From d9e396ce4a124529fa92ad967f2b3ff72534079b Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Sun, 21 Feb 2010 02:26:25 +0100 Subject: nv30, nv40: non-trivially partially unify nv[34]0_shader.h shader.h is similar, except for the following differences: 1. The instruction sets are not exactly the same, but mostly similar 2. Vertex program fields are in different bit positions This patch unifies all parts of nv[34]0_shader.h except the vertex program fields. Vertex opcodes are also changed so that the constant names includes SCA if it is a scalar opcode and VEC if it is a vector opcode. --- src/gallium/drivers/nv30/nv30_shader.h | 323 +-------------------------------- 1 file changed, 1 insertion(+), 322 deletions(-) (limited to 'src/gallium/drivers/nv30/nv30_shader.h') diff --git a/src/gallium/drivers/nv30/nv30_shader.h b/src/gallium/drivers/nv30/nv30_shader.h index dd3a36f78f..f19efb5aa4 100644 --- a/src/gallium/drivers/nv30/nv30_shader.h +++ b/src/gallium/drivers/nv30/nv30_shader.h @@ -72,14 +72,6 @@ #define NV30_VP_INST_COND_TEST_ENABLE (1<<14) #define NV30_VP_INST_COND_SHIFT 11 #define NV30_VP_INST_COND_MASK (0x07 << 11) -# define NV30_VP_INST_COND_FL 0 /* guess */ -# define NV30_VP_INST_COND_LT 1 -# define NV30_VP_INST_COND_EQ 2 -# define NV30_VP_INST_COND_LE 3 -# define NV30_VP_INST_COND_GT 4 -# define NV30_VP_INST_COND_NE 5 -# define NV30_VP_INST_COND_GE 6 -# define NV30_VP_INST_COND_TR 7 /* guess */ #define NV30_VP_INST_COND_SWZ_X_SHIFT 9 #define NV30_VP_INST_COND_SWZ_X_MASK (0x03 << 9) #define NV30_VP_INST_COND_SWZ_Y_SHIFT 7 @@ -98,59 +90,12 @@ /* DWORD 1 */ #define NV30_VP_INST_SCA_OPCODEL_SHIFT 28 #define NV30_VP_INST_SCA_OPCODEL_MASK (0x0F << 28) -# define NV30_VP_INST_OP_NOP 0x00 -# define NV30_VP_INST_OP_RCP 0x02 -# define NV30_VP_INST_OP_RCC 0x03 -# define NV30_VP_INST_OP_RSQ 0x04 -# define NV30_VP_INST_OP_EXP 0x05 -# define NV30_VP_INST_OP_LOG 0x06 -# define NV30_VP_INST_OP_LIT 0x07 -# define NV30_VP_INST_OP_BRA 0x09 -# define NV30_VP_INST_OP_CAL 0x0B -# define NV30_VP_INST_OP_RET 0x0C -# define NV30_VP_INST_OP_LG2 0x0D -# define NV30_VP_INST_OP_EX2 0x0E -# define NV30_VP_INST_OP_SIN 0x0F -# define NV30_VP_INST_OP_COS 0x10 #define NV30_VP_INST_VEC_OPCODE_SHIFT 23 #define NV30_VP_INST_VEC_OPCODE_MASK (0x1F << 23) -# define NV30_VP_INST_OP_NOPV 0x00 -# define NV30_VP_INST_OP_MOV 0x01 -# define NV30_VP_INST_OP_MUL 0x02 -# define NV30_VP_INST_OP_ADD 0x03 -# define NV30_VP_INST_OP_MAD 0x04 -# define NV30_VP_INST_OP_DP3 0x05 -# define NV30_VP_INST_OP_DP4 