From: Russell King <r...@arm.linux.org.uk> Add support for lowering FLR and CEIL to FRC/SUB and FRC/ADD instructions for GPUs that support FRC but not FLR or CEIL. Since these uses FRC, it is invalid to ask for FLR or CEIL to be lowered along with FRC, so add an assert to catch this invalid configuration.
We also need to deal with FLR instructions emitted by the lowering code. Fix these up with the FRC+SUB equivalent when FLR lowering is enabled. Signed-off-by: Russell King <r...@arm.linux.org.uk> Reviewed-by: Rob Clark <robdcl...@gmail.com> --- src/gallium/auxiliary/tgsi/tgsi_lowering.c | 167 +++++++++++++++++++++++++---- src/gallium/auxiliary/tgsi/tgsi_lowering.h | 2 + 2 files changed, 149 insertions(+), 20 deletions(-) diff --git a/src/gallium/auxiliary/tgsi/tgsi_lowering.c b/src/gallium/auxiliary/tgsi/tgsi_lowering.c index 0ffd855..b2dd37e 100644 --- a/src/gallium/auxiliary/tgsi/tgsi_lowering.c +++ b/src/gallium/auxiliary/tgsi/tgsi_lowering.c @@ -676,14 +676,19 @@ transform_lit(struct tgsi_transform_context *tctx, * dst.w = 1.0 * * ; needs: 1 tmp, imm{1.0} - * FLR tmpA.x, src.x + * if (lowering FLR) { + * FRC tmpA.x, src.x + * SUB tmpA.x, src.x, tmpA.x + * } else { + * FLR tmpA.x, src.x + * } * EX2 tmpA.y, src.x * SUB dst.y, src.x, tmpA.x * EX2 dst.x, tmpA.x * MOV dst.z, tmpA.y * MOV dst.w, imm{1.0} */ -#define EXP_GROW (NINST(1) + NINST(1) + NINST(2) + NINST(1) + \ +#define EXP_GROW (NINST(1) + NINST(2) + NINST(1) + NINST(2) + NINST(1) + \ NINST(1)+ NINST(1) - OINST(1)) #define EXP_TMP 1 static void @@ -696,14 +701,35 @@ transform_exp(struct tgsi_transform_context *tctx, struct tgsi_full_instruction new_inst; if (dst->Register.WriteMask & TGSI_WRITEMASK_XY) { - /* FLR tmpA.x, src.x */ - new_inst = tgsi_default_full_instruction(); - new_inst.Instruction.Opcode = TGSI_OPCODE_FLR; - new_inst.Instruction.NumDstRegs = 1; - reg_dst(&new_inst.Dst[0], &ctx->tmp[A].dst, TGSI_WRITEMASK_X); - new_inst.Instruction.NumSrcRegs = 1; - reg_src(&new_inst.Src[0], src, SWIZ(X, _, _, _)); - tctx->emit_instruction(tctx, &new_inst); + if (ctx->config->lower_FLR) { + /* FRC tmpA.x, src.x */ + new_inst = tgsi_default_full_instruction(); + new_inst.Instruction.Opcode = TGSI_OPCODE_FRC; + new_inst.Instruction.NumDstRegs = 1; + reg_dst(&new_inst.Dst[0], &ctx->tmp[A].dst, TGSI_WRITEMASK_X); + new_inst.Instruction.NumSrcRegs = 1; + reg_src(&new_inst.Src[0], src, SWIZ(X, _, _, _)); + tctx->emit_instruction(tctx, &new_inst); + + /* SUB tmpA.x, src.x, tmpA.x */ + new_inst = tgsi_default_full_instruction(); + new_inst.Instruction.Opcode = TGSI_OPCODE_SUB; + new_inst.Instruction.NumDstRegs = 1; + reg_dst(&new_inst.Dst[0], &ctx->tmp[A].dst, TGSI_WRITEMASK_X); + new_inst.Instruction.NumSrcRegs = 2; + reg_src(&new_inst.Src[0], src, SWIZ(X, _, _, _)); + reg_src(&new_inst.Src[1], &ctx->tmp[A].src, SWIZ(X, _, _, _)); + tctx->emit_instruction(tctx, &new_inst); + } else { + /* FLR tmpA.x, src.x */ + new_inst = tgsi_default_full_instruction(); + new_inst.Instruction.Opcode = TGSI_OPCODE_FLR; + new_inst.Instruction.NumDstRegs = 1; + reg_dst(&new_inst.Dst[0], &ctx->tmp[A].dst, TGSI_WRITEMASK_X); + new_inst.Instruction.NumSrcRegs = 1; + reg_src(&new_inst.Src[0], src, SWIZ(X, _, _, _)); + tctx->emit_instruction(tctx, &new_inst); + } } if (dst->Register.WriteMask & TGSI_WRITEMASK_Z) { @@ -771,14 +797,19 @@ transform_exp(struct tgsi_transform_context *tctx, * * ; needs: 1 tmp, imm{1.0} * LG2 tmpA.x, |src.x| - * FLR tmpA.y, tmpA.x + * if (lowering FLR) { + * FRC tmpA.y, tmpA.x + * SUB tmpA.y, tmpA.x, tmpA.y + * } else { + * FLR tmpA.y, tmpA.x + * } * EX2 tmpA.z, tmpA.y * RCP tmpA.z, tmpA.z * MUL dst.y, |src.x|, tmpA.z * MOV dst.xz, tmpA.yx * MOV dst.w, imm{1.0} */ -#define LOG_GROW (NINST(1) + NINST(1) + NINST(1) + NINST(1) + \ +#define LOG_GROW (NINST(1) + NINST(1) + NINST(2) + NINST(1) + NINST(1) + \ NINST(2) + NINST(1) + NINST(1) - OINST(1)) #define LOG_TMP 1 static void @@ -803,14 +834,35 @@ transform_log(struct tgsi_transform_context *tctx, } if (dst->Register.WriteMask & TGSI_WRITEMASK_XY) { - /* FLR tmpA.y, tmpA.x */ - new_inst = tgsi_default_full_instruction(); - new_inst.Instruction.Opcode = TGSI_OPCODE_FLR; - new_inst.Instruction.NumDstRegs = 1; - reg_dst(&new_inst.Dst[0], &ctx->tmp[A].dst, TGSI_WRITEMASK_Y); - new_inst.Instruction.NumSrcRegs = 1; - reg_src(&new_inst.Src[0], &ctx->tmp[A].src, SWIZ(_, X, _, _)); - tctx->emit_instruction(tctx, &new_inst); + if (ctx->config->lower_FLR) { + /* FRC tmpA.y, tmpA.x */ + new_inst = tgsi_default_full_instruction(); + new_inst.Instruction.Opcode = TGSI_OPCODE_FRC; + new_inst.Instruction.NumDstRegs = 1; + reg_dst(&new_inst.Dst[0], &ctx->tmp[A].dst, TGSI_WRITEMASK_Y); + new_inst.Instruction.NumSrcRegs = 1; + reg_src(&new_inst.Src[0], &ctx->tmp[A].src, SWIZ(_, X, _, _)); + tctx->emit_instruction(tctx, &new_inst); + + /* SUB tmpA.y, tmpA.x, tmpA.y */ + new_inst = tgsi_default_full_instruction(); + new_inst.Instruction.Opcode = TGSI_OPCODE_SUB; + new_inst.Instruction.NumDstRegs = 1; + reg_dst(&new_inst.Dst[0], &ctx->tmp[A].dst, TGSI_WRITEMASK_Y); + new_inst.Instruction.NumSrcRegs = 2; + reg_src(&new_inst.Src[0], &ctx->tmp[A].src, SWIZ(_, X, _, _)); + reg_src(&new_inst.Src[1], &ctx->tmp[A].src, SWIZ(_, Y, _, _)); + tctx->emit_instruction(tctx, &new_inst); + } else { + /* FLR tmpA.y, tmpA.x */ + new_inst = tgsi_default_full_instruction(); + new_inst.Instruction.Opcode = TGSI_OPCODE_FLR; + new_inst.Instruction.NumDstRegs = 1; + reg_dst(&new_inst.Dst[0], &ctx->tmp[A].dst, TGSI_WRITEMASK_Y); + new_inst.Instruction.NumSrcRegs = 1; + reg_src(&new_inst.Src[0], &ctx->tmp[A].src, SWIZ(_, X, _, _)); + tctx->emit_instruction(tctx, &new_inst); + } } if (dst->Register.WriteMask & TGSI_WRITEMASK_Y) { @@ -1005,6 +1057,58 @@ transform_dotp(struct tgsi_transform_context *tctx, } } +/* FLR - floor, CEIL - ceil + * ; needs: 1 tmp + * if (CEIL) { + * FRC tmpA, -src + * ADD dst, src, tmpA + * } else { + * FRC tmpA, src + * SUB dst, src, tmpA + * } + */ +#define FLR_GROW (NINST(1) + NINST(2) - OINST(1)) +#define CEIL_GROW (NINST(1) + NINST(2) - OINST(1)) +#define FLR_TMP 1 +#define CEIL_TMP 1 +static void +transform_flr_ceil(struct tgsi_transform_context *tctx, + struct tgsi_full_instruction *inst) +{ + struct tgsi_lowering_context *ctx = tgsi_lowering_context(tctx); + struct tgsi_full_dst_register *dst = &inst->Dst[0]; + struct tgsi_full_src_register *src0 = &inst->Src[0]; + struct tgsi_full_instruction new_inst; + unsigned opcode = inst->Instruction.Opcode; + + if (dst->Register.WriteMask & TGSI_WRITEMASK_XYZW) { + /* FLR: FRC tmpA, src CEIL: FRC tmpA, -src */ + new_inst = tgsi_default_full_instruction(); + new_inst.Instruction.Opcode = TGSI_OPCODE_FRC; + new_inst.Instruction.NumDstRegs = 1; + reg_dst(&new_inst.Dst[0], &ctx->tmp[A].dst, TGSI_WRITEMASK_XYZW); + new_inst.Instruction.NumSrcRegs = 1; + reg_src(&new_inst.Src[0], src0, SWIZ(X, Y, Z, W)); + + if (opcode == TGSI_OPCODE_CEIL) + new_inst.Src[0].Register.Negate = !new_inst.Src[0].Register.Negate; + tctx->emit_instruction(tctx, &new_inst); + + /* FLR: SUB dst, src, tmpA CEIL: ADD dst, src, tmpA */ + new_inst = tgsi_default_full_instruction(); + if (opcode == TGSI_OPCODE_CEIL) + new_inst.Instruction.Opcode = TGSI_OPCODE_ADD; + else + new_inst.Instruction.Opcode = TGSI_OPCODE_SUB; + new_inst.Instruction.NumDstRegs = 1; + reg_dst(&new_inst.Dst[0], dst, TGSI_WRITEMASK_XYZW); + new_inst.Instruction.NumSrcRegs = 2; + reg_src(&new_inst.Src[0], src0, SWIZ(X, Y, Z, W)); + reg_src(&new_inst.Src[1], &ctx->tmp[A].src, SWIZ(X, Y, Z, W)); + tctx->emit_instruction(tctx, &new_inst); + } +} + /* Inserts a MOV_SAT for the needed components of tex coord. Note that * in the case of TXP, the clamping must happen *after* projection, so * we need to lower TXP to TEX. @@ -1401,6 +1505,16 @@ transform_instr(struct tgsi_transform_context *tctx, goto skip; transform_dotp(tctx, inst); break; + case TGSI_OPCODE_FLR: + if (!ctx->config->lower_FLR) + goto skip; + transform_flr_ceil(tctx, inst); + break; + case TGSI_OPCODE_CEIL: + if (!ctx->config->lower_CEIL) + goto skip; + transform_flr_ceil(tctx, inst); + break; case TGSI_OPCODE_TEX: case TGSI_OPCODE_TXP: case TGSI_OPCODE_TXB: @@ -1432,6 +1546,9 @@ tgsi_transform_lowering(const struct tgsi_lowering_config *config, /* sanity check in case limit is ever increased: */ STATIC_ASSERT((sizeof(config->saturate_s) * 8) >= PIPE_MAX_SAMPLERS); + /* sanity check the lowering */ + assert(!(config->lower_FRC && (config->lower_FLR || config->lower_CEIL))); + memset(&ctx, 0, sizeof(ctx)); ctx.base.transform_instruction = transform_instr; ctx.info = info; @@ -1473,6 +1590,8 @@ tgsi_transform_lowering(const struct tgsi_lowering_config *config, OPCS(DPH) || OPCS(DP2) || OPCS(DP2A) || + OPCS(FLR) || + OPCS(CEIL) || OPCS(TXP) || ctx.two_side_colors || ctx.saturate)) @@ -1541,6 +1660,14 @@ tgsi_transform_lowering(const struct tgsi_lowering_config *config, newlen += DP2A_GROW * OPCS(DP2A); numtmp = MAX2(numtmp, DOTP_TMP); } + if (OPCS(FLR)) { + newlen += FLR_GROW * OPCS(FLR); + numtmp = MAX2(numtmp, FLR_TMP); + } + if (OPCS(CEIL)) { + newlen += CEIL_GROW * OPCS(CEIL); + numtmp = MAX2(numtmp, CEIL_TMP); + } if (ctx.saturate || config->lower_TXP) { int n = 0; diff --git a/src/gallium/auxiliary/tgsi/tgsi_lowering.h b/src/gallium/auxiliary/tgsi/tgsi_lowering.h index 52c204f..a96d85d 100644 --- a/src/gallium/auxiliary/tgsi/tgsi_lowering.h +++ b/src/gallium/auxiliary/tgsi/tgsi_lowering.h @@ -68,6 +68,8 @@ struct tgsi_lowering_config unsigned lower_DPH:1; unsigned lower_DP2:1; unsigned lower_DP2A:1; + unsigned lower_FLR:1; + unsigned lower_CEIL:1; /* bitmask of (1 << TGSI_TEXTURE_type): */ unsigned lower_TXP; -- 2.5.5 _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev