e->register_idx = s->reg.reg.idx[0].offset;
    e->mask = s->reg.write_mask;
}
 
static void shader_signature_from_usage(struct wined3d_shader_signature_element *e,
        enum wined3d_decl_usage usage, UINT usage_idx, UINT reg_idx, DWORD write_mask)
{
    e->semantic_name = shader_semantic_name_from_usage(usage);
    e->semantic_idx = usage_idx;
    e->stream_idx = 0;
    e->sysval_semantic = shader_sysval_semantic_from_usage(usage);
    e->component_type = WINED3D_TYPE_FLOAT;
    e->register_idx = reg_idx;
    e->mask = write_mask;
}
 
static const struct wined3d_shader_frontend *shader_select_frontend(enum vkd3d_shader_source_type source_type)
{
    switch (source_type)
    {
        case VKD3D_SHADER_SOURCE_D3D_BYTECODE:
            return &sm1_shader_frontend;
 
        case VKD3D_SHADER_SOURCE_DXBC_TPF:
            return &sm4_shader_frontend;
 
        default:
            WARN("Invalid source type %#x specified.\n", source_type);
            return NULL;
    }
}
 
void string_buffer_clear(struct wined3d_string_buffer *buffer)
{
    buffer->buffer[0] = '\0';
    buffer->content_size = 0;
}
 
BOOL string_buffer_init(struct wined3d_string_buffer *buffer)
{
    buffer->buffer_size = 32;
    if (!(buffer->buffer = malloc(buffer->buffer_size)))
    {
        ERR("Failed to allocate shader buffer memory.\n");
        return FALSE;
    }
 
    string_buffer_clear(buffer);
    return TRUE;
}
 
void string_buffer_free(struct wined3d_string_buffer *buffer)
{
    free(buffer->buffer);
}
 
BOOL string_buffer_resize(struct wined3d_string_buffer *buffer, int rc)
{
    char *new_buffer;
    unsigned int new_buffer_size = buffer->buffer_size * 2;
 
    while (rc > 0 && (unsigned int)rc >= new_buffer_size - buffer->content_size)
        new_buffer_size *= 2;
    if (!(new_buffer = realloc(buffer->buffer, new_buffer_size)))
    {
        ERR("Failed to grow buffer.\n");
        buffer->buffer[buffer->content_size] = '\0';
        return FALSE;
    }
    buffer->buffer = new_buffer;
    buffer->buffer_size = new_buffer_size;
    return TRUE;
}
 
int shader_vaddline(struct wined3d_string_buffer *buffer, const char *format, va_list args)
{
    unsigned int rem;
    int rc;
 
    rem = buffer->buffer_size - buffer->content_size;
    rc = vsnprintf(&buffer->buffer[buffer->content_size], rem, format, args);
    if (rc < 0 /* C89 */ || (unsigned int)rc >= rem /* C99 */)
        return rc;
 
    buffer->content_size += rc;
    return 0;
}
 
int shader_addline(struct wined3d_string_buffer *buffer, const char *format, ...)
{
    va_list args;
    int ret;
 
    for (;;)
    {
        va_start(args, format);
        ret = shader_vaddline(buffer, format, args);
        va_end(args);
        if (!ret)
            return ret;
        if (!string_buffer_resize(buffer, ret))
            return -1;
    }
}
 
struct wined3d_string_buffer *string_buffer_get(struct wined3d_string_buffer_list *list)
{
    struct wined3d_string_buffer *buffer;
 
    if (list_empty(&list->list))
    {
        buffer = malloc(sizeof(*buffer));
        if (!buffer || !string_buffer_init(buffer))
        {
            ERR("Couldn't allocate buffer for temporary string.\n");
            free(buffer);
            return NULL;
        }
    }
    else
    {
        buffer = LIST_ENTRY(list_head(&list->list), struct wined3d_string_buffer, entry);
        list_remove(&buffer->entry);
    }
    string_buffer_clear(buffer);
    return buffer;
}
 
static int string_buffer_vsprintf(struct wined3d_string_buffer *buffer, const char *format, va_list args)
{
    if (!buffer)
        return 0;
    string_buffer_clear(buffer);
    return shader_vaddline(buffer, format, args);
}
 
void string_buffer_sprintf(struct wined3d_string_buffer *buffer, const char *format, ...)
{
    va_list args;
    int ret;
 
    for (;;)
    {
        va_start(args, format);
        ret = string_buffer_vsprintf(buffer, format, args);
        va_end(args);
        if (!ret)
            return;
        if (!string_buffer_resize(buffer, ret))
            return;
    }
}
 
void string_buffer_release(struct wined3d_string_buffer_list *list, struct wined3d_string_buffer *buffer)
{
    if (!buffer)
        return;
    list_add_head(&list->list, &buffer->entry);
}
 
void string_buffer_list_init(struct wined3d_string_buffer_list *list)
{
    list_init(&list->list);
}
 
void string_buffer_list_cleanup(struct wined3d_string_buffer_list *list)
{
    struct wined3d_string_buffer *buffer, *buffer_next;
 
    LIST_FOR_EACH_ENTRY_SAFE(buffer, buffer_next, &list->list, struct wined3d_string_buffer, entry)
    {
        string_buffer_free(buffer);
        free(buffer);
    }
    list_init(&list->list);
}
 
static void shader_delete_constant_list(struct list *clist)
{
    struct wined3d_shader_lconst *constant, *constant_next;
 
    LIST_FOR_EACH_ENTRY_SAFE(constant, constant_next, clist, struct wined3d_shader_lconst, entry)
        free(constant);
    list_init(clist);
}
 
static void shader_set_limits(struct wined3d_shader *shader)
{
    static const struct limits_entry
    {
        unsigned int min_version;
        unsigned int max_version;
        struct wined3d_shader_limits limits;
    }
    vs_limits[] =
    {
        /* min_version, max_version, sampler, constant_int, constant_float, constant_bool, packed_output, packed_input */
        {WINED3D_SHADER_VERSION(1, 0), WINED3D_SHADER_VERSION(1, 1), { 0,  0, 256,  0, 12,  0}},
        {WINED3D_SHADER_VERSION(2, 0), WINED3D_SHADER_VERSION(2, 1), { 0, 16, 256, 16, 12,  0}},
        /* DX10 cards on Windows advertise a D3D9 constant limit of 256
         * even though they are capable of supporting much more (GL
         * drivers advertise 1024). d3d9.dll and d3d8.dll clamp the
         * wined3d-advertised maximum. Clamp the constant limit for <= 3.0
         * shaders to 256. */
        {WINED3D_SHADER_VERSION(3, 0), WINED3D_SHADER_VERSION(3, 0), { 4, 16, 256, 16, 12,  0}},
        {WINED3D_SHADER_VERSION(4, 0), WINED3D_SHADER_VERSION(4, 0), {16,  0,   0,  0, 16,  0}},
        {WINED3D_SHADER_VERSION(4, 1), WINED3D_SHADER_VERSION(5, 0), {16,  0,   0,  0, 32,  0}},
        {0}
    },
    hs_limits[] =
    {
        /* min_version, max_version, sampler, constant_int, constant_float, constant_bool, packed_output, packet_input */
        {WINED3D_SHADER_VERSION(5, 0), WINED3D_SHADER_VERSION(5, 0), {16,  0,   0,  0, 32, 32}},
    },
    ds_limits[] =
    {
        /* min_version, max_version, sampler, constant_int, constant_float, constant_bool, packed_output, packet_input */
        {WINED3D_SHADER_VERSION(5, 0), WINED3D_SHADER_VERSION(5, 0), {16,  0,   0,  0, 32, 32}},
    },
    gs_limits[] =
    {
        /* min_version, max_version, sampler, constant_int, constant_float, constant_bool, packed_output, packed_input */
        {WINED3D_SHADER_VERSION(4, 0), WINED3D_SHADER_VERSION(4, 0), {16,  0,   0,  0, 32, 16}},
        {WINED3D_SHADER_VERSION(4, 1), WINED3D_SHADER_VERSION(5, 0), {16,  0,   0,  0, 32, 32}},
        {0}
    },
    ps_limits[] =
    {
        /* min_version, max_version, sampler, constant_int, constant_float, constant_bool, packed_output, packed_input */
        {WINED3D_SHADER_VERSION(1, 0), WINED3D_SHADER_VERSION(1, 3), { 4,  0,   8,  0,  0,  0}},
        {WINED3D_SHADER_VERSION(1, 4), WINED3D_SHADER_VERSION(1, 4), { 6,  0,   8,  0,  0,  0}},
        {WINED3D_SHADER_VERSION(2, 0), WINED3D_SHADER_VERSION(2, 0), {16,  0,  32,  0,  0,  0}},
        {WINED3D_SHADER_VERSION(2, 1), WINED3D_SHADER_VERSION(2, 1), {16, 16,  32, 16,  0,  0}},
        {WINED3D_SHADER_VERSION(3, 0), WINED3D_SHADER_VERSION(3, 0), {16, 16, 224, 16,  0, 10}},
        {WINED3D_SHADER_VERSION(4, 0), WINED3D_SHADER_VERSION(5, 0), {16,  0,   0,  0,  0, 32}},
        {0}
    },
    cs_limits[] =
    {
        /* min_version, max_version, sampler, constant_int, constant_float, constant_bool, packed_output, packed_input */
        {WINED3D_SHADER_VERSION(5, 0), WINED3D_SHADER_VERSION(5, 0), {16,  0,   0,  0,  0,  0}},
    };
    const struct limits_entry *limits_array;
    DWORD shader_version = WINED3D_SHADER_VERSION(shader->reg_maps.shader_version.major,
            shader->reg_maps.shader_version.minor);
    int i = 0;
 
    switch (shader->reg_maps.shader_version.type)
    {
        default:
            FIXME("Unexpected shader type %u found.\n", shader->reg_maps.shader_version.type);
            /* Fall-through. */
        case WINED3D_SHADER_TYPE_VERTEX:
            limits_array = vs_limits;
            break;
        case WINED3D_SHADER_TYPE_HULL:
            limits_array = hs_limits;
            break;
        case WINED3D_SHADER_TYPE_DOMAIN:
            limits_array = ds_limits;
            break;
        case WINED3D_SHADER_TYPE_GEOMETRY:
            limits_array = gs_limits;
            break;
        case WINED3D_SHADER_TYPE_PIXEL:
            limits_array = ps_limits;
            break;
        case WINED3D_SHADER_TYPE_COMPUTE:
            limits_array = cs_limits;
            break;
    }
 
    while (limits_array[i].min_version && limits_array[i].min_version <= shader_version)
    {
        if (shader_version <= limits_array[i].max_version)
        {
            shader->limits = &limits_array[i].limits;
            break;
        }
        ++i;
    }
    if (!shader->limits)
    {
        FIXME("Unexpected shader version \"%u.%u\" (shader type %u).\n",
                shader->reg_maps.shader_version.major,
                shader->reg_maps.shader_version.minor,
                shader->reg_maps.shader_version.type);
        shader->limits = &limits_array[max(0, i - 1)].limits;
    }
}
 
static BOOL shader_record_register_usage(struct wined3d_shader *shader, struct wined3d_shader_reg_maps *reg_maps,
        const struct wined3d_shader_register *reg, enum wined3d_shader_type shader_type, unsigned int constf_size)
{
    switch (reg->type)
    {
        case WINED3DSPR_TEXTURE: /* WINED3DSPR_ADDR */
            if (shader_type == WINED3D_SHADER_TYPE_PIXEL)
                reg_maps->texcoord |= 1u << reg->idx[0].offset;
            else
                reg_maps->address |= 1u << reg->idx[0].offset;
            break;
 
        case WINED3DSPR_TEMP:
            reg_maps->temporary |= 1u << reg->idx[0].offset;
            break;
 
        case WINED3DSPR_INPUT:
            if (reg->idx[0].rel_addr)
                reg_maps->input_rel_addressing = 1;
            if (shader_type == WINED3D_SHADER_TYPE_PIXEL)
            {
                /* If relative addressing is used, we must assume that all
                 * registers are used. Even if it is a construct like v3[aL],
                 * we can't assume that v0, v1 and v2 aren't read because aL
                 * can be negative. */
                if (reg->idx[0].rel_addr)
                    shader->u.ps.input_reg_used = ~0u;
                else
                    shader->u.ps.input_reg_used |= 1u << reg->idx[0].offset;
            }
            else
            {
                reg_maps->input_registers |= 1u << reg->idx[0].offset;
            }
            break;
 
        case WINED3DSPR_RASTOUT:
            if (reg->idx[0].offset == 1)
                reg_maps->fog = 1;
            if (reg->idx[0].offset == 2)
                reg_maps->point_size = 1;
            break;
 
        case WINED3DSPR_MISCTYPE:
            if (shader_type == WINED3D_SHADER_TYPE_PIXEL)
            {
                if (!reg->idx[0].offset)
                    reg_maps->vpos = 1;
                else if (reg->idx[0].offset == 1)
                    reg_maps->usesfacing = 1;
            }
            break;
 
        case WINED3DSPR_CONST:
            if (reg->idx[0].rel_addr)
            {
                if (reg->idx[0].offset < reg_maps->min_rel_offset)
                    reg_maps->min_rel_offset = reg->idx[0].offset;
                if (reg->idx[0].offset > reg_maps->max_rel_offset)
                    reg_maps->max_rel_offset = reg->idx[0].offset;
                reg_maps->usesrelconstF = TRUE;
            }
            else
            {
                if (reg->idx[0].offset >= min(shader->limits->constant_float, constf_size))
                {
                    WARN("Shader using float constant %u which is not supported.\n", reg->idx[0].offset);
                    return FALSE;
                }
                else
                {
                    wined3d_insert_bits(reg_maps->constf, reg->idx[0].offset, 1, 0x1);
                }
            }
            break;
 
        case WINED3DSPR_CONSTINT:
            if (reg->idx[0].offset >= shader->limits->constant_int)
            {
                WARN("Shader using integer constant %u which is not supported.\n", reg->idx[0].offset);
                return FALSE;
            }
            else
            {
                reg_maps->integer_constants |= (1u << reg->idx[0].offset);
            }
            break;
 
        case WINED3DSPR_CONSTBOOL:
            if (reg->idx[0].offset >= shader->limits->constant_bool)
            {
                WARN("Shader using bool constant %u which is not supported.\n", reg->idx[0].offset);
                return FALSE;
            }
            else
            {
                reg_maps->boolean_constants |= (1u << reg->idx[0].offset);
            }
            break;
 
        case WINED3DSPR_COLOROUT:
            reg_maps->rt_mask |= (1u << reg->idx[0].offset);
            break;
 
        case WINED3DSPR_OUTCONTROLPOINT:
            reg_maps->vocp = 1;
            break;
 
        case WINED3DSPR_SAMPLEMASK:
            reg_maps->sample_mask = 1;
            break;
 
        case WINED3DSPR_STENCILREF:
            reg_maps->stencil_ref = 1;
            break;
 