0x07 -# define NV30_VP_INST_OP_DPH 0x06 -# define NV30_VP_INST_OP_DST 0x08 -# define NV30_VP_INST_OP_MIN 0x09 -# define NV30_VP_INST_OP_MAX 0x0A -# define NV30_VP_INST_OP_SLT 0x0B -# define NV30_VP_INST_OP_SGE 0x0C -# define NV30_VP_INST_OP_ARL 0x0D -# define NV30_VP_INST_OP_FRC 0x0E -# define NV30_VP_INST_OP_FLR 0x0F -# define NV30_VP_INST_OP_SEQ 0x10 -# define NV30_VP_INST_OP_SFL 0x11 -# define NV30_VP_INST_OP_SGT 0x12 -# define NV30_VP_INST_OP_SLE 0x13 -# define NV30_VP_INST_OP_SNE 0x14 -# define NV30_VP_INST_OP_STR 0x15 -# define NV30_VP_INST_OP_SSG 0x16 -# define NV30_VP_INST_OP_ARR 0x17 -# define NV30_VP_INST_OP_ARA 0x18 #define NV30_VP_INST_CONST_SRC_SHIFT 14 #define NV30_VP_INST_CONST_SRC_MASK (0xFF << 14) #define NV30_VP_INST_INPUT_SRC_SHIFT 9 /*NV20*/ #define NV30_VP_INST_INPUT_SRC_MASK (0x0F << 9) /*NV20*/ -# define NV30_VP_INST_IN_POS 0 /* These seem to match the bindings specified in */ -# define NV30_VP_INST_IN_WEIGHT 1 /* the ARB_v_p spec (2.14.3.1) */ -# define NV30_VP_INST_IN_NORMAL 2 -# define NV30_VP_INST_IN_COL0 3 /* Should probably confirm them all though */ -# define NV30_VP_INST_IN_COL1 4 -# define NV30_VP_INST_IN_FOGC 5 -# define NV30_VP_INST_IN_TC0 8 -# define NV30_VP_INST_IN_TC(n) (8+n) #define NV30_VP_INST_SRC0H_SHIFT 0 /*NV20*/ #define NV30_VP_INST_SRC0H_MASK (0x1FF << 0) /*NV20*/ @@ -190,8 +135,6 @@ # define NV30_VP_INST_DEST_PSZ 6 # define NV30_VP_INST_DEST_TC(n) (8+n) -#define NV30_VP_INST_LAST (1 << 0) - /* Useful to split the source selection regs into their pieces */ #define NV30_VP_SRC0_HIGH_SHIFT 6 #define NV30_VP_SRC0_HIGH_MASK 0x00007FC0 @@ -221,270 +164,6 @@ #define NV30_VP_SRC_REG_TYPE_INPUT 2 #define NV30_VP_SRC_REG_TYPE_CONST 3 /* guess */ -/* - * Each fragment program opcode appears to be comprised of 4 32-bit values. - * - * 0 - Opcode, output reg/mask, ATTRIB source - * 1 - Source 0 - * 2 - Source 1 - * 3 - Source 2 - * - * There appears to be no special difference between result regs and temp regs. - * result.color == R0.xyzw - * result.depth == R1.z - * When the fragprog contains instructions to write depth, NV30_TCL_PRIMITIVE_3D_UNK1D78=0 - * otherwise it is set to 1. - * - * Constants are inserted directly after the instruction that uses them. - * - * It appears that it's not possible to use two input registers in one - * instruction as the input sourcing is done in the instruction dword - * and not the source selection dwords. As such instructions such as: - * - * ADD result.color, fragment.color, fragment.texcoord[0]; - * - * must be split into two MOV's and then an ADD (nvidia does this) but - * I'm not sure why it's not just one MOV and then source the second input - * in the ADD instruction.. - * - * Negation of the full source is done with NV30_FP_REG_NEGATE, arbitrary - * negation requires multiplication with a const. - * - * Arbitrary swizzling is supported with the exception of SWIZZLE_ZERO/SWIZZLE_ONE - * The temp/result regs appear to be initialised to (0.0, 0.0, 0.0, 0.0) as SWIZZLE_ZERO - * is implemented simply by not writing to the relevant components of the destination. - * - * Conditional execution - * TODO - * - * Non-native instructions: - * LIT - * LRP - MAD+MAD - * SUB - ADD, negate second source - * RSQ - LG2 + EX2 - * POW - LG2 + MUL + EX2 - * SCS - COS + SIN - * XPD - */ - -//== Opcode / Destination selection == -#define NV30_FP_OP_PROGRAM_END (1 << 0) -#define NV30_FP_OP_OUT_REG_SHIFT 1 -#define NV30_FP_OP_OUT_REG_MASK (31 << 1) /* uncertain */ -/* Needs to be set when writing outputs to get expected result.. */ -#define NV30_FP_OP_OUT_REG_HALF (1 << 7) -#define NV30_FP_OP_COND_WRITE_ENABLE (1 << 8) -#define NV30_FP_OP_OUTMASK_SHIFT 9 -#define NV30_FP_OP_OUTMASK_MASK (0xF << 9) -# define NV30_FP_OP_OUT_X (1<<9) -# define NV30_FP_OP_OUT_Y (1<<10) -# define NV30_FP_OP_OUT_Z (1<<11) -# define NV30_FP_OP_OUT_W (1<<12) -/* Uncertain about these, especially the input_src values.. it's possible that - * they can be dynamically changed. - */ -#define NV30_FP_OP_INPUT_SRC_SHIFT 13 -#define NV30_FP_OP_INPUT_SRC_MASK (15 << 13) -# define NV30_FP_OP_INPUT_SRC_POSITION 0x0 -# define NV30_FP_OP_INPUT_SRC_COL0 0x1 -# define NV30_FP_OP_INPUT_SRC_COL1 0x2 -# define NV30_FP_OP_INPUT_SRC_FOGC 0x3 -# define NV30_FP_OP_INPUT_SRC_TC0 0x4 -# define NV30_FP_OP_INPUT_SRC_TC(n) (0x4 + n) -#define NV30_FP_OP_TEX_UNIT_SHIFT 17 -#define NV30_FP_OP_TEX_UNIT_MASK (0xF << 17) /* guess */ -#define NV30_FP_OP_PRECISION_SHIFT 22 -#define NV30_FP_OP_PRECISION_MASK (3 << 22) -# define NV30_FP_PRECISION_FP32 0 -# define NV30_FP_PRECISION_FP16 1 -# define NV30_FP_PRECISION_FX12 2 -#define NV30_FP_OP_OPCODE_SHIFT 24 -#define NV30_FP_OP_OPCODE_MASK (0x3F << 24) -# define NV30_FP_OP_OPCODE_NOP 0x00 -# define NV30_FP_OP_OPCODE_MOV 0x01 -# define NV30_FP_OP_OPCODE_MUL 0x02 -# define NV30_FP_OP_OPCODE_ADD 0x03 -# define NV30_FP_OP_OPCODE_MAD 0x04 -# define NV30_FP_OP_OPCODE_DP3 0x05 -# define NV30_FP_OP_OPCODE_DP4 0x06 -# define NV30_FP_OP_OPCODE_DST 0x07 -# define NV30_FP_OP_OPCODE_MIN 0x08 -# define NV30_FP_OP_OPCODE_MAX 0x09 -# define NV30_FP_OP_OPCODE_SLT 0x0A -# define NV30_FP_OP_OPCODE_SGE 0x0B -# define NV30_FP_OP_OPCODE_SLE 0x0C -# define NV30_FP_OP_OPCODE_SGT 0x0D -# define NV30_FP_OP_OPCODE_SNE 0x0E -# define NV30_FP_OP_OPCODE_SEQ 0x0F -# define NV30_FP_OP_OPCODE_FRC 0x10 -# define NV30_FP_OP_OPCODE_FLR 0x11 -# define NV30_FP_OP_OPCODE_KIL 0x12 -# define NV30_FP_OP_OPCODE_PK4B 0x13 -# define NV30_FP_OP_OPCODE_UP4B 0x14 -# define NV30_FP_OP_OPCODE_DDX 0x15 /* can only write XY */ -# define NV30_FP_OP_OPCODE_DDY 0x16 /* can only write XY */ -# define NV30_FP_OP_OPCODE_TEX 0x17 -# define NV30_FP_OP_OPCODE_TXP 0x18 -# define NV30_FP_OP_OPCODE_TXD 0x19 -# define NV30_FP_OP_OPCODE_RCP 0x1A -# define NV30_FP_OP_OPCODE_RSQ 0x1B -# define NV30_FP_OP_OPCODE_EX2 0x1C -# define NV30_FP_OP_OPCODE_LG2 0x1D -# define NV30_FP_OP_OPCODE_LIT 0x1E -# define NV30_FP_OP_OPCODE_LRP 0x1F -# define NV30_FP_OP_OPCODE_STR 0x20 -# define NV30_FP_OP_OPCODE_SFL 0x21 -# define NV30_FP_OP_OPCODE_COS 0x22 -# define NV30_FP_OP_OPCODE_SIN 0x23 -# define NV30_FP_OP_OPCODE_PK2H 0x24 -# define NV30_FP_OP_OPCODE_UP2H 0x25 -# define NV30_FP_OP_OPCODE_POW 0x26 -# define NV30_FP_OP_OPCODE_PK4UB 0x27 -# define NV30_FP_OP_OPCODE_UP4UB 0x28 -# define NV30_FP_OP_OPCODE_PK2US 0x29 -# define NV30_FP_OP_OPCODE_UP2US 0x2A -# define NV30_FP_OP_OPCODE_DP2A 0x2E -# define NV30_FP_OP_OPCODE_TXB 0x31 -# define NV30_FP_OP_OPCODE_RFL 0x36 -# define NV30_FP_OP_OPCODE_DIV 0x3A -#define NV30_FP_OP_OUT_SAT (1 << 31) - -/* high order bits of SRC0 */ -#define NV30_FP_OP_OUT_ABS (1 << 29) -#define NV30_FP_OP_COND_SWZ_W_SHIFT 27 -#define NV30_FP_OP_COND_SWZ_W_MASK (3 << 27) -#define NV30_FP_OP_COND_SWZ_Z_SHIFT 25 -#define NV30_FP_OP_COND_SWZ_Z_MASK (3 << 25) -#define NV30_FP_OP_COND_SWZ_Y_SHIFT 23 -#define NV30_FP_OP_COND_SWZ_Y_MASK (3 << 23) -#define NV30_FP_OP_COND_SWZ_X_SHIFT 21 -#define NV30_FP_OP_COND_SWZ_X_MASK (3 << 21) -#define NV30_FP_OP_COND_SWZ_ALL_SHIFT 21 -#define NV30_FP_OP_COND_SWZ_ALL_MASK (0xFF << 21) -#define NV30_FP_OP_COND_SHIFT 18 -#define NV30_FP_OP_COND_MASK (0x07 << 18) -# define NV30_FP_OP_COND_FL 0 -# define NV30_FP_OP_COND_LT 1 -# define NV30_FP_OP_COND_EQ 2 -# define NV30_FP_OP_COND_LE 3 -# define NV30_FP_OP_COND_GT 4 -# define NV30_FP_OP_COND_NE 5 -# define NV30_FP_OP_COND_GE 6 -# define NV30_FP_OP_COND_TR 7 - -/* high order bits of SRC1 */ -#define NV30_FP_OP_DST_SCALE_SHIFT 28 -#define NV30_FP_OP_DST_SCALE_MASK (3 << 28) -#define NV30_FP_OP_DST_SCALE_1X 0 -#define