nv40: implement user clip planes

It turns out the user planes handed to the driver are already in clip space.
Hence, we no longer need to transform incoming vertices before computing the
clip distance, and no longer need to change the interface provided by
gallium.  Yay :)

The clip state change handling could be better, but this works.

1 files changed, 81 insertions, 78 deletions

diff --git a/src/gallium/drivers/nv40/nv40_vertprog.c b/src/gallium/drivers/nv40/nv40_vertprog.c
index 08d3f387e09..e10250528e2 100644
--- a/src/gallium/drivers/nv40/nv40_vertprog.c
+++ b/src/gallium/drivers/nv40/nv40_vertprog.c
@@ -52,6 +52,8 @@ struct nv40_vpc {
 
 	struct nv40_sreg *imm;
 	unsigned nr_imm;
+
+	unsigned hpos_idx;
 };
 
 static struct nv40_sreg
@@ -423,11 +425,6 @@ nv40_vertprog_parse_instruction(struct nv40_vpc *vpc,
 	int ai = -1, ci = -1, ii = -1;
 	int i;
 
-	struct {
-		struct nv40_sreg dst;
-		unsigned m;
-	} clip;
-
 	if (finst->Instruction.Opcode == TGSI_OPCODE_END)
 		return TRUE;
 
@@ -501,47 +498,6 @@ nv40_vertprog_parse_instruction(struct nv40_vpc *vpc,
 	dst  = tgsi_dst(vpc, &finst->FullDstRegisters[0]);
 	mask = tgsi_mask(finst->FullDstRegisters[0].DstRegister.WriteMask);
 
-	/* If writing to clip distance regs, need to modify instruction to
-	 * change which component is written to.  On NV40 the clip regs
-	 * are the unused components (yzw) of FOGC/PSZ.
-	 */
-	clip.dst = none;
-	if (dst.type == NV40SR_OUTPUT &&
-	    dst.index >= NV40_VP_INST_DEST_CLIP(0) &&
-	    dst.index <= NV40_VP_INST_DEST_CLIP(5)) {
-		unsigned n = dst.index - NV40_VP_INST_DEST_CLIP(0);
-		unsigned m[] =
-			{ MASK_Y, MASK_Z, MASK_W, MASK_Y, MASK_Z, MASK_W };
-
-		/* Some instructions we can get away with swizzling and/or
-		 * changing the writemask.  Others, we'll use a temp reg.
-		 */
-		switch (finst->Instruction.Opcode) {
-		case TGSI_OPCODE_DST:
-		case TGSI_OPCODE_EXP:
-		case TGSI_OPCODE_LIT:
-		case TGSI_OPCODE_LOG:
-		case TGSI_OPCODE_XPD:
-			clip.dst = dst;
-			clip.m = m[n];
-			dst = temp(vpc);
-			break;
-		case TGSI_OPCODE_DP3:
-		case TGSI_OPCODE_DP4:
-		case TGSI_OPCODE_DPH:
-		case TGSI_OPCODE_POW:
-		case TGSI_OPCODE_RCP:
-		case TGSI_OPCODE_RSQ:
-			mask = m[n];
-			break;
-		default:
-			for (i = 0; i < finst->Instruction.NumSrcRegs; i++)
-				src[i] = swz(src[i], X, X, X, X);
-			mask = m[n];
-			break;
-		}
-	}
-
 	switch (finst->Instruction.Opcode) {
 	case TGSI_OPCODE_ABS:
 		arith(vpc, 0, OP_MOV, dst, mask, abs(src[0]), none, none);
@@ -639,11 +595,6 @@ nv40_vertprog_parse_instruction(struct nv40_vpc *vpc,
 		return FALSE;
 	}
 