        default:
            TRACE("Not recording register of type %#x and [%#x][%#x].\n",
                    reg->type, reg->idx[0].offset, reg->idx[1].offset);
            break;
    }
    return TRUE;
}
 
static void shader_record_sample(struct wined3d_shader_reg_maps *reg_maps,
        unsigned int resource_idx, unsigned int sampler_idx, unsigned int bind_idx)
{
    struct wined3d_shader_sampler_map_entry *entries, *entry;
    struct wined3d_shader_sampler_map *map;
    unsigned int i;
 
    map = &reg_maps->sampler_map;
    entries = map->entries;
    for (i = 0; i < map->count; ++i)
    {
        if (entries[i].resource_idx == resource_idx && entries[i].sampler_idx == sampler_idx)
            return;
    }
 
    if (!map->size)
    {
        if (!(entries = calloc(4, sizeof(*entries))))
        {
            ERR("Failed to allocate sampler map entries.\n");
            return;
        }
        map->size = 4;
        map->entries = entries;
    }
    else if (map->count == map->size)
    {
        size_t new_size = map->size * 2;
 
        if (sizeof(*entries) * new_size <= sizeof(*entries) * map->size
                || !(entries = realloc(entries, sizeof(*entries) * new_size)))
        {
            ERR("Failed to resize sampler map entries.\n");
            return;
        }
        map->size = new_size;
        map->entries = entries;
    }
 
    entry = &entries[map->count++];
    entry->resource_idx = resource_idx;
    entry->sampler_idx = sampler_idx;
    entry->bind_idx = bind_idx;
}
 
static unsigned int get_instr_extra_regcount(enum WINED3D_SHADER_INSTRUCTION_HANDLER instr, unsigned int param)
{
    switch (instr)
    {
        case WINED3DSIH_M4x4:
        case WINED3DSIH_M3x4:
            return param == 1 ? 3 : 0;
 
        case WINED3DSIH_M4x3:
        case WINED3DSIH_M3x3:
            return param == 1 ? 2 : 0;
 
        case WINED3DSIH_M3x2:
            return param == 1 ? 1 : 0;
 
        default:
            return 0;
    }
}
 
static HRESULT shader_reg_maps_add_tgsm(struct wined3d_shader_reg_maps *reg_maps,
        unsigned int register_idx, unsigned int size, unsigned int stride)
{
    struct wined3d_shader_tgsm *tgsm;
 
    if (register_idx >= MAX_TGSM_REGISTERS)
    {
        ERR("Invalid TGSM register index %u.\n", register_idx);
        return S_OK;
    }
    if (reg_maps->shader_version.type != WINED3D_SHADER_TYPE_COMPUTE)
    {
        FIXME("TGSM declarations are allowed only in compute shaders.\n");
        return S_OK;
    }
 
    if (!wined3d_array_reserve((void **)&reg_maps->tgsm, &reg_maps->tgsm_capacity,
            register_idx + 1, sizeof(*reg_maps->tgsm)))
        return E_OUTOFMEMORY;
 
    reg_maps->tgsm_count = max(register_idx + 1, reg_maps->tgsm_count);
    tgsm = &reg_maps->tgsm[register_idx];
    tgsm->size = size;
    tgsm->stride = stride;
    return S_OK;
}
 
static HRESULT shader_record_shader_phase(struct wined3d_shader *shader,
        struct wined3d_shader_phase **current_phase, const struct wined3d_shader_instruction *ins,
        const DWORD *current_instruction_ptr, const DWORD *previous_instruction_ptr)
{
    struct wined3d_shader_phase *phase;
 
    if ((phase = *current_phase))
    {
        phase->end = previous_instruction_ptr;
        *current_phase = NULL;
    }
 
    if (shader->reg_maps.shader_version.type != WINED3D_SHADER_TYPE_HULL)
    {
        ERR("Unexpected shader type %s.\n", debug_shader_type(shader->reg_maps.shader_version.type));
        return E_FAIL;
    }
 
    switch (ins->handler_idx)
    {
        case WINED3DSIH_HS_CONTROL_POINT_PHASE:
            if (shader->u.hs.phases.control_point)
            {
                FIXME("Multiple control point phases.\n");
                free(shader->u.hs.phases.control_point);
            }
            if (!(shader->u.hs.phases.control_point = calloc(1, sizeof(*shader->u.hs.phases.control_point))))
                return E_OUTOFMEMORY;
            phase = shader->u.hs.phases.control_point;
            break;
        case WINED3DSIH_HS_FORK_PHASE:
            if (!wined3d_array_reserve((void **)&shader->u.hs.phases.fork,
                    &shader->u.hs.phases.fork_size, shader->u.hs.phases.fork_count + 1,
                    sizeof(*shader->u.hs.phases.fork)))
                return E_OUTOFMEMORY;
            phase = &shader->u.hs.phases.fork[shader->u.hs.phases.fork_count++];
            break;
        case WINED3DSIH_HS_JOIN_PHASE:
            if (!wined3d_array_reserve((void **)&shader->u.hs.phases.join,
                    &shader->u.hs.phases.join_size, shader->u.hs.phases.join_count + 1,
                    sizeof(*shader->u.hs.phases.join)))
                return E_OUTOFMEMORY;
            phase = &shader->u.hs.phases.join[shader->u.hs.phases.join_count++];
            break;
        default:
            ERR("Unexpected opcode %s.\n", debug_d3dshaderinstructionhandler(ins->handler_idx));
            return E_FAIL;
    }
 
    phase->start = current_instruction_ptr;
    *current_phase = phase;
 
    return WINED3D_OK;
}
 
static HRESULT shader_calculate_clip_or_cull_distance_mask(
        const struct wined3d_shader_signature_element *e, unsigned int *mask)
{
    /* Clip and cull distances are packed in 4 component registers. 0 and 1 are
     * the only allowed semantic indices.
     */
    if (e->semantic_idx >= WINED3D_MAX_CLIP_DISTANCES / 4)
    {
        *mask = 0;
        WARN("Invalid clip/cull distance index %u.\n", e->semantic_idx);
        return WINED3DERR_INVALIDCALL;
    }
 
    *mask = (e->mask & WINED3DSP_WRITEMASK_ALL) << (4 * e->semantic_idx);
    return WINED3D_OK;
}
 
static void wined3d_insert_interpolation_mode(uint32_t *packed_interpolation_mode,
        unsigned int register_idx, enum wined3d_shader_interpolation_mode mode)
{
    if (mode > WINED3DSIM_LINEAR_NOPERSPECTIVE_SAMPLE)
        FIXME("Unexpected interpolation mode %#x.\n", mode);
 
    wined3d_insert_bits(packed_interpolation_mode,
            register_idx * WINED3D_PACKED_INTERPOLATION_BIT_COUNT, WINED3D_PACKED_INTERPOLATION_BIT_COUNT, mode);
}
 
static HRESULT shader_scan_output_signature(struct wined3d_shader *shader)
{
    const struct wined3d_shader_signature *output_signature = &shader->output_signature;
    struct wined3d_shader_reg_maps *reg_maps = &shader->reg_maps;
    unsigned int i;
    HRESULT hr;
 
    for (i = 0; i < output_signature->element_count; ++i)
    {
        const struct wined3d_shader_signature_element *e = &output_signature->elements[i];
        unsigned int mask;
 
        reg_maps->output_registers |= 1u << e->register_idx;
        if (e->sysval_semantic == WINED3D_SV_CLIP_DISTANCE)
        {
            if (FAILED(hr = shader_calculate_clip_or_cull_distance_mask(e, &mask)))
                return hr;
            reg_maps->clip_distance_mask |= mask;
        }
        else if (e->sysval_semantic == WINED3D_SV_CULL_DISTANCE)
        {
            if (FAILED(hr = shader_calculate_clip_or_cull_distance_mask(e, &mask)))
                return hr;
            reg_maps->cull_distance_mask |= mask;
        }
        else if (e->sysval_semantic == WINED3D_SV_VIEWPORT_ARRAY_INDEX)
        {
            reg_maps->viewport_array = 1;
        }
    }
 
    return WINED3D_OK;
}
 
/* Note that this does not count the loop register as an address register. */
static HRESULT shader_get_registers_used(struct wined3d_shader *shader, DWORD constf_size)
{
    struct wined3d_shader_signature_element input_signature_elements[max(MAX_ATTRIBS, MAX_REG_INPUT)];
    struct wined3d_shader_signature_element output_signature_elements[MAX_REG_OUTPUT];
    struct wined3d_shader_signature *output_signature = &shader->output_signature;
    struct wined3d_shader_signature *input_signature = &shader->input_signature;
    struct wined3d_shader_reg_maps *reg_maps = &shader->reg_maps;
    const struct wined3d_shader_frontend *fe = shader->frontend;
    unsigned int cur_loop_depth = 0, max_loop_depth = 0;
    struct wined3d_shader_version shader_version;
    struct wined3d_shader_phase *phase = NULL;
    const DWORD *ptr, *prev_ins, *current_ins;
    void *fe_data = shader->frontend_data;
    unsigned int i;
    HRESULT hr;
 
    memset(reg_maps, 0, sizeof(*reg_maps));
    memset(input_signature_elements, 0, sizeof(input_signature_elements));
    memset(output_signature_elements, 0, sizeof(output_signature_elements));
    reg_maps->min_rel_offset = ~0U;
    list_init(&reg_maps->indexable_temps);
 
    fe->shader_read_header(fe_data, &ptr, &shader_version);
    prev_ins = current_ins = ptr;
    reg_maps->shader_version = shader_version;
 
    shader_set_limits(shader);
 
    if (!(reg_maps->constf = calloc(((min(shader->limits->constant_float, constf_size) + 31) / 32),
            sizeof(*reg_maps->constf))))
    {
        ERR("Failed to allocate constant map memory.\n");
        return E_OUTOFMEMORY;
    }
 
    while (!fe->shader_is_end(fe_data, &ptr))
    {
        struct wined3d_shader_instruction ins;
 
        current_ins = ptr;
        /* Fetch opcode. */
        fe->shader_read_instruction(fe_data, &ptr, &ins);
 
        /* Unhandled opcode, and its parameters. */
        if (ins.handler_idx == WINED3DSIH_TABLE_SIZE)
        {
            WARN("Encountered unrecognised or invalid instruction.\n");
            return WINED3DERR_INVALIDCALL;
        }
 
        /* Handle declarations. */
        if (ins.handler_idx == WINED3DSIH_DCL
                || ins.handler_idx == WINED3DSIH_DCL_UAV_TYPED)
        {
            struct wined3d_shader_semantic *semantic = &ins.declaration.semantic;
            unsigned int reg_idx = semantic->reg.reg.idx[0].offset;
 
            switch (semantic->reg.reg.type)
            {
                /* Mark input registers used. */
                case WINED3DSPR_INPUT:
                    if (reg_idx >= MAX_REG_INPUT)
                    {
                        ERR("Invalid input register index %u.\n", reg_idx);
                        break;
                    }
                    if (shader_version.type == WINED3D_SHADER_TYPE_PIXEL && shader_version.major == 3
                            && semantic->usage == WINED3D_DECL_USAGE_POSITION && !semantic->usage_idx)
                        return WINED3DERR_INVALIDCALL;
                    reg_maps->input_registers |= 1u << reg_idx;
                    shader_signature_from_semantic(&input_signature_elements[reg_idx], semantic);
                    break;
 
                /* Vertex shader: mark 3.0 output registers used, save token. */
                case WINED3DSPR_OUTPUT:
                    if (reg_idx >= MAX_REG_OUTPUT)
                    {
                        ERR("Invalid output register index %u.\n", reg_idx);
                        break;
                    }
                    reg_maps->output_registers |= 1u << reg_idx;
                    shader_signature_from_semantic(&output_signature_elements[reg_idx], semantic);
                    if (semantic->usage == WINED3D_DECL_USAGE_FOG)
                        reg_maps->fog = 1;
                    if (semantic->usage == WINED3D_DECL_USAGE_PSIZE)
                        reg_maps->point_size = 1;
                    break;
 
                case WINED3DSPR_SAMPLER:
                    shader_record_sample(reg_maps, reg_idx, reg_idx, reg_idx);
                case WINED3DSPR_RESOURCE:
                    if (reg_idx >= ARRAY_SIZE(reg_maps->resource_info))
                    {
                        ERR("Invalid resource index %u.\n", reg_idx);
                        break;
                    }
                    reg_maps->resource_info[reg_idx].type = semantic->resource_type;
                    if (semantic->resource_type == WINED3D_SHADER_RESOURCE_TEXTURE_2DMS && semantic->sample_count == 1)
                        reg_maps->resource_info[reg_idx].type = WINED3D_SHADER_RESOURCE_TEXTURE_2D;
                    if (semantic->resource_type == WINED3D_SHADER_RESOURCE_TEXTURE_2DMSARRAY && semantic->sample_count == 1)
                        reg_maps->resource_info[reg_idx].type = WINED3D_SHADER_RESOURCE_TEXTURE_2DARRAY;
                    reg_maps->resource_info[reg_idx].data_type = semantic->resource_data_type;
                    wined3d_bitmap_set(reg_maps->resource_map, reg_idx);
                    break;
 
                case WINED3DSPR_UAV:
                    if (reg_idx >= ARRAY_SIZE(reg_maps->uav_resource_info))
                    {
                        ERR("Invalid UAV resource index %u.\n", reg_idx);
                        break;
                    }
                    reg_maps->uav_resource_info[reg_idx].type = semantic->resource_type;
                    reg_maps->uav_resource_info[reg_idx].data_type = semantic->resource_data_type;
                    if (ins.flags)
                        FIXME("Ignoring typed UAV flags %#x.\n", ins.flags);
                    break;
 
                default:
                    TRACE("Not recording DCL register type %#x.\n", semantic->reg.reg.type);
                    break;
            }
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_CONSTANT_BUFFER)
        {
            struct wined3d_shader_register *reg = &ins.declaration.src.reg;
            if (reg->idx[0].offset >= WINED3D_MAX_CBS)
            {
                ERR("Invalid CB index %u.\n", reg->idx[0].offset);
            }
            else
            {
                reg_maps->cb_sizes[reg->idx[0].offset] = reg->idx[1].offset;
                wined3d_bitmap_set(&reg_maps->cb_map, reg->idx[0].offset);
            }
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_GLOBAL_FLAGS)
        {
            if (ins.flags & WINED3DSGF_FORCE_EARLY_DEPTH_STENCIL)
            {
                if (shader_version.type == WINED3D_SHADER_TYPE_PIXEL)
                    shader->u.ps.force_early_depth_stencil = TRUE;
                else
                    FIXME("Invalid instruction %#x for shader type %#x.\n",
                            ins.handler_idx, shader_version.type);
            }
            else
            {
                WARN("Ignoring global flags %#x.\n", ins.flags);
            }
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_GS_INSTANCES)
        {
            if (shader_version.type == WINED3D_SHADER_TYPE_GEOMETRY)
                shader->u.gs.instance_count = ins.declaration.count;
            else
                FIXME("Invalid instruction %#x for shader type %#x.\n",
                        ins.handler_idx, shader_version.type);
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_HS_FORK_PHASE_INSTANCE_COUNT
                || ins.handler_idx == WINED3DSIH_DCL_HS_JOIN_PHASE_INSTANCE_COUNT)
        {
            if (phase)
                phase->instance_count = ins.declaration.count;
            else
                FIXME("Instruction %s outside of shader phase.\n",
                        debug_d3dshaderinstructionhandler(ins.handler_idx));
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_IMMEDIATE_CONSTANT_BUFFER)
        {
            if (reg_maps->icb)
                FIXME("Multiple immediate constant buffers.\n");
            reg_maps->icb = ins.declaration.icb;
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_INDEXABLE_TEMP)
        {
            if (phase)
            {
                FIXME("Indexable temporary registers not supported.\n");
            }
            else
            {
                struct wined3d_shader_indexable_temp *reg;
 
                if (!(reg = malloc(sizeof(*reg))))
                    return E_OUTOFMEMORY;
 
                *reg = ins.declaration.indexable_temp;
                list_add_tail(&reg_maps->indexable_temps, &reg->entry);
            }
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_INPUT_PRIMITIVE)
        {
            if (shader_version.type == WINED3D_SHADER_TYPE_GEOMETRY)
                shader->u.gs.input_type = ins.declaration.primitive_type.type;
            else
                FIXME("Invalid instruction %#x for shader type %#x.\n",
                        ins.handler_idx, shader_version.type);
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_INPUT_PS)
        {
            unsigned int reg_idx = ins.declaration.dst.reg.idx[0].offset;
            if (reg_idx >= MAX_REG_INPUT)
            {
                ERR("Invalid register index %u.\n", reg_idx);
                break;
            }
            if (shader_version.type == WINED3D_SHADER_TYPE_PIXEL)
                wined3d_insert_interpolation_mode(shader->u.ps.interpolation_mode, reg_idx, ins.flags);
            else
                FIXME("Invalid instruction %#x for shader type %#x.\n",
                        ins.handler_idx, shader_version.type);
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_OUTPUT)
        {
            if (ins.declaration.dst.reg.type == WINED3DSPR_DEPTHOUT
                    || ins.declaration.dst.reg.type == WINED3DSPR_DEPTHOUTGE
                    || ins.declaration.dst.reg.type == WINED3DSPR_DEPTHOUTLE)
            {
                if (shader_version.type == WINED3D_SHADER_TYPE_PIXEL)
                    shader->u.ps.depth_output = ins.declaration.dst.reg.type;
                else
                    FIXME("Invalid instruction %#x for shader type %#x.\n",
                            ins.handler_idx, shader_version.type);
            }
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_OUTPUT_CONTROL_POINT_COUNT)
        {
            if (shader_version.type == WINED3D_SHADER_TYPE_HULL)
                shader->u.hs.output_vertex_count = ins.declaration.count;
            else
                FIXME("Invalid instruction %#x for shader type %#x.\n", ins.handler_idx, shader_version.type);
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_OUTPUT_TOPOLOGY)
        {
            if (shader_version.type == WINED3D_SHADER_TYPE_GEOMETRY)
                shader->u.gs.output_type = ins.declaration.primitive_type.type;
            else
                FIXME("Invalid instruction %#x for shader type %#x.\n",
                        ins.handler_idx, shader_version.type);
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_RESOURCE_RAW)
        {
            unsigned int reg_idx = ins.declaration.dst.reg.idx[0].offset;
            if (reg_idx >= ARRAY_SIZE(reg_maps->resource_info))
            {
                ERR("Invalid resource index %u.\n", reg_idx);
                break;
            }
            reg_maps->resource_info[reg_idx].type = WINED3D_SHADER_RESOURCE_BUFFER;
            reg_maps->resource_info[reg_idx].data_type = WINED3D_DATA_UINT;
            reg_maps->resource_info[reg_idx].flags = WINED3D_VIEW_BUFFER_RAW;
            wined3d_bitmap_set(reg_maps->resource_map, reg_idx);
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_RESOURCE_STRUCTURED)
        {
            unsigned int reg_idx = ins.declaration.structured_resource.reg.reg.idx[0].offset;
            if (reg_idx >= ARRAY_SIZE(reg_maps->resource_info))
            {
                ERR("Invalid resource index %u.\n", reg_idx);
                break;
            }
            reg_maps->resource_info[reg_idx].type = WINED3D_SHADER_RESOURCE_BUFFER;
            reg_maps->resource_info[reg_idx].data_type = WINED3D_DATA_UINT;
            reg_maps->resource_info[reg_idx].flags = 0;
            reg_maps->resource_info[reg_idx].stride = ins.declaration.structured_resource.byte_stride / 4;
            wined3d_bitmap_set(reg_maps->resource_map, reg_idx);
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_SAMPLER)
        {
            if (ins.flags & WINED3DSI_SAMPLER_COMPARISON_MODE)
                reg_maps->sampler_comparison_mode |= (1u << ins.declaration.dst.reg.idx[0].offset);
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_TEMPS)
        {
            if (phase)
                phase->temporary_count = ins.declaration.count;
            else
                reg_maps->temporary_count = ins.declaration.count;
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_TESSELLATOR_DOMAIN)
        {
            if (shader_version.type == WINED3D_SHADER_TYPE_DOMAIN)
                shader->u.ds.tessellator_domain = ins.declaration.tessellator_domain;
            else if (shader_version.type != WINED3D_SHADER_TYPE_HULL)
                FIXME("Invalid instruction %#x for shader type %#x.\n", ins.handler_idx, shader_version.type);
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_TESSELLATOR_OUTPUT_PRIMITIVE)
        {
            if (shader_version.type == WINED3D_SHADER_TYPE_HULL)
                shader->u.hs.tessellator_output_primitive = ins.declaration.tessellator_output_primitive;
            else
                FIXME("Invalid instruction %#x for shader type %#x.\n", ins.handler_idx, shader_version.type);
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_TESSELLATOR_PARTITIONING)
        {
            if (shader_version.type == WINED3D_SHADER_TYPE_HULL)
                shader->u.hs.tessellator_partitioning = ins.declaration.tessellator_partitioning;
            else
                FIXME("Invalid instruction %#x for shader type %#x.\n", ins.handler_idx, shader_version.type);
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_TGSM_RAW)
        {
            if (FAILED(hr = shader_reg_maps_add_tgsm(reg_maps, ins.declaration.tgsm_raw.reg.reg.idx[0].offset,
                    ins.declaration.tgsm_raw.byte_count / 4, 0)))
                return hr;
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_TGSM_STRUCTURED)
        {
            unsigned int stride = ins.declaration.tgsm_structured.byte_stride / 4;
            unsigned int size = stride * ins.declaration.tgsm_structured.structure_count;
            if (FAILED(hr = shader_reg_maps_add_tgsm(reg_maps,
                    ins.declaration.tgsm_structured.reg.reg.idx[0].offset, size, stride)))
                return hr;
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_THREAD_GROUP)
        {
            if (shader_version.type == WINED3D_SHADER_TYPE_COMPUTE)
            {
                shader->u.cs.thread_group_size = ins.declaration.thread_group_size;
            }
            else
            {
                FIXME("Invalid instruction %#x for shader type %#x.\n",
                        ins.handler_idx, shader_version.type);
            }
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_UAV_RAW)
        {
            unsigned int reg_idx = ins.declaration.dst.reg.idx[0].offset;
            if (reg_idx >= ARRAY_SIZE(reg_maps->uav_resource_info))
            {
                ERR("Invalid UAV resource index %u.\n", reg_idx);
                break;
            }
            if (ins.flags)
                FIXME("Ignoring raw UAV flags %#x.\n", ins.flags);
            reg_maps->uav_resource_info[reg_idx].type = WINED3D_SHADER_RESOURCE_BUFFER;
            reg_maps->uav_resource_info[reg_idx].data_type = WINED3D_DATA_UINT;
            reg_maps->uav_resource_info[reg_idx].flags = WINED3D_VIEW_BUFFER_RAW;
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_UAV_STRUCTURED)
        {
            unsigned int reg_idx = ins.declaration.structured_resource.reg.reg.idx[0].offset;
            if (reg_idx >= ARRAY_SIZE(reg_maps->uav_resource_info))
            {
                ERR("Invalid UAV resource index %u.\n", reg_idx);
                break;
            }
            if (ins.flags)
                FIXME("Ignoring structured UAV flags %#x.\n", ins.flags);
            reg_maps->uav_resource_info[reg_idx].type = WINED3D_SHADER_RESOURCE_BUFFER;
            reg_maps->uav_resource_info[reg_idx].data_type = WINED3D_DATA_UINT;
            reg_maps->uav_resource_info[reg_idx].flags = 0;
            reg_maps->uav_resource_info[reg_idx].stride = ins.declaration.structured_resource.byte_stride / 4;
        }
        else if (ins.handler_idx == WINED3DSIH_DCL_VERTICES_OUT)
        {
            if (shader_version.type == WINED3D_SHADER_TYPE_GEOMETRY)
                shader->u.gs.vertices_out = ins.declaration.count;
            else
                FIXME("Invalid instruction %#x for shader type %#x.\n",
                        ins.handler_idx, shader_version.type);
        }
        else if (ins.handler_idx == WINED3DSIH_DEF)
        {
            struct wined3d_shader_lconst *lconst;
            float *value;
 
            if (!(lconst = malloc(sizeof(*lconst))))
                return E_OUTOFMEMORY;
 
            lconst->idx = ins.dst[0].reg.idx[0].offset;
            memcpy(lconst->value, ins.src[0].reg.u.immconst_data, 4 * sizeof(DWORD));
            value = (float *)lconst->value;
 
            /* In pixel shader 1.X shaders, the constants are clamped between [-1;1] */
            if (shader_version.major == 1 && shader_version.type == WINED3D_SHADER_TYPE_PIXEL)
            {
                if (value[0] < -1.0f) value[0] = -1.0f;
                else if (value[0] > 1.0f) value[0] = 1.0f;
                if (value[1] < -1.0f) value[1] = -1.0f;
                else if (value[1] > 1.0f) value[1] = 1.0f;
                if (value[2] < -1.0f) value[2] = -1.0f;
                else if (value[2] > 1.0f) value[2] = 1.0f;
                if (value[3] < -1.0f) value[3] = -1.0f;
                else if (value[3] > 1.0f) value[3] = 1.0f;
            }
 
            list_add_head(&shader->constantsF, &lconst->entry);
 
            if (isinf(value[0]) || isnan(value[0]) || isinf(value[1]) || isnan(value[1])
                    || isinf(value[2]) || isnan(value[2]) || isinf(value[3]) || isnan(value[3]))
            {
                shader->lconst_inf_or_nan = TRUE;
            }
        }
        else if (ins.handler_idx == WINED3DSIH_DEFI)
        {
            struct wined3d_shader_lconst *lconst;
 
            if (!(lconst = malloc(sizeof(*lconst))))
                return E_OUTOFMEMORY;
 
            lconst->idx = ins.dst[0].reg.idx[0].offset;
            memcpy(lconst->value, ins.src[0].reg.u.immconst_data, 4 * sizeof(DWORD));
 
            list_add_head(&shader->constantsI, &lconst->entry);
            reg_maps->local_int_consts |= (1u << lconst->idx);
        }
        else if (ins.handler_idx == WINED3DSIH_DEFB)
        {
            struct wined3d_shader_lconst *lconst;
 
            if (!(lconst = malloc(sizeof(*lconst))))
                return E_OUTOFMEMORY;
 
            lconst->idx = ins.dst[0].reg.idx[0].offset;
            memcpy(lconst->value, ins.src[0].reg.u.immconst_data, sizeof(DWORD));
 
            list_add_head(&shader->constantsB, &lconst->entry);
            reg_maps->local_bool_consts |= (1u << lconst->idx);
        }
        /* Handle shader phases. */
        else if (ins.handler_idx == WINED3DSIH_HS_CONTROL_POINT_PHASE
                || ins.handler_idx == WINED3DSIH_HS_FORK_PHASE
                || ins.handler_idx == WINED3DSIH_HS_JOIN_PHASE)
        {
            if (FAILED(hr = shader_record_shader_phase(shader, &phase, &ins, current_ins, prev_ins)))
                return hr;
        }
        /* For subroutine prototypes. */
        else if (ins.handler_idx == WINED3DSIH_LABEL)
        {
            reg_maps->labels |= 1u << ins.src[0].reg.idx[0].offset;
        }
        /* Set texture, address, temporary registers. */
        else
        {
            BOOL color0_mov = FALSE;
            unsigned int i;
 
            /* This will loop over all the registers and try to
             * make a bitmask of the ones we're interested in.
             *
             * Relative addressing tokens are ignored, but that's
             * okay, since we'll catch any address registers when
             * they are initialized (required by spec). */
            for (i = 0; i < ins.dst_count; ++i)
            {
                if (!shader_record_register_usage(shader, reg_maps, &ins.dst[i].reg,
                        shader_version.type, constf_size))
                    return WINED3DERR_INVALIDCALL;
 
                if (shader_version.type == WINED3D_SHADER_TYPE_VERTEX)
                {
                    UINT idx = ins.dst[i].reg.idx[0].offset;
 
                    switch (ins.dst[i].reg.type)
                    {
                        case WINED3DSPR_RASTOUT:
                            if (shader_version.major >= 3)
                                break;
                            switch (idx)
                            {
                                case 0: /* oPos */
                                    reg_maps->output_registers |= 1u << 10;
                                    shader_signature_from_usage(&output_signature_elements[10],
                                            WINED3D_DECL_USAGE_POSITION, 0, 10, WINED3DSP_WRITEMASK_ALL);
                                    break;
 
                                case 1: /* oFog */
                                    reg_maps->output_registers |= 1u << 11;
                                    shader_signature_from_usage(&output_signature_elements[11],
                                            WINED3D_DECL_USAGE_FOG, 0, 11, WINED3DSP_WRITEMASK_0);
                                    break;
 
                                case 2: /* oPts */
                                    reg_maps->output_registers |= 1u << 11;
                                    shader_signature_from_usage(&output_signature_elements[12],
                                            WINED3D_DECL_USAGE_PSIZE, 0, 11, WINED3DSP_WRITEMASK_1);
                                    break;
                            }
                            break;
 
                        case WINED3DSPR_ATTROUT:
                            if (shader_version.major >= 3)
                                break;
                            if (idx < 2)
                            {
                                idx += 8;
                                if (reg_maps->output_registers & (1u << idx))
                                {
                                    output_signature_elements[idx].mask |= ins.dst[i].write_mask;
                                }
                                else
                                {
                                    reg_maps->output_registers |= 1u << idx;
                                    shader_signature_from_usage(&output_signature_elements[idx],
                                            WINED3D_DECL_USAGE_COLOR, idx - 8, idx, ins.dst[i].write_mask);
                                }
                            }
                            break;
 
                        case WINED3DSPR_TEXCRDOUT: /* WINED3DSPR_OUTPUT */
                            if (shader_version.major >= 3)
                            {
                                if (idx >= ARRAY_SIZE(reg_maps->u.output_registers_mask))
                                {
                                    WARN("Invalid output register index %u.\n", idx);
                                    break;
                                }
                                reg_maps->u.output_registers_mask[idx] |= ins.dst[i].write_mask;
                                break;
                            }
                            if (idx >= ARRAY_SIZE(reg_maps->u.texcoord_mask))
                            {
                                WARN("Invalid texcoord index %u.\n", idx);
                                break;
                            }
                            reg_maps->u.texcoord_mask[idx] |= ins.dst[i].write_mask;
                            if (reg_maps->output_registers & (1u << idx))
                            {
                                output_signature_elements[idx].mask |= ins.dst[i].write_mask;
                            }
                            else
                            {
                                reg_maps->output_registers |= 1u << idx;
                                shader_signature_from_usage(&output_signature_elements[idx],
                                        WINED3D_DECL_USAGE_TEXCOORD, idx, idx, ins.dst[i].write_mask);
                            }
                            break;
 
                        default:
                            break;
                    }
                }
 
                if (shader_version.type == WINED3D_SHADER_TYPE_PIXEL)
                {
                    if (ins.dst[i].reg.type == WINED3DSPR_COLOROUT && !ins.dst[i].reg.idx[0].offset)
                    {
                        /* Many 2.0 and 3.0 pixel shaders end with a MOV from a temp register to
                         * COLOROUT 0. If we know this in advance, the ARB shader backend can skip
                         * the mov and perform the sRGB write correction from the source register.
                         *
                         * However, if the mov is only partial, we can't do this, and if the write
                         * comes from an instruction other than MOV it is hard to do as well. If
                         * COLOROUT 0 is overwritten partially later, the marker is dropped again. */
                        shader->u.ps.color0_mov = FALSE;
                        if (ins.handler_idx == WINED3DSIH_MOV
                                && ins.dst[i].write_mask == WINED3DSP_WRITEMASK_ALL)
                        {
                            /* Used later when the source register is read. */
                            color0_mov = TRUE;
                        }
                    }
                    /* Also drop the MOV marker if the source register is overwritten prior to the shader
                     * end
                     */
                    else if (ins.dst[i].reg.type == WINED3DSPR_TEMP
                            && ins.dst[i].reg.idx[0].offset == shader->u.ps.color0_reg)
                    {
                        shader->u.ps.color0_mov = FALSE;
                    }
                }
 
                /* Declare 1.x samplers implicitly, based on the destination reg. number. */
                if (shader_version.major == 1
                        && (ins.handler_idx == WINED3DSIH_TEX
                            || ins.handler_idx == WINED3DSIH_TEXBEM
                            || ins.handler_idx == WINED3DSIH_TEXBEML
                            || ins.handler_idx == WINED3DSIH_TEXDP3TEX
                            || ins.handler_idx == WINED3DSIH_TEXM3x2TEX
                            || ins.handler_idx == WINED3DSIH_TEXM3x3SPEC
                            || ins.handler_idx == WINED3DSIH_TEXM3x3TEX
                            || ins.handler_idx == WINED3DSIH_TEXM3x3VSPEC
                            || ins.handler_idx == WINED3DSIH_TEXREG2AR
                            || ins.handler_idx == WINED3DSIH_TEXREG2GB
                            || ins.handler_idx == WINED3DSIH_TEXREG2RGB))
                {
                    unsigned int reg_idx = ins.dst[i].reg.idx[0].offset;
 
                    if (reg_idx >= ARRAY_SIZE(reg_maps->resource_info))
                    {
                        WARN("Invalid 1.x sampler index %u.\n", reg_idx);
                        continue;
                    }
 
                    TRACE("Setting fake 2D resource for 1.x pixelshader.\n");
                    reg_maps->resource_info[reg_idx].type = WINED3D_SHADER_RESOURCE_TEXTURE_2D;
                    reg_maps->resource_info[reg_idx].data_type = WINED3D_DATA_FLOAT;
                    shader_record_sample(reg_maps, reg_idx, reg_idx, reg_idx);
                    wined3d_bitmap_set(reg_maps->resource_map, reg_idx);
 
                    /* texbem is only valid with < 1.4 pixel shaders */
                    if (ins.handler_idx == WINED3DSIH_TEXBEM
                            || ins.handler_idx == WINED3DSIH_TEXBEML)
                    {
                        reg_maps->bumpmat |= 1u << reg_idx;
                        if (ins.handler_idx == WINED3DSIH_TEXBEML)
                        {
                            reg_maps->luminanceparams |= 1u << reg_idx;
                        }
                    }
                }
                else if (ins.handler_idx == WINED3DSIH_BEM)
                {
                    reg_maps->bumpmat |= 1u << ins.dst[i].reg.idx[0].offset;
                }
            }
 
            if (ins.handler_idx == WINED3DSIH_IMM_ATOMIC_ALLOC || ins.handler_idx == WINED3DSIH_IMM_ATOMIC_CONSUME)
            {
                unsigned int reg_idx = ins.src[0].reg.idx[0].offset;
                if (reg_idx >= MAX_UNORDERED_ACCESS_VIEWS)
                {
                    ERR("Invalid UAV index %u.\n", reg_idx);
                    break;
                }
                reg_maps->uav_counter_mask |= (1u << reg_idx);
            }
            else if ((WINED3DSIH_ATOMIC_AND <= ins.handler_idx && ins.handler_idx <= WINED3DSIH_ATOMIC_XOR)
                    || (WINED3DSIH_IMM_ATOMIC_AND <= ins.handler_idx && ins.handler_idx <= WINED3DSIH_IMM_ATOMIC_XOR)
                    || (ins.handler_idx == WINED3DSIH_BUFINFO && ins.src[0].reg.type == WINED3DSPR_UAV)
                    || ins.handler_idx == WINED3DSIH_LD_UAV_TYPED
                    || (ins.handler_idx == WINED3DSIH_LD_RAW && ins.src[1].reg.type == WINED3DSPR_UAV)
                    || (ins.handler_idx == WINED3DSIH_LD_STRUCTURED && ins.src[2].reg.type == WINED3DSPR_UAV))
            {
                const struct wined3d_shader_register *reg;
 
                if (ins.handler_idx == WINED3DSIH_LD_UAV_TYPED || ins.handler_idx == WINED3DSIH_LD_RAW)
                    reg = &ins.src[1].reg;
                else if (ins.handler_idx == WINED3DSIH_LD_STRUCTURED)
                    reg = &ins.src[2].reg;
                else if (WINED3DSIH_ATOMIC_AND <= ins.handler_idx && ins.handler_idx <= WINED3DSIH_ATOMIC_XOR)
                    reg = &ins.dst[0].reg;
                else if (ins.handler_idx == WINED3DSIH_BUFINFO)
                    reg = &ins.src[0].reg;
                else
                    reg = &ins.dst[1].reg;
 
                if (reg->type == WINED3DSPR_UAV)
                {
                    if (reg->idx[0].offset >= MAX_UNORDERED_ACCESS_VIEWS)
                    {
                        ERR("Invalid UAV index %u.\n", reg->idx[0].offset);
                        break;
                    }
                    reg_maps->uav_read_mask |= (1u << reg->idx[0].offset);
                }
            }
            else if (ins.handler_idx == WINED3DSIH_NRM)
            {
                reg_maps->usesnrm = 1;
            }
            else if (ins.handler_idx == WINED3DSIH_DSY
                    || ins.handler_idx == WINED3DSIH_DSY_COARSE
                    || ins.handler_idx == WINED3DSIH_DSY_FINE)
            {
                reg_maps->usesdsy = 1;
            }
            else if (ins.handler_idx == WINED3DSIH_DSX
                    || ins.handler_idx == WINED3DSIH_DSX_COARSE
                    || ins.handler_idx == WINED3DSIH_DSX_FINE)
            {
                reg_maps->usesdsx = 1;
            }
            else if (ins.handler_idx == WINED3DSIH_TEXLDD) reg_maps->usestexldd = 1;
            else if (ins.handler_idx == WINED3DSIH_TEXLDL) reg_maps->usestexldl = 1;
            else if (ins.handler_idx == WINED3DSIH_MOVA) reg_maps->usesmova = 1;
            else if (ins.handler_idx == WINED3DSIH_IFC) reg_maps->usesifc = 1;
            else if (ins.handler_idx == WINED3DSIH_CALL) reg_maps->usescall = 1;
            else if (ins.handler_idx == WINED3DSIH_POW) reg_maps->usespow = 1;
            else if (ins.handler_idx == WINED3DSIH_LOOP
                    || ins.handler_idx == WINED3DSIH_REP)
            {
                ++cur_loop_depth;
                if (cur_loop_depth > max_loop_depth)
                    max_loop_depth = cur_loop_depth;
            }
            else if (ins.handler_idx == WINED3DSIH_ENDLOOP
                    || ins.handler_idx == WINED3DSIH_ENDREP)
            {
                --cur_loop_depth;
            }
            else if (ins.handler_idx == WINED3DSIH_GATHER4
                    || ins.handler_idx == WINED3DSIH_GATHER4_C
                    || ins.handler_idx == WINED3DSIH_SAMPLE
                    || ins.handler_idx == WINED3DSIH_SAMPLE_B
                    || ins.handler_idx == WINED3DSIH_SAMPLE_C
                    || ins.handler_idx == WINED3DSIH_SAMPLE_C_LZ
                    || ins.handler_idx == WINED3DSIH_SAMPLE_GRAD
                    || ins.handler_idx == WINED3DSIH_SAMPLE_LOD)
            {
                shader_record_sample(reg_maps, ins.src[1].reg.idx[0].offset,
                        ins.src[2].reg.idx[0].offset, reg_maps->sampler_map.count);
            }
            else if (ins.handler_idx == WINED3DSIH_GATHER4_PO
                    || ins.handler_idx == WINED3DSIH_GATHER4_PO_C)
            {
                shader_record_sample(reg_maps, ins.src[2].reg.idx[0].offset,
                        ins.src[3].reg.idx[0].offset, reg_maps->sampler_map.count);
            }
            else if ((ins.handler_idx == WINED3DSIH_BUFINFO && ins.src[0].reg.type == WINED3DSPR_RESOURCE)
                    || (ins.handler_idx == WINED3DSIH_SAMPLE_INFO && ins.src[0].reg.type == WINED3DSPR_RESOURCE))
            {
                shader_record_sample(reg_maps, ins.src[0].reg.idx[0].offset,
                        WINED3D_SAMPLER_DEFAULT, reg_maps->sampler_map.count);
            }
            else if (ins.handler_idx == WINED3DSIH_LD
                    || ins.handler_idx == WINED3DSIH_LD2DMS
                    || (ins.handler_idx == WINED3DSIH_LD_RAW && ins.src[1].reg.type == WINED3DSPR_RESOURCE)
                    || (ins.handler_idx == WINED3DSIH_RESINFO && ins.src[1].reg.type == WINED3DSPR_RESOURCE))
            {
                shader_record_sample(reg_maps, ins.src[1].reg.idx[0].offset,
                        WINED3D_SAMPLER_DEFAULT, reg_maps->sampler_map.count);
            }
            else if (ins.handler_idx == WINED3DSIH_LD_STRUCTURED
                    && ins.src[2].reg.type == WINED3DSPR_RESOURCE)
            {
                shader_record_sample(reg_maps, ins.src[2].reg.idx[0].offset,
                        WINED3D_SAMPLER_DEFAULT, reg_maps->sampler_map.count);
            }
 
            if (ins.predicate)
                if (!shader_record_register_usage(shader, reg_maps, &ins.predicate->reg,
                        shader_version.type, constf_size))
                    return WINED3DERR_INVALIDCALL;
 
            for (i = 0; i < ins.src_count; ++i)
            {
                unsigned int count = get_instr_extra_regcount(ins.handler_idx, i);
                struct wined3d_shader_register reg = ins.src[i].reg;
 
                if (!shader_record_register_usage(shader, reg_maps, &ins.src[i].reg,
                        shader_version.type, constf_size))
                    return WINED3DERR_INVALIDCALL;
                while (count)
                {
                    ++reg.idx[0].offset;
                    if (!shader_record_register_usage(shader, reg_maps, &reg,
                            shader_version.type, constf_size))
                        return WINED3DERR_INVALIDCALL;
                    --count;
                }
 
                if (color0_mov)
                {
                    if (ins.src[i].reg.type == WINED3DSPR_TEMP
                            && ins.src[i].swizzle == WINED3DSP_NOSWIZZLE)
                    {
                        shader->u.ps.color0_mov = TRUE;
                        shader->u.ps.color0_reg = ins.src[i].reg.idx[0].offset;
                    }
                }
            }
        }
 
        prev_ins = current_ins;
    }
    reg_maps->loop_depth = max_loop_depth;
 
    if (phase)
    {
        phase->end = prev_ins;
        phase = NULL;
    }
 
    /* PS before 2.0 don't have explicit color outputs. Instead the value of
     * R0 is written to the render target. */
    if (shader_version.major < 2 && shader_version.type == WINED3D_SHADER_TYPE_PIXEL)
        reg_maps->rt_mask |= (1u << 0);
 
    if (input_signature->elements)
    {
        for (i = 0; i < input_signature->element_count; ++i)
        {
            if (shader_version.type == WINED3D_SHADER_TYPE_VERTEX)
            {
                if (input_signature->elements[i].register_idx >= ARRAY_SIZE(shader->u.vs.attributes))
                {
                    WARN("Invalid input signature register index %u.\n", input_signature->elements[i].register_idx);
                    return WINED3DERR_INVALIDCALL;
                }
            }
            else if (shader_version.type == WINED3D_SHADER_TYPE_PIXEL)
            {
                if (input_signature->elements[i].sysval_semantic == WINED3D_SV_POSITION)
                    reg_maps->vpos = 1;
                else if (input_signature->elements[i].sysval_semantic == WINED3D_SV_IS_FRONT_FACE)
                    reg_maps->usesfacing = 1;
            }
            reg_maps->input_registers |= 1u << input_signature->elements[i].register_idx;
        }
    }
    else if (!input_signature->elements && reg_maps->input_registers)
    {
        unsigned int count = wined3d_popcount(reg_maps->input_registers);
        struct wined3d_shader_signature_element *e;
        unsigned int i;
 
        if (!(input_signature->elements = calloc(count, sizeof(*input_signature->elements))))
            return E_OUTOFMEMORY;
        input_signature->element_count = count;
 
        e = input_signature->elements;
        for (i = 0; i < ARRAY_SIZE(input_signature_elements); ++i)
        {
            if (!(reg_maps->input_registers & (1u << i)))
                continue;
            input_signature_elements[i].register_idx = i;
            *e++ = input_signature_elements[i];
        }
    }
 
    if (output_signature->elements)
    {
        if (FAILED(hr = shader_scan_output_signature(shader)))
            return hr;
    }
    else if (reg_maps->output_registers)
    {
        struct wined3d_shader_signature_element *e;
        unsigned int count = 0;
 
        for (i = 0; i < ARRAY_SIZE(output_signature_elements); ++i)
        {
            if (output_signature_elements[i].semantic_name)
                ++count;
        }
 
        if (!(output_signature->elements = calloc(count, sizeof(*output_signature->elements))))
            return E_OUTOFMEMORY;
        output_signature->element_count = count;
 
        e = output_signature->elements;
        for (i = 0; i < ARRAY_SIZE(output_signature_elements); ++i)
        {
            if (output_signature_elements[i].semantic_name)
                *e++ = output_signature_elements[i];
        }
    }
 
    return WINED3D_OK;
}
 
static void shader_cleanup_reg_maps(struct wined3d_shader_reg_maps *reg_maps)
{
    struct wined3d_shader_indexable_temp *reg, *reg_next;
 
    free(reg_maps->constf);
    free(reg_maps->sampler_map.entries);
 
    LIST_FOR_EACH_ENTRY_SAFE(reg, reg_next, &reg_maps->indexable_temps, struct wined3d_shader_indexable_temp, entry)
        free(reg);
    list_init(&reg_maps->indexable_temps);
 
    free(reg_maps->tgsm);
}
 
unsigned int shader_find_free_input_register(const struct wined3d_shader_reg_maps *reg_maps, unsigned int max)
{
    DWORD map = 1u << max;
    map |= map - 1;
    map &= reg_maps->shader_version.major < 3 ? ~reg_maps->texcoord : ~reg_maps->input_registers;
 
    return wined3d_log2i(map);
}
 
/* Shared code in order to generate the bulk of the shader string. */
HRESULT shader_generate_code(const struct wined3d_shader *shader, struct wined3d_string_buffer *buffer,
        const struct wined3d_shader_reg_maps *reg_maps, void *backend_ctx,
        const DWORD *start, const DWORD *end)
{
    struct wined3d_device *device = shader->device;
    const struct wined3d_shader_frontend *fe = shader->frontend;
    void *fe_data = shader->frontend_data;
    struct wined3d_shader_version shader_version;
    struct wined3d_shader_parser_state state;
    struct wined3d_shader_instruction ins;
    struct wined3d_shader_tex_mx tex_mx;
    struct wined3d_shader_context ctx;
    const DWORD *ptr;
 
    /* Initialize current parsing state. */
    tex_mx.current_row = 0;
    state.current_loop_depth = 0;
    state.current_loop_reg = 0;
    state.in_subroutine = FALSE;
 
    ctx.shader = shader;
    ctx.reg_maps = reg_maps;
    ctx.buffer = buffer;
    ctx.tex_mx = &tex_mx;
    ctx.state = &state;
    ctx.backend_data = backend_ctx;
    ins.ctx = &ctx;
 
    fe->shader_read_header(fe_data, &ptr, &shader_version);
    if (start)
        ptr = start;
 
    while (!fe->shader_is_end(fe_data, &ptr) && ptr != end)
    {
        /* Read opcode. */
        fe->shader_read_instruction(fe_data, &ptr, &ins);
 
        /* Unknown opcode and its parameters. */
        if (ins.handler_idx == WINED3DSIH_TABLE_SIZE)
        {
            WARN("Encountered unrecognised or invalid instruction.\n");
            return WINED3DERR_INVALIDCALL;
        }
 
        if (ins.predicate)
            FIXME("Predicates not implemented.\n");
 
        /* Call appropriate function for output target */
        device->shader_backend->shader_handle_instruction(&ins);
    }
 
    return WINED3D_OK;
}
 
static void shader_cleanup(struct wined3d_shader *shader)
{
    if (shader->reg_maps.shader_version.type == WINED3D_SHADER_TYPE_HULL)
    {
        free(shader->u.hs.phases.control_point);
        free(shader->u.hs.phases.fork);
        free(shader->u.hs.phases.join);
    }
 
    free(shader->patch_constant_signature.elements);
    free(shader->output_signature.elements);
    free(shader->input_signature.elements);
    shader->device->shader_backend->shader_destroy(shader);
    shader_cleanup_reg_maps(&shader->reg_maps);
    free(shader->byte_code);
    shader_delete_constant_list(&shader->constantsF);
    shader_delete_constant_list(&shader->constantsB);
    shader_delete_constant_list(&shader->constantsI);
    list_remove(&shader->shader_list_entry);
 
    if (shader->frontend && shader->frontend_data)
        shader->frontend->shader_free(shader->frontend_data);
}
 
struct shader_none_priv
{
    const struct wined3d_vertex_pipe_ops *vertex_pipe;
    const struct wined3d_fragment_pipe_ops *fragment_pipe;
};
 
static void shader_none_handle_instruction(const struct wined3d_shader_instruction *ins) {}
static void shader_none_precompile(void *shader_priv, struct wined3d_shader *shader) {}
static void shader_none_apply_compute_state(void *shader_priv, struct wined3d_context *context,
        const struct wined3d_state *state) {}
static void shader_none_update_float_vertex_constants(struct wined3d_device *device, UINT start, UINT count) {}
static void shader_none_update_float_pixel_constants(struct wined3d_device *device, UINT start, UINT count) {}
static void shader_none_destroy(struct wined3d_shader *shader) {}
static void shader_none_free_context_data(struct wined3d_context *context) {}
static void shader_none_init_context_state(struct wined3d_context *context) {}
 
/* Context activation is done by the caller. */
static void shader_none_apply_draw_state(void *shader_priv, struct wined3d_context *context,
        const struct wined3d_state *state)
{
    struct shader_none_priv *priv = shader_priv;
 
    priv->vertex_pipe->vp_apply_draw_state(context, state);
    priv->fragment_pipe->fp_apply_draw_state(context, state);
}
 
/* Context activation is done by the caller. */
static void shader_none_disable(void *shader_priv, struct wined3d_context *context)
{
    struct shader_none_priv *priv = shader_priv;
 
    priv->vertex_pipe->vp_disable(context);
    priv->fragment_pipe->fp_disable(context);
 
    context->shader_update_mask = (1u << WINED3D_SHADER_TYPE_PIXEL)
            | (1u << WINED3D_SHADER_TYPE_VERTEX)
            | (1u << WINED3D_SHADER_TYPE_GEOMETRY)
            | (1u << WINED3D_SHADER_TYPE_HULL)
            | (1u << WINED3D_SHADER_TYPE_DOMAIN)
            | (1u << WINED3D_SHADER_TYPE_COMPUTE);
}
 
static HRESULT shader_none_alloc(struct wined3d_device *device, const struct wined3d_vertex_pipe_ops *vertex_pipe,
        const struct wined3d_fragment_pipe_ops *fragment_pipe)
{
    void *vertex_priv, *fragment_priv;
    struct shader_none_priv *priv;
 
    if (!(priv = malloc(sizeof(*priv))))
        return E_OUTOFMEMORY;
 
    if (!(vertex_priv = vertex_pipe->vp_alloc(&none_shader_backend, priv)))
    {
        ERR("Failed to initialize vertex pipe.\n");
        free(priv);
        return E_FAIL;
    }
 
    if (!(fragment_priv = fragment_pipe->alloc_private(&none_shader_backend, priv)))
    {
        ERR("Failed to initialize fragment pipe.\n");
        vertex_pipe->vp_free(device, NULL);
        free(priv);
        return E_FAIL;
    }
 
    priv->vertex_pipe = vertex_pipe;
    priv->fragment_pipe = fragment_pipe;
 
    device->vertex_priv = vertex_priv;
    device->fragment_priv = fragment_priv;
    device->shader_priv = priv;
 
    return WINED3D_OK;
}
 
static void shader_none_free(struct wined3d_device *device, struct wined3d_context *context)
{
    struct shader_none_priv *priv = device->shader_priv;
 
    priv->fragment_pipe->free_private(device, context);
    priv->vertex_pipe->vp_free(device, context);
    free(priv);
}
 
static BOOL shader_none_allocate_context_data(struct wined3d_context *context)
{
    return TRUE;
}
 
static void shader_none_get_caps(const struct wined3d_adapter *adapter, struct shader_caps *caps)
{
    /* Set the shader caps to 0 for the none shader backend */
    memset(caps, 0, sizeof(*caps));
}
 
static BOOL shader_none_color_fixup_supported(struct color_fixup_desc fixup)
{
    /* We "support" every possible fixup, since we don't support any shader
     * model, and will never have to actually sample a texture. */
    return TRUE;
}
 
static uint64_t shader_none_shader_compile(struct wined3d_context *context, const struct wined3d_shader_desc *shader_desc,
        enum wined3d_shader_type shader_type)
{
    return 0;
}
 
const struct wined3d_shader_backend_ops none_shader_backend =
{
    shader_none_handle_instruction,
    shader_none_precompile,
    shader_none_apply_draw_state,
    shader_none_apply_compute_state,
    shader_none_disable,
    shader_none_update_float_vertex_constants,
    shader_none_update_float_pixel_constants,
    shader_none_destroy,
    shader_none_alloc,
    shader_none_free,
    shader_none_allocate_context_data,
    shader_none_free_context_data,
    shader_none_init_context_state,
    shader_none_get_caps,
    shader_none_color_fixup_supported,
    shader_none_shader_compile,
};
 
static unsigned int shader_max_version_from_feature_level(enum wined3d_feature_level level)
{
    switch (level)
    {
        case WINED3D_FEATURE_LEVEL_11_1:
        case WINED3D_FEATURE_LEVEL_11:
            return 5;
        case WINED3D_FEATURE_LEVEL_10_1:
        case WINED3D_FEATURE_LEVEL_10:
            return 4;
        case WINED3D_FEATURE_LEVEL_9_3:
            return 3;
        case WINED3D_FEATURE_LEVEL_9_2:
        case WINED3D_FEATURE_LEVEL_9_1:
            return 2;
        default:
            return 1;
    }
}
 
static struct wined3d_shader_signature_element *shader_find_signature_element(const struct wined3d_shader_signature *s,
        unsigned int stream_idx, const char *semantic_name, unsigned int semantic_idx)
{
    struct wined3d_shader_signature_element *e = s->elements;
    unsigned int i;
 
    for (i = 0; i < s->element_count; ++i)
    {
        if (e[i].stream_idx == stream_idx
                && !stricmp(e[i].semantic_name, semantic_name)
                && e[i].semantic_idx == semantic_idx)
            return &e[i];
    }
 
    return NULL;
}
 
BOOL shader_get_stream_output_register_info(const struct wined3d_shader *shader,
        const struct wined3d_stream_output_element *so_element, unsigned int *register_idx, unsigned int *component_idx)
{
    const struct wined3d_shader_signature_element *output;
    unsigned int idx;
 
    if (!(output = shader_find_signature_element(&shader->output_signature,
            so_element->stream_idx, so_element->semantic_name, so_element->semantic_idx)))
        return FALSE;
 
    for (idx = 0; idx < 4; ++idx)
    {
        if (output->mask & (1u << idx))
            break;
    }
    idx += so_element->component_idx;
 
    *register_idx = output->register_idx;
    *component_idx = idx;
    return TRUE;
}
 
static HRESULT geometry_shader_init_so_desc(struct wined3d_geometry_shader *gs, struct wined3d_device *device,
        const struct wined3d_stream_output_desc *so_desc)
{
    struct wined3d_so_desc_entry *s;
    struct wine_rb_entry *entry;
    unsigned int i;
    size_t size;
    char *name;
 
    if ((entry = wine_rb_get(&device->so_descs, so_desc)))
    {
        gs->so_desc = &WINE_RB_ENTRY_VALUE(entry, struct wined3d_so_desc_entry, entry)->desc;
        return WINED3D_OK;
    }
 
    size = FIELD_OFFSET(struct wined3d_so_desc_entry, elements[so_desc->element_count]);
    for (i = 0; i < so_desc->element_count; ++i)
    {
        const char *n = so_desc->elements[i].semantic_name;
 
        if (n)
            size += strlen(n) + 1;
    }
    if (!(s = malloc(size)))
        return E_OUTOFMEMORY;
 
    s->desc = *so_desc;
 
    memcpy(s->elements, so_desc->elements, so_desc->element_count * sizeof(*s->elements));
    s->desc.elements = s->elements;
 
    name = (char *)&s->elements[s->desc.element_count];
    for (i = 0; i < so_desc->element_count; ++i)
    {
        struct wined3d_stream_output_element *e = &s->elements[i];
 
        if (!e->semantic_name)
            continue;
 
        size = strlen(e->semantic_name) + 1;
        memcpy(name, e->semantic_name, size);
        e->semantic_name = name;
        name += size;
    }
 
    if (wine_rb_put(&device->so_descs, &s->desc, &s->entry) == -1)
    {
        free(s);
        return E_FAIL;
    }
    gs->so_desc = &s->desc;
 
    return WINED3D_OK;
}
 
static HRESULT geometry_shader_init_stream_output(struct wined3d_shader *shader,
        const struct wined3d_stream_output_desc *so_desc)
{
    const struct wined3d_shader_frontend *fe = shader->frontend;
    const struct wined3d_shader_signature_element *output;
    unsigned int i, component_idx, register_idx, mask;
    struct wined3d_shader_version shader_version;
    const DWORD *ptr;
    void *fe_data;
    HRESULT hr;
 
    if (!so_desc)
        return WINED3D_OK;
 
    if (!(fe_data = fe->shader_init(shader->function, shader->functionLength, &shader->output_signature)))
    {
        WARN("Failed to initialise frontend data.\n");
        return WINED3DERR_INVALIDCALL;
    }
    fe->shader_read_header(fe_data, &ptr, &shader_version);
    fe->shader_free(fe_data);
 
    switch (shader_version.type)
    {
        case WINED3D_SHADER_TYPE_VERTEX:
        case WINED3D_SHADER_TYPE_DOMAIN:
            shader->function = NULL;
            shader->functionLength = 0;
            break;
        case WINED3D_SHADER_TYPE_GEOMETRY:
            break;
        default:
            WARN("Wrong shader type %s.\n", debug_shader_type(shader_version.type));
            return E_INVALIDARG;
    }
 
    if (!shader->function)
    {
        shader->reg_maps.shader_version = shader_version;
        shader->reg_maps.shader_version.type = WINED3D_SHADER_TYPE_GEOMETRY;
        shader_set_limits(shader);
        if (FAILED(hr = shader_scan_output_signature(shader)))
            return hr;
    }
 
    for (i = 0; i < so_desc->element_count; ++i)
    {
        const struct wined3d_stream_output_element *e = &so_desc->elements[i];
 
        if (!e->semantic_name)
            continue;
        if (!(output = shader_find_signature_element(&shader->output_signature,
                e->stream_idx, e->semantic_name, e->semantic_idx))
                || !shader_get_stream_output_register_info(shader, e, &register_idx, &component_idx))
        {
            WARN("Failed to find output signature element for stream output entry.\n");
            return E_INVALIDARG;
        }
 
        mask = wined3d_mask_from_size(e->component_count) << component_idx;
        if ((output->mask & 0xff & mask) != mask)
        {
            WARN("Invalid component range %u-%u (mask %#x), output mask %#x.\n",
                    component_idx, e->component_count, mask, output->mask & 0xff);
            return E_INVALIDARG;
        }
    }
 
    if (FAILED(hr = geometry_shader_init_so_desc(&shader->u.gs, shader->device, so_desc)))
    {
        WARN("Failed to initialise stream output description, hr %#lx.\n", hr);
        return hr;
    }
 
    return WINED3D_OK;
}
 
static void shader_trace(const void *code, size_t size, enum vkd3d_shader_source_type source_type)
{
    struct vkd3d_shader_compile_info info = {.type = VKD3D_SHADER_STRUCTURE_TYPE_COMPILE_INFO};
    struct vkd3d_shader_code d3d_asm;
    const char *ptr, *end, *line;
    char *messages;
    int ret;
 
    static const struct vkd3d_shader_compile_option compile_options[] =
    {
        {VKD3D_SHADER_COMPILE_OPTION_API_VERSION, VKD3D_SHADER_API_VERSION_1_6},
    };
 
    info.source.code = code;
    info.source.size = size;
    info.source_type = source_type;
    info.target_type = VKD3D_SHADER_TARGET_D3D_ASM;
    info.options = compile_options;
    info.option_count = ARRAY_SIZE(compile_options);
    info.log_level = VKD3D_SHADER_LOG_WARNING;
 
    ret = vkd3d_shader_compile(&info, &d3d_asm, &messages);
    if (messages && *messages && FIXME_ON(d3d_shader))
    {
        FIXME("Shader log:\n");
        ptr = messages;
        end = ptr + strlen(ptr);
        while ((line = wined3d_get_line(&ptr, end)))
            FIXME("    %.*s", (int)(ptr - line), line);
        FIXME("\n");
    }
    vkd3d_shader_free_messages(messages);
 
    if (ret < 0)
    {
        ERR("Failed to disassemble, ret %d.\n", ret);
        return;
    }
 
    ptr = d3d_asm.code;
    end = ptr + d3d_asm.size;
    while ((line = wined3d_get_line(&ptr, end)))
        TRACE("    %.*s", (int)(ptr - line), line);
    TRACE("\n");
 
    vkd3d_shader_free_shader_code(&d3d_asm);
}
 
static HRESULT shader_set_function(struct wined3d_shader *shader, const struct wined3d_shader_desc *desc,
        enum wined3d_shader_type type, const struct wined3d_stream_output_desc *so_desc, unsigned int float_const_count)
{
    const struct wined3d_d3d_info *d3d_info = &shader->device->adapter->d3d_info;
    struct wined3d_shader_reg_maps *reg_maps = &shader->reg_maps;
    const struct wined3d_shader_version *version = &reg_maps->shader_version;
    unsigned int highest_reg_used = 0, num_regs_used = 0;
    const struct wined3d_shader_frontend *fe;
    unsigned int backend_version;
    HRESULT hr;
 
    TRACE("shader %p, byte_code %p, size %#Ix, type %s, float_const_count %u.\n",
            shader, desc->byte_code, desc->byte_code_size, debug_shader_type(type), float_const_count);
 
    if (!desc->byte_code)
        return WINED3DERR_INVALIDCALL;
 
    if (desc->byte_code_size == ~(size_t)0)
    {
        struct wined3d_shader_version shader_version;
        const struct wined3d_shader_frontend *fe;
        struct wined3d_shader_instruction ins;
        const DWORD *ptr;
        void *fe_data;
 
        shader->source_type = VKD3D_SHADER_SOURCE_D3D_BYTECODE;
        if (!(shader->frontend = shader_select_frontend(shader->source_type)))
        {
            FIXME("Unable to find frontend for shader.\n");
            shader_cleanup(shader);
            return WINED3DERR_INVALIDCALL;
        }
 
        fe = shader->frontend;
        if (!(fe_data = fe->shader_init(desc->byte_code, desc->byte_code_size, &shader->output_signature)))
        {
            WARN("Failed to initialise frontend data.\n");
            shader_cleanup(shader);
            return WINED3DERR_INVALIDCALL;
        }
 
        fe->shader_read_header(fe_data, &ptr, &shader_version);
        while (!fe->shader_is_end(fe_data, &ptr))
            fe->shader_read_instruction(fe_data, &ptr, &ins);
 
        fe->shader_free(fe_data);
 
        shader->byte_code_size = (ptr - desc->byte_code) * sizeof(*ptr);
 
        if (!(shader->byte_code = malloc(shader->byte_code_size)))
        {
            shader_cleanup(shader);
            return E_OUTOFMEMORY;
        }
        memcpy(shader->byte_code, desc->byte_code, shader->byte_code_size);
 
        shader->function = shader->byte_code;
        shader->functionLength = shader->byte_code_size;
    }
    else
    {
        unsigned int max_version;
 
        if (!(shader->byte_code = malloc(desc->byte_code_size)))
        {
            shader_cleanup(shader);
            return E_OUTOFMEMORY;
        }
        memcpy(shader->byte_code, desc->byte_code, desc->byte_code_size);
        shader->byte_code_size = desc->byte_code_size;
 
        max_version = shader_max_version_from_feature_level(shader->device->cs->c.state->feature_level);
        if (FAILED(hr = wined3d_shader_extract_from_dxbc(shader, max_version, &shader->source_type)))
        {
            shader_cleanup(shader);
            return hr;
        }
 
        if (!(shader->frontend = shader_select_frontend(shader->source_type)))
        {
            FIXME("Unable to find frontend for shader.\n");
            shader_cleanup(shader);
            return WINED3DERR_INVALIDCALL;
        }
    }
 
    if (TRACE_ON(d3d_shader))
    {
        if (shader->source_type == VKD3D_SHADER_SOURCE_D3D_BYTECODE)
            shader_trace(shader->function, shader->functionLength, shader->source_type);
        else
            shader_trace(shader->byte_code, shader->byte_code_size, shader->source_type);
    }
 
    if (type == WINED3D_SHADER_TYPE_GEOMETRY)
    {
        if (FAILED(hr = geometry_shader_init_stream_output(shader, so_desc)))
            return hr;
        if (!shader->function)
            return S_OK;
    }
 
    fe = shader->frontend;
    if (!(shader->frontend_data = fe->shader_init(shader->function,
            shader->functionLength, &shader->output_signature)))
    {
        FIXME("Failed to initialize frontend.\n");
        return WINED3DERR_INVALIDCALL;
    }
 
    if (FAILED(hr = shader_get_registers_used(shader, float_const_count)))
        return hr;
 
    shader->load_local_constsF = shader->lconst_inf_or_nan;
 
    if (version->type != type)
    {
        WARN("Wrong shader type %s.\n", debug_shader_type(reg_maps->shader_version.type));
        return WINED3DERR_INVALIDCALL;
    }
    if (!shader->is_ffp_vs && !shader->is_ffp_ps
            && version->major > shader_max_version_from_feature_level(shader->device->cs->c.state->feature_level))
    {
        WARN("Shader version %u not supported by this device.\n", version->major);
        return WINED3DERR_INVALIDCALL;
    }
    switch (type)
    {
        case WINED3D_SHADER_TYPE_VERTEX:
            for (unsigned int i = 0; i < shader->input_signature.element_count; ++i)
            {
                const struct wined3d_shader_signature_element *input = &shader->input_signature.elements[i];
 
                if (!(reg_maps->input_registers & (1u << input->register_idx)) || !input->semantic_name)
                    continue;
 
                shader->u.vs.attributes[input->register_idx].usage =
                        shader_usage_from_semantic_name(input->semantic_name);
                shader->u.vs.attributes[input->register_idx].usage_idx = input->semantic_idx;
            }
 
            if (reg_maps->usesrelconstF && !list_empty(&shader->constantsF))
                shader->load_local_constsF = TRUE;
 
            backend_version = d3d_info->limits.vs_version;
            break;
 
        case WINED3D_SHADER_TYPE_HULL:
            backend_version = d3d_info->limits.hs_version;
            break;
        case WINED3D_SHADER_TYPE_DOMAIN:
            backend_version = d3d_info->limits.ds_version;
            break;
        case WINED3D_SHADER_TYPE_GEOMETRY:
            backend_version = d3d_info->limits.gs_version;
            break;
        case WINED3D_SHADER_TYPE_PIXEL:
            for (unsigned int i = 0; i < MAX_REG_INPUT; ++i)
            {
                if (shader->u.ps.input_reg_used & (1u << i))
                {
                    ++num_regs_used;
                    highest_reg_used = i;
                }
            }
 
            /* Don't do any register mapping magic if it is not needed,
             * or if we can't achieve anything anyway. */
            if (highest_reg_used < (d3d_info->limits.varying_count / 4)
                    || num_regs_used > (d3d_info->limits.varying_count / 4)
                    || shader->reg_maps.shader_version.major >= 4)
            {
                if (num_regs_used > (d3d_info->limits.varying_count / 4))
                {
                    /* This happens with relative addressing. The input mapper
                     * function warns about this if the higher registers are
                     * declared too, so don't write a FIXME here. */
                    WARN("More varying registers used than supported.\n");
                }
 
                for (unsigned int i = 0; i < MAX_REG_INPUT; ++i)
                    shader->u.ps.input_reg_map[i] = i;
 
                shader->u.ps.declared_in_count = highest_reg_used + 1;
            }
            else
            {
                shader->u.ps.declared_in_count = 0;
                for (unsigned int i = 0; i < MAX_REG_INPUT; ++i)
                {
                    if (shader->u.ps.input_reg_used & (1u << i))
                        shader->u.ps.input_reg_map[i] = shader->u.ps.declared_in_count++;
                    else
                        shader->u.ps.input_reg_map[i] = ~0u;
                }
            }
 
            backend_version = d3d_info->limits.ps_version;
            break;
        case WINED3D_SHADER_TYPE_COMPUTE:
            backend_version = d3d_info->limits.cs_version;
            break;
        default:
            FIXME("No backend version-checking for this shader type.\n");
            backend_version = 0;
    }
    if (version->major > backend_version)
    {
        WARN("Shader version %u.%u not supported by the current shader backend.\n",
                version->major, version->minor);
        return WINED3DERR_INVALIDCALL;
    }
 
    return WINED3D_OK;
}
 
ULONG CDECL wined3d_shader_incref(struct wined3d_shader *shader)
{
    unsigned int refcount = InterlockedIncrement(&shader->ref);
 
    TRACE("%p increasing refcount to %u.\n", shader, refcount);
 
    return refcount;
}
 
static void wined3d_shader_init_object(void *object)
{
    struct wined3d_shader *shader = object;
    struct wined3d_device *device = shader->device;
 
    TRACE("shader %p.\n", shader);
 
    list_add_head(&device->shaders, &shader->shader_list_entry);
 
    if (shader->is_ffp_vs)
    {
        struct wined3d_ffp_vs_settings *settings = shader->byte_code;
        struct wined3d_shader_desc desc;
 
        if (!ffp_hlsl_compile_vs(settings, &desc, device))
            return;
        free(settings);
        shader_set_function(shader, &desc, WINED3D_SHADER_TYPE_VERTEX, NULL,
                device->adapter->d3d_info.limits.vs_uniform_count);
    }
 
    if (shader->is_ffp_ps)
    {
        struct ffp_frag_settings *settings = shader->byte_code;
        struct wined3d_shader_desc desc;
 
        if (!ffp_hlsl_compile_ps(settings, &desc))
            return;
        free(settings);
        shader_set_function(shader, &desc, WINED3D_SHADER_TYPE_PIXEL, NULL,
                device->adapter->d3d_info.limits.ps_uniform_count);
    }
 
    device->shader_backend->shader_precompile(device->shader_priv, shader);
}
 
static void wined3d_shader_destroy_object(void *object)
{
    TRACE("object %p.\n", object);
 
    shader_cleanup(object);
    free(object);
}
 
ULONG CDECL wined3d_shader_decref(struct wined3d_shader *shader)
{
    unsigned int refcount = InterlockedDecrement(&shader->ref);
 
    TRACE("%p decreasing refcount to %u.\n", shader, refcount);
 
    if (!refcount)
    {
        wined3d_mutex_lock();
        shader->parent_ops->wined3d_object_destroyed(shader->parent);
        wined3d_cs_destroy_object(shader->device->cs, wined3d_shader_destroy_object, shader);
        wined3d_mutex_unlock();
    }
 
    return refcount;
}
 
void * CDECL wined3d_shader_get_parent(const struct wined3d_shader *shader)
{
    TRACE("shader %p.\n", shader);
 
    return shader->parent;
}
 
HRESULT CDECL wined3d_shader_get_byte_code(const struct wined3d_shader *shader,
        void *byte_code, UINT *byte_code_size)
{
    TRACE("shader %p, byte_code %p, byte_code_size %p.\n", shader, byte_code, byte_code_size);
 
    if (!byte_code)
    {
        *byte_code_size = shader->byte_code_size;
        return WINED3D_OK;
    }
 
    if (*byte_code_size < shader->byte_code_size)
    {
        /* MSDN claims (for d3d8 at least) that if *byte_code_size is smaller
         * than the required size we should write the required size and
         * return D3DERR_MOREDATA. That's not actually true. */
        return WINED3DERR_INVALIDCALL;
    }
 
    memcpy(byte_code, shader->byte_code, shader->byte_code_size);
 
    return WINED3D_OK;
}
 
/* Set local constants for d3d8 shaders. */
HRESULT CDECL wined3d_shader_set_local_constants_float(struct wined3d_shader *shader,
        UINT start_idx, const float *src_data, UINT count)
{
    UINT end_idx = start_idx + count;
    UINT i;
 
    TRACE("shader %p, start_idx %u, src_data %p, count %u.\n", shader, start_idx, src_data, count);
 
    if (end_idx > shader->limits->constant_float)
    {
        WARN("end_idx %u > float constants limit %u.\n",
                end_idx, shader->limits->constant_float);
        end_idx = shader->limits->constant_float;
    }
 
    for (i = start_idx; i < end_idx; ++i)
    {
        struct wined3d_shader_lconst *lconst;
        float *value;
 
        if (!(lconst = malloc(sizeof(*lconst))))
            return E_OUTOFMEMORY;
 
        lconst->idx = i;
        value = (float *)lconst->value;
        memcpy(value, src_data + (i - start_idx) * 4 /* 4 components */, 4 * sizeof(float));
        list_add_head(&shader->constantsF, &lconst->entry);
 
        if (isinf(value[0]) || isnan(value[0]) || isinf(value[1]) || isnan(value[1])
                || isinf(value[2]) || isnan(value[2]) || isinf(value[3]) || isnan(value[3]))
        {
            shader->lconst_inf_or_nan = TRUE;
        }
    }
 
    return WINED3D_OK;
}
 
static void init_interpolation_compile_args(uint32_t *interpolation_args,
        const struct wined3d_shader *pixel_shader, const struct wined3d_d3d_info *d3d_info)
{
    if (!d3d_info->shader_output_interpolation || !pixel_shader
            || pixel_shader->reg_maps.shader_version.major < 4)
    {
        memset(interpolation_args, 0, sizeof(pixel_shader->u.ps.interpolation_mode));
        return;
    }
 
    memcpy(interpolation_args, pixel_shader->u.ps.interpolation_mode,
            sizeof(pixel_shader->u.ps.interpolation_mode));
}
 
void find_vs_compile_args(const struct wined3d_state *state, const struct wined3d_shader *shader,
        struct vs_compile_args *args, const struct wined3d_context *context)
{
    const struct wined3d_shader *geometry_shader = state->shader[WINED3D_SHADER_TYPE_GEOMETRY];
    const struct wined3d_shader *pixel_shader = state->shader[WINED3D_SHADER_TYPE_PIXEL];
    const struct wined3d_shader *hull_shader = state->shader[WINED3D_SHADER_TYPE_HULL];
    const struct wined3d_d3d_info *d3d_info = context->d3d_info;
    WORD swizzle_map = context->stream_info.swizzle_map;
 
    if (state->render_states[WINED3D_RS_FOGTABLEMODE] != WINED3D_FOG_NONE)
    {
        if (state->transforms[WINED3D_TS_PROJECTION]._14 == 0.0f
                && state->transforms[WINED3D_TS_PROJECTION]._24 == 0.0f
                && state->transforms[WINED3D_TS_PROJECTION]._34 == 0.0f
                && state->transforms[WINED3D_TS_PROJECTION]._44 == 1.0f)
        {
            /* Fog source is vertex output Z.
             *
             * However, if drawing RHW (which means we are using an HLSL
             * replacement shader, since we got here), and depth testing is
             * disabled, primitives are not supposed to be clipped by the
             * viewport. We handle this in the vertex shader by essentially
             * flushing output Z to zero.
             *
             * Fog needs to still read from the original Z, however. In this
             * case we read from oFog, which contains the original Z. */
 
            if (state->vertex_declaration->position_transformed)
                args->fog_src = VS_FOG_COORD;
            else
                args->fog_src = VS_FOG_Z;
        }
        else
        {
            args->fog_src = VS_FOG_W;
        }
    }
    else
    {
        args->fog_src = VS_FOG_COORD;
    }
 
    args->clip_enabled = state->render_states[WINED3D_RS_CLIPPING]
            && state->render_states[WINED3D_RS_CLIPPLANEENABLE];
    args->point_size = state->primitive_type == WINED3D_PT_POINTLIST;
    args->next_shader_type = hull_shader ? WINED3D_SHADER_TYPE_HULL
            : geometry_shader ? WINED3D_SHADER_TYPE_GEOMETRY : WINED3D_SHADER_TYPE_PIXEL;
    if (shader->reg_maps.shader_version.major >= 4)
        args->next_shader_input_count = hull_shader ? hull_shader->limits->packed_input
                : geometry_shader ? geometry_shader->limits->packed_input
                : pixel_shader ? pixel_shader->limits->packed_input : 0;
    else
        args->next_shader_input_count = 0;
    args->swizzle_map = swizzle_map;
    if (d3d_info->emulated_flatshading)
        args->flatshading = state->render_states[WINED3D_RS_SHADEMODE] == WINED3D_SHADE_FLAT;
    else
        args->flatshading = 0;
 
    init_interpolation_compile_args(args->interpolation_mode,
            args->next_shader_type == WINED3D_SHADER_TYPE_PIXEL ? pixel_shader : NULL, d3d_info);
}
 
static BOOL match_usage(BYTE usage1, BYTE usage_idx1, BYTE usage2, BYTE usage_idx2)
{
    if (usage_idx1 != usage_idx2)
        return FALSE;
    if (usage1 == usage2)
        return TRUE;
    if (usage1 == WINED3D_DECL_USAGE_POSITION && usage2 == WINED3D_DECL_USAGE_POSITIONT)
        return TRUE;
    if (usage2 == WINED3D_DECL_USAGE_POSITION && usage1 == WINED3D_DECL_USAGE_POSITIONT)
        return TRUE;
 
    return FALSE;
}
 
bool vshader_get_input(const struct wined3d_shader *shader,
        uint8_t usage_req, uint8_t usage_idx_req, unsigned int *regnum)
{
    uint32_t map = shader->reg_maps.input_registers & 0xffff;
    unsigned int i;
 
    while (map)
    {
        i = wined3d_bit_scan(&map);
        if (match_usage(shader->u.vs.attributes[i].usage,
                shader->u.vs.attributes[i].usage_idx, usage_req, usage_idx_req))
        {
            *regnum = i;
            return true;
        }
    }
 
    return false;
}
 
static void shader_init(struct wined3d_shader *shader, struct wined3d_device *device,
        void *parent, const struct wined3d_parent_ops *parent_ops)
{
    shader->ref = 1;
    shader->device = device;
    shader->parent = parent;
    shader->parent_ops = parent_ops;
 
    list_init(&shader->linked_programs);
    list_init(&shader->constantsF);
    list_init(&shader->constantsB);
    list_init(&shader->constantsI);
    shader->lconst_inf_or_nan = FALSE;
    list_init(&shader->reg_maps.indexable_temps);
    list_init(&shader->shader_list_entry);
}
 
static HRESULT vertex_shader_init(struct wined3d_shader *shader, struct wined3d_device *device,
        const struct wined3d_shader_desc *desc, void *parent, const struct wined3d_parent_ops *parent_ops)
{
    HRESULT hr;
 
    shader_init(shader, device, parent, parent_ops);
 
    if (FAILED(hr = shader_set_function(shader, desc,
            WINED3D_SHADER_TYPE_VERTEX, NULL, device->adapter->d3d_info.limits.vs_uniform_count)))
    {
        shader_cleanup(shader);
        return hr;
    }
 
    return WINED3D_OK;
}
 
static HRESULT geometry_shader_init(struct wined3d_shader *shader, struct wined3d_device *device,
        const struct wined3d_shader_desc *desc, const struct wined3d_stream_output_desc *so_desc,
        void *parent, const struct wined3d_parent_ops *parent_ops)
{
    HRESULT hr;
 
    shader_init(shader, device, parent, parent_ops);
 
    if (FAILED(hr = shader_set_function(shader, desc, WINED3D_SHADER_TYPE_GEOMETRY, so_desc, 0)))
        goto fail;
 
    return WINED3D_OK;
 
fail:
    shader_cleanup(shader);
    return hr;
}
 
void find_ds_compile_args(const struct wined3d_state *state, const struct wined3d_shader *shader,
        struct ds_compile_args *args, const struct wined3d_context *context)
{
    const struct wined3d_shader *geometry_shader = state->shader[WINED3D_SHADER_TYPE_GEOMETRY];
    const struct wined3d_shader *pixel_shader = state->shader[WINED3D_SHADER_TYPE_PIXEL];
    const struct wined3d_shader *hull_shader = state->shader[WINED3D_SHADER_TYPE_HULL];
 
    args->tessellator_output_primitive = hull_shader->u.hs.tessellator_output_primitive;
    args->tessellator_partitioning = hull_shader->u.hs.tessellator_partitioning;
 
    args->output_count = geometry_shader ? geometry_shader->limits->packed_input
            : pixel_shader ? pixel_shader->limits->packed_input : shader->limits->packed_output;
    args->next_shader_type = geometry_shader ? WINED3D_SHADER_TYPE_GEOMETRY : WINED3D_SHADER_TYPE_PIXEL;
 
    init_interpolation_compile_args(args->interpolation_mode,
            args->next_shader_type == WINED3D_SHADER_TYPE_PIXEL ? pixel_shader : NULL, context->d3d_info);
 
    args->padding = 0;
}
 
void find_gs_compile_args(const struct wined3d_state *state, const struct wined3d_shader *shader,
        struct gs_compile_args *args, const struct wined3d_context *context)
{
    const struct wined3d_shader *pixel_shader = state->shader[WINED3D_SHADER_TYPE_PIXEL];
 
    args->output_count = pixel_shader ? pixel_shader->limits->packed_input : shader->limits->packed_output;
 
    if (!(args->primitive_type = shader->u.gs.input_type))
        args->primitive_type = state->primitive_type;
 
    init_interpolation_compile_args(args->interpolation_mode, pixel_shader, context->d3d_info);
}
 
void find_ps_compile_args(const struct wined3d_state *state, const struct wined3d_shader *shader,
        BOOL position_transformed, struct ps_compile_args *args, const struct wined3d_context *context)
{
    const struct wined3d_shader *vs = state->shader[WINED3D_SHADER_TYPE_VERTEX];
    const struct wined3d_d3d_info *d3d_info = context->d3d_info;
    struct wined3d_shader_resource_view *view;
    struct wined3d_texture *texture;
    unsigned int i;
 
    memset(args, 0, sizeof(*args)); /* FIXME: Make sure all bits are set. */
    if (!d3d_info->srgb_write_control && needs_srgb_write(d3d_info, state, &state->fb))
    {
        static unsigned int warned = 0;
 
        args->srgb_correction = 1;
        if (state->blend_state && state->blend_state->desc.rt[0].enable && !warned++)
            WARN("Blending into a sRGB render target with no GL_ARB_framebuffer_sRGB "
                    "support, expect rendering artifacts.\n");
    }
 
    if (shader->reg_maps.shader_version.major == 1
            && shader->reg_maps.shader_version.minor <= 3)
    {
        for (i = 0; i < shader->limits->sampler; ++i)
        {
            uint32_t flags = state->texture_states[i][WINED3D_TSS_TEXTURE_TRANSFORM_FLAGS];
 
            if (flags & WINED3D_TTFF_PROJECTED)
            {
                uint32_t tex_transform = flags & ~WINED3D_TTFF_PROJECTED;
 
                if (!vs || vs->is_ffp_vs)
                {
                    enum wined3d_shader_resource_type resource_type = shader->reg_maps.resource_info[i].type;
                    unsigned int j;
                    unsigned int index = state->texture_states[i][WINED3D_TSS_TEXCOORD_INDEX];
                    uint32_t max_valid = WINED3D_TTFF_COUNT4;
 
                    for (j = 0; j < state->vertex_declaration->element_count; ++j)
                    {
                        struct wined3d_vertex_declaration_element *element =
                                &state->vertex_declaration->elements[j];
 
                        if (element->usage == WINED3D_DECL_USAGE_TEXCOORD
                                && element->usage_idx == index)
                        {
                            max_valid = element->format->component_count;
                            break;
                        }
                    }
                    if (!tex_transform || tex_transform > max_valid)
                    {
                        WARN("Fixing up projected texture transform flags from %#x to %#x.\n",
                                tex_transform, max_valid);
                        tex_transform = max_valid;
                    }
                    if ((resource_type == WINED3D_SHADER_RESOURCE_TEXTURE_1D && tex_transform > WINED3D_TTFF_COUNT1)
                            || (resource_type == WINED3D_SHADER_RESOURCE_TEXTURE_2D
                            && tex_transform > WINED3D_TTFF_COUNT2)
                            || (resource_type == WINED3D_SHADER_RESOURCE_TEXTURE_3D
                            && tex_transform > WINED3D_TTFF_COUNT3))
                        tex_transform |= WINED3D_PSARGS_PROJECTED;
                    else
                    {
                        WARN("Application requested projected texture with unsuitable texture coordinates.\n");
                        WARN("(texture unit %u, transform flags %#x, sampler type %u).\n",
                                i, tex_transform, resource_type);
                    }
                }
                else
                    tex_transform = WINED3D_TTFF_COUNT4 | WINED3D_PSARGS_PROJECTED;
 
                args->tex_transform |= tex_transform << i * WINED3D_PSARGS_TEXTRANSFORM_SHIFT;
            }
        }
    }
    if (shader->reg_maps.shader_version.major == 1
            && shader->reg_maps.shader_version.minor <= 4)
    {
        for (i = 0; i < shader->limits->sampler; ++i)
        {
            enum wined3d_shader_tex_types type = WINED3D_SHADER_TEX_2D;
 
            if (!shader->reg_maps.resource_info[i].type)
                continue;
 
            /* Treat unbound textures as 2D. The dummy texture will provide
             * the proper sample value. The tex_types bitmap defaults to
             * 2D because of the memset. */
            if (!(view = state->shader_resource_view[WINED3D_SHADER_TYPE_PIXEL][i]))
                continue;
            texture = texture_from_resource(view->resource);
 
            switch (texture->resource.type)
            {
                case WINED3D_RTYPE_NONE:
                case WINED3D_RTYPE_BUFFER:
                case WINED3D_RTYPE_TEXTURE_1D:
                    assert(0);
 
                case WINED3D_RTYPE_TEXTURE_2D:
                    if (texture->resource.usage & WINED3DUSAGE_LEGACY_CUBEMAP)
                        type = WINED3D_SHADER_TEX_CUBE;
                    break;
 
                case WINED3D_RTYPE_TEXTURE_3D:
                    type = WINED3D_SHADER_TEX_3D;
                    break;
            }
 
            args->tex_types |= (type << (i * WINED3D_PSARGS_TEXTYPE_SHIFT));
        }
    }
    else if (shader->reg_maps.shader_version.major <= 3)
    {
        for (i = 0; i < shader->limits->sampler; ++i)
        {
            enum wined3d_shader_resource_type resource_type;
            enum wined3d_shader_tex_types tex_type;
 
            if (!(resource_type = shader->reg_maps.resource_info[i].type))
                continue;
 
            switch (resource_type)
            {
                case WINED3D_SHADER_RESOURCE_TEXTURE_3D:
                    tex_type = WINED3D_SHADER_TEX_3D;
                    break;
                case WINED3D_SHADER_RESOURCE_TEXTURE_CUBE:
                    tex_type = WINED3D_SHADER_TEX_CUBE;
                    break;
                default:
                    tex_type = WINED3D_SHADER_TEX_2D;
                    break;
            }
 
            if ((view = state->shader_resource_view[WINED3D_SHADER_TYPE_PIXEL][i]))
            {
                texture = texture_from_resource(view->resource);
                /* Star Wars: The Old Republic uses mismatched samplers for rendering water. */
                if (texture->resource.type == WINED3D_RTYPE_TEXTURE_2D
                        && resource_type == WINED3D_SHADER_RESOURCE_TEXTURE_3D
                        && !(texture->resource.usage & WINED3DUSAGE_LEGACY_CUBEMAP))
                    tex_type = WINED3D_SHADER_TEX_2D;
                else if (texture->resource.type == WINED3D_RTYPE_TEXTURE_3D
                        && resource_type == WINED3D_SHADER_RESOURCE_TEXTURE_2D)
                    tex_type = WINED3D_SHADER_TEX_3D;
            }
            args->tex_types |= tex_type << i * WINED3D_PSARGS_TEXTYPE_SHIFT;
        }
    }
 
    if (shader->reg_maps.shader_version.major >= 4)
    {
        /* In SM4+ we use dcl_sampler in order to determine if we should use shadow sampler. */
        args->shadow = 0;
        for (i = 0 ; i < WINED3D_MAX_FRAGMENT_SAMPLERS; ++i)
            args->color_fixup[i] = COLOR_FIXUP_IDENTITY;
    }
    else
    {
        for (i = 0; i < WINED3D_MAX_FRAGMENT_SAMPLERS; ++i)
        {
            if (!shader->reg_maps.resource_info[i].type)
                continue;
 
            if (!(view = state->shader_resource_view[WINED3D_SHADER_TYPE_PIXEL][i]))
            {
                args->color_fixup[i] = COLOR_FIXUP_IDENTITY;
                continue;
            }
            texture = texture_from_resource(view->resource);
 
            if (can_use_texture_swizzle(d3d_info, texture->resource.format))
                args->color_fixup[i] = COLOR_FIXUP_IDENTITY;
            else
                args->color_fixup[i] = texture->resource.format->color_fixup;
 
            if (texture->resource.format_caps & WINED3D_FORMAT_CAP_SHADOW)
                args->shadow |= 1u << i;
        }
    }
 
    if (shader->reg_maps.shader_version.major >= 3)
    {
        if (position_transformed)
            args->vp_mode = WINED3D_VP_MODE_NONE;
        else if (use_vs(state))
            args->vp_mode = WINED3D_VP_MODE_SHADER;
        else
            args->vp_mode = WINED3D_VP_MODE_FF;
        args->fog = WINED3D_FFP_PS_FOG_OFF;
    }
    else
    {
        args->vp_mode = WINED3D_VP_MODE_SHADER;
        if (state->render_states[WINED3D_RS_FOGENABLE])
        {
            switch (state->render_states[WINED3D_RS_FOGTABLEMODE])
            {
                case WINED3D_FOG_NONE:
                case WINED3D_FOG_LINEAR: args->fog = WINED3D_FFP_PS_FOG_LINEAR; break;
                case WINED3D_FOG_EXP:    args->fog = WINED3D_FFP_PS_FOG_EXP;    break;
                case WINED3D_FOG_EXP2:   args->fog = WINED3D_FFP_PS_FOG_EXP2;   break;
            }
        }
        else
        {
            args->fog = WINED3D_FFP_PS_FOG_OFF;
        }
    }
 
    if (!d3d_info->full_ffp_varyings)
    {
        const struct wined3d_shader *vs = state->shader[WINED3D_SHADER_TYPE_VERTEX];
 
        args->texcoords_initialized = 0;
        for (i = 0; i < WINED3D_MAX_FFP_TEXTURES; ++i)
        {
            if (vs)
            {
                if (state->shader[WINED3D_SHADER_TYPE_VERTEX]->reg_maps.output_registers & (1u << i))
                    args->texcoords_initialized |= 1u << i;
            }
            else
            {
                const struct wined3d_stream_info *si = &context->stream_info;
                unsigned int coord_idx = state->texture_states[i][WINED3D_TSS_TEXCOORD_INDEX];
 
                if ((state->texture_states[i][WINED3D_TSS_TEXCOORD_INDEX] >> WINED3D_FFP_TCI_SHIFT)
                        & WINED3D_FFP_TCI_MASK
                        || (coord_idx < WINED3D_MAX_FFP_TEXTURES && (si->use_map & (1u << (WINED3D_FFP_TEXCOORD0 + coord_idx)))))
                    args->texcoords_initialized |= 1u << i;
            }
        }
    }
    else
    {
        args->texcoords_initialized = wined3d_mask_from_size(WINED3D_MAX_FFP_TEXTURES);
    }
 
    args->pointsprite = state->render_states[WINED3D_RS_POINTSPRITEENABLE]
            && state->primitive_type == WINED3D_PT_POINTLIST;
 
    if (d3d_info->ffp_alpha_test)
        args->alpha_test_func = WINED3D_CMP_ALWAYS - 1;
    else
        args->alpha_test_func = (state->render_states[WINED3D_RS_ALPHATESTENABLE]
                ? wined3d_sanitize_cmp_func(state->render_states[WINED3D_RS_ALPHAFUNC])
                : WINED3D_CMP_ALWAYS) - 1;
 
    if (d3d_info->emulated_flatshading)
        args->flatshading = state->render_states[WINED3D_RS_SHADEMODE] == WINED3D_SHADE_FLAT;
 
    for (i = 0; i < ARRAY_SIZE(state->fb.render_targets); ++i)
    {
        struct wined3d_rendertarget_view *rtv = state->fb.render_targets[i];
        if (rtv && rtv->format->id == WINED3DFMT_A8_UNORM && !is_identity_fixup(rtv->format->color_fixup))
            args->rt_alpha_swizzle |= 1u << i;
    }
 
    args->dual_source_blend = state->blend_state && state->blend_state->dual_source;
}
 
static HRESULT pixel_shader_init(struct wined3d_shader *shader, struct wined3d_device *device,
        const struct wined3d_shader_desc *desc, void *parent, const struct wined3d_parent_ops *parent_ops)
{
    const struct wined3d_d3d_info *d3d_info = &device->adapter->d3d_info;
    HRESULT hr;
 
    shader_init(shader, device, parent, parent_ops);
 
    if (FAILED(hr = shader_set_function(shader, desc,
            WINED3D_SHADER_TYPE_PIXEL, NULL, d3d_info->limits.ps_uniform_count)))
    {
        shader_cleanup(shader);
        return hr;
    }
 
    return WINED3D_OK;
}
 
enum wined3d_shader_resource_type pixelshader_get_resource_type(const struct wined3d_shader_reg_maps *reg_maps,
        unsigned int resource_idx, DWORD tex_types)
{
    static enum wined3d_shader_resource_type shader_resource_type_from_shader_tex_types[] =
    {
        WINED3D_SHADER_RESOURCE_TEXTURE_2D,     /* WINED3D_SHADER_TEX_2D     */
        WINED3D_SHADER_RESOURCE_TEXTURE_3D,     /* WINED3D_SHADER_TEX_3D     */
        WINED3D_SHADER_RESOURCE_TEXTURE_CUBE,   /* WINED3D_SHADER_TEX_CUBE   */
    };
 
    unsigned int idx;
 
    if (reg_maps->shader_version.major > 3)
        return reg_maps->resource_info[resource_idx].type;
 
    if (!reg_maps->resource_info[resource_idx].type)
        return 0;
 
    idx = (tex_types >> resource_idx * WINED3D_PSARGS_TEXTYPE_SHIFT) & WINED3D_PSARGS_TEXTYPE_MASK;
    assert(idx < ARRAY_SIZE(shader_resource_type_from_shader_tex_types));
    return shader_resource_type_from_shader_tex_types[idx];
}
 
HRESULT CDECL wined3d_shader_create_cs(struct wined3d_device *device, const struct wined3d_shader_desc *desc,
        void *parent, const struct wined3d_parent_ops *parent_ops, struct wined3d_shader **shader)
{
    struct wined3d_shader *object;
    HRESULT hr;
 
    TRACE("device %p, desc %p, parent %p, parent_ops %p, shader %p.\n",
            device, desc, parent, parent_ops, shader);
 
    if (!(object = calloc(1, sizeof(*object))))
        return E_OUTOFMEMORY;
 
    shader_init(object, device, parent, parent_ops);
 
    if (FAILED(hr = shader_set_function(object, desc, WINED3D_SHADER_TYPE_COMPUTE, NULL, 0)))
    {
        shader_cleanup(object);
        free(object);
        return hr;
    }
 
    wined3d_cs_init_object(device->cs, wined3d_shader_init_object, object);
 
    TRACE("Created compute shader %p.\n", object);
    *shader = object;
 
    return WINED3D_OK;
}
 
HRESULT CDECL wined3d_shader_create_ds(struct wined3d_device *device, const struct wined3d_shader_desc *desc,
        void *parent, const struct wined3d_parent_ops *parent_ops, struct wined3d_shader **shader)
{
    struct wined3d_shader *object;
    HRESULT hr;
 
    TRACE("device %p, desc %p, parent %p, parent_ops %p, shader %p.\n",
            device, desc, parent, parent_ops, shader);
 
    if (!(object = calloc(1, sizeof(*object))))
        return E_OUTOFMEMORY;
 
    shader_init(object, device, parent, parent_ops);
 
    if (FAILED(hr = shader_set_function(object, desc, WINED3D_SHADER_TYPE_DOMAIN, NULL, 0)))
    {
        shader_cleanup(object);
        free(object);
        return hr;
    }
 
    wined3d_cs_init_object(device->cs, wined3d_shader_init_object, object);
 
    TRACE("Created domain shader %p.\n", object);
    *shader = object;
 
    return WINED3D_OK;
}
 
HRESULT CDECL wined3d_shader_create_gs(struct wined3d_device *device, const struct wined3d_shader_desc *desc,
        const struct wined3d_stream_output_desc *so_desc, void *parent,
        const struct wined3d_parent_ops *parent_ops, struct wined3d_shader **shader)
{
    struct wined3d_shader *object;
    HRESULT hr;
 
    TRACE("device %p, desc %p, so_desc %p, parent %p, parent_ops %p, shader %p.\n",
            device, desc, so_desc, parent, parent_ops, shader);
 
    if (!(object = calloc(1, sizeof(*object))))
        return E_OUTOFMEMORY;
 
    if (FAILED(hr = geometry_shader_init(object, device, desc, so_desc, parent, parent_ops)))
    {
        WARN("Failed to initialize geometry shader, hr %#lx.\n", hr);
        free(object);
        return hr;
    }
 
    wined3d_cs_init_object(device->cs, wined3d_shader_init_object, object);
 
    TRACE("Created geometry shader %p.\n", object);
    *shader = object;
 
    return WINED3D_OK;
}
 
HRESULT CDECL wined3d_shader_create_hs(struct wined3d_device *device, const struct wined3d_shader_desc *desc,
        void *parent, const struct wined3d_parent_ops *parent_ops, struct wined3d_shader **shader)
{
    struct wined3d_shader *object;
    HRESULT hr;
 
    TRACE("device %p, desc %p, parent %p, parent_ops %p, shader %p.\n",
            device, desc, parent, parent_ops, shader);
 
    if (!(object = calloc(1, sizeof(*object))))
        return E_OUTOFMEMORY;
 
    shader_init(object, device, parent, parent_ops);
 
    if (FAILED(hr = shader_set_function(object, desc, WINED3D_SHADER_TYPE_HULL, NULL, 0)))
    {
        shader_cleanup(object);
        free(object);
        return hr;
    }
 
    wined3d_cs_init_object(device->cs, wined3d_shader_init_object, object);
 
    TRACE("Created hull shader %p.\n", object);
    *shader = object;
 
    return WINED3D_OK;
}
 
HRESULT CDECL wined3d_shader_create_ps(struct wined3d_device *device, const struct wined3d_shader_desc *desc,
        void *parent, const struct wined3d_parent_ops *parent_ops, struct wined3d_shader **shader)
{
    struct wined3d_shader *object;
    HRESULT hr;
 
    TRACE("device %p, desc %p, parent %p, parent_ops %p, shader %p.\n",
            device, desc, parent, parent_ops, shader);
 
    if (!(object = calloc(1, sizeof(*object))))
        return E_OUTOFMEMORY;
 
    if (FAILED(hr = pixel_shader_init(object, device, desc, parent, parent_ops)))
    {
        WARN("Failed to initialize pixel shader, hr %#lx.\n", hr);
        free(object);
        return hr;
    }
 
    wined3d_cs_init_object(device->cs, wined3d_shader_init_object, object);
 
    TRACE("Created pixel shader %p.\n", object);
    *shader = object;
 
    return WINED3D_OK;
}
 
HRESULT CDECL wined3d_shader_create_vs(struct wined3d_device *device, const struct wined3d_shader_desc *desc,
        void *parent, const struct wined3d_parent_ops *parent_ops, struct wined3d_shader **shader)
{
    struct wined3d_shader *object;
    HRESULT hr;
 
    TRACE("device %p, desc %p, parent %p, parent_ops %p, shader %p.\n",
            device, desc, parent, parent_ops, shader);
 
    if (!(object = calloc(1, sizeof(*object))))
        return E_OUTOFMEMORY;
 
    if (FAILED(hr = vertex_shader_init(object, device, desc, parent, parent_ops)))
    {
        WARN("Failed to initialize vertex shader, hr %#lx.\n", hr);
        free(object);
        return hr;
    }
 
    wined3d_cs_init_object(device->cs, wined3d_shader_init_object, object);
 
    TRACE("Created vertex shader %p.\n", object);
    *shader = object;
 
    return WINED3D_OK;
}
 
HRESULT wined3d_shader_create_ffp_vs(struct wined3d_device *device,
        const struct wined3d_ffp_vs_settings *settings, struct wined3d_shader **shader)
{
    struct wined3d_shader *object;
 
    if (!(object = calloc(1, sizeof(*object))))
        return E_OUTOFMEMORY;
 
    shader_init(object, device, NULL, &wined3d_null_parent_ops);
    object->is_ffp_vs = true;
    if (!(object->byte_code = malloc(sizeof(*settings))))
    {
        free(object);
        return E_OUTOFMEMORY;
    }
    memcpy(object->byte_code, settings, sizeof(*settings));
 
    wined3d_cs_init_object(device->cs, wined3d_shader_init_object, object);
 
    TRACE("Created FFP vertex shader %p.\n", object);
    *shader = object;
 
    return WINED3D_OK;
}
 
HRESULT wined3d_shader_create_ffp_ps(struct wined3d_device *device,
        const struct ffp_frag_settings *settings, struct wined3d_shader **shader)
{
    struct wined3d_shader *object;
 
    if (!(object = calloc(1, sizeof(*object))))
        return E_OUTOFMEMORY;
 
    shader_init(object, device, NULL, &wined3d_null_parent_ops);
    object->is_ffp_ps = true;
    if (!(object->byte_code = malloc(sizeof(*settings))))
    {
        free(object);
        return E_OUTOFMEMORY;
    }
    memcpy(object->byte_code, settings, sizeof(*settings));
 
    wined3d_cs_init_object(device->cs, wined3d_shader_init_object, object);
 
    TRACE("Created FFP pixel shader %p.\n", object);
    *shader = object;
 
    return WINED3D_OK;
}