NV30_FP_OP_DST_SCALE_2X 1 -#define NV30_FP_OP_DST_SCALE_4X 2 -#define NV30_FP_OP_DST_SCALE_8X 3 -#define NV30_FP_OP_DST_SCALE_INV_2X 5 -#define NV30_FP_OP_DST_SCALE_INV_4X 6 -#define NV30_FP_OP_DST_SCALE_INV_8X 7 - - -/* high order bits of SRC2 */ -#define NV30_FP_OP_INDEX_INPUT (1 << 30) - -//== Register selection == -#define NV30_FP_REG_TYPE_SHIFT 0 -#define NV30_FP_REG_TYPE_MASK (3 << 0) -# define NV30_FP_REG_TYPE_TEMP 0 -# define NV30_FP_REG_TYPE_INPUT 1 -# define NV30_FP_REG_TYPE_CONST 2 -#define NV30_FP_REG_SRC_SHIFT 2 /* uncertain */ -#define NV30_FP_REG_SRC_MASK (31 << 2) -#define NV30_FP_REG_SRC_HALF (1 << 8) -#define NV30_FP_REG_SWZ_ALL_SHIFT 9 -#define NV30_FP_REG_SWZ_ALL_MASK (255 << 9) -#define NV30_FP_REG_SWZ_X_SHIFT 9 -#define NV30_FP_REG_SWZ_X_MASK (3 << 9) -#define NV30_FP_REG_SWZ_Y_SHIFT 11 -#define NV30_FP_REG_SWZ_Y_MASK (3 << 11) -#define NV30_FP_REG_SWZ_Z_SHIFT 13 -#define NV30_FP_REG_SWZ_Z_MASK (3 << 13) -#define NV30_FP_REG_SWZ_W_SHIFT 15 -#define NV30_FP_REG_SWZ_W_MASK (3 << 15) -# define NV30_FP_SWIZZLE_X 0 -# define NV30_FP_SWIZZLE_Y 1 -# define NV30_FP_SWIZZLE_Z 2 -# define NV30_FP_SWIZZLE_W 3 -#define NV30_FP_REG_NEGATE (1 << 17) - -#define NV30SR_NONE 0 -#define NV30SR_OUTPUT 1 -#define NV30SR_INPUT 2 -#define NV30SR_TEMP 3 -#define NV30SR_CONST 4 - -struct nv30_sreg { - int type; - int index; - - int dst_scale; - - int negate; - int abs; - int swz[4]; - - int cc_update; - int cc_update_reg; - int cc_test; - int cc_test_reg; - int cc_swz[4]; -}; - -static INLINE struct nv30_sreg -nv30_sr(int type, int index) -{ - struct nv30_sreg temp = { - .type = type, - .index = index, - .dst_scale = DEF_SCALE, - .abs = 0, - .negate = 0, - .swz = { 0, 1, 2, 3 }, - .cc_update = 0, - .cc_update_reg = 0, - .cc_test = DEF_CTEST, - .cc_test_reg = 0, - .cc_swz = { 0, 1, 2, 3 }, - }; - return temp; -} - -static INLINE struct nv30_sreg -nv30_sr_swz(struct nv30_sreg src, int x, int y, int z, int w) -{ - struct nv30_sreg dst = src; - - dst.swz[SWZ_X] = src.swz[x]; - dst.swz[SWZ_Y] = src.swz[y]; - dst.swz[SWZ_Z] = src.swz[z]; - dst.swz[SWZ_W] = src.swz[w]; - return dst; -} - -static INLINE struct nv30_sreg -nv30_sr_neg(struct nv30_sreg src) -{ - src.negate = !src.negate; - return src; -} - -static INLINE struct nv30_sreg -nv30_sr_abs(struct nv30_sreg src) -{ - src.abs = 1; - return src; -} - -static INLINE struct nv30_sreg -nv30_sr_scale(struct nv30_sreg src, int scale) -{ - src.dst_scale = scale; - return src; -} +#include "nvfx_shader.h" #endif -- cgit v1.2.3