-	if (clip.dst.type != NV40SR_NONE) {
-		arith(vpc, 0, OP_MOV, clip.dst, clip.m,
-		      swz(dst, X, X, X, X), none, none);
-	}
-
 	release_temps(vpc);
 	return TRUE;
 }
@@ -658,6 +609,7 @@ nv40_vertprog_parse_decl_output(struct nv40_vpc *vpc,
 	switch (fdec->Semantic.SemanticName) {
 	case TGSI_SEMANTIC_POSITION:
 		hw = NV40_VP_INST_DEST_POS;
+		vpc->hpos_idx = idx;
 		break;
 	case TGSI_SEMANTIC_COLOR:
 		if (fdec->Semantic.SemanticIndex == 0) {
@@ -695,15 +647,6 @@ nv40_vertprog_parse_decl_output(struct nv40_vpc *vpc,
 			return FALSE;
 		}
 		break;
-#if 0
-	case TGSI_SEMANTIC_CLIP:
-		if (fdec->Semantic.SemanticIndex >= 6) {
-			NOUVEAU_ERR("bad clip distance index\n");
-			return FALSE;
-		}
-		hw = NV40_VP_INST_DEST_CLIP(fdec->Semantic.SemanticIndex);
-		break;
-#endif
 	default:
 		NOUVEAU_ERR("bad output semantic\n");
 		return FALSE;
@@ -748,6 +691,10 @@ nv40_vertprog_prepare(struct nv40_vpc *vpc)
 				}
 				break;
 #endif
+			case TGSI_FILE_OUTPUT:
+				if (!nv40_vertprog_parse_decl_output(vpc, fdec))
+					return FALSE;
+				break;
 			default:
 				break;
 			}
@@ -803,6 +750,8 @@ nv40_vertprog_translate(struct nv40_context *nv40,
 {
 	struct tgsi_parse_context parse;
 	struct nv40_vpc *vpc = NULL;
+	struct nv40_sreg none = nv40_sr(NV40SR_NONE, 0);
+	int i;
 
 	vpc = CALLOC(1, sizeof(struct nv40_vpc));
 	if (!vpc)
@@ -814,26 +763,21 @@ nv40_vertprog_translate(struct nv40_context *nv40,
 		return;
 	}
 
+	/* Redirect post-transform vertex position to a temp if user clip
+	 * planes are enabled.  We need to append code the the vtxprog
+	 * to handle clip planes later.
+	 */
+	if (vp->ucp.nr)  {
+		vpc->r_result[vpc->hpos_idx] = temp(vpc);
+		vpc->r_temps_discard = 0;
+	}
+
 	tgsi_parse_init(&parse, vp->pipe.tokens);
 
 	while (!tgsi_parse_end_of_tokens(&parse)) {
 		tgsi_parse_token(&parse);
 
 		switch (parse.FullToken.Token.Type) {
-		case TGSI_TOKEN_TYPE_DECLARATION:
-		{
-			const struct tgsi_full_declaration *fdec;
-			fdec = &parse.FullToken.FullDeclaration;
-			switch (fdec->Declaration.File) {
-			case TGSI_FILE_OUTPUT:
-				if (!nv40_vertprog_parse_decl_output(vpc, fdec))
-					goto out_err;
-				break;
-			default:
-				break;
-			}
-		}
-			break;
 		case TGSI_TOKEN_TYPE_IMMEDIATE:
 		{
 			const struct tgsi_full_immediate *imm;
@@ -862,6 +806,39 @@ nv40_vertprog_translate(struct nv40_context *nv40,
 		}
 	}
 
+	/* Write out HPOS if it was redirected to a temp earlier */
+	if (vpc->r_result[vpc->hpos_idx].type != NV40SR_OUTPUT) {
+		struct nv40_sreg hpos = nv40_sr(NV40SR_OUTPUT,
+						NV40_VP_INST_DEST_POS);
+		struct nv40_sreg htmp = vpc->r_result[vpc->hpos_idx];
+
+		arith(vpc, 0, OP_MOV, hpos, MASK_ALL, htmp, none, none);
+	}
+
+	/* Insert code to handle user clip planes */
+	for (i = 0; i < vp->ucp.nr; i++) {
+		struct nv40_sreg cdst = nv40_sr(NV40SR_OUTPUT,
+						NV40_VP_INST_DEST_CLIP(i));
+		struct nv40_sreg ceqn = constant(vpc, -1,
+						 nv40->clip.ucp[i][0],
+						 nv40->clip.ucp[i][1],
+						 nv40->clip.ucp[i][2],
+						 nv40->clip.ucp[i][3]);
+		struct nv40_sreg htmp = vpc->r_result[vpc->hpos_idx];
+		unsigned mask;
+
+		switch (i) {
+		case 0: case 3: mask = MASK_Y; break;
+		case 1: case 4: mask = MASK_Z; break;
+		case 2: case 5: mask = MASK_W; break;
+		default:
+			NOUVEAU_ERR("invalid clip dist #%d\n", i);
+			goto out_err;
+		}
+
+		arith(vpc, 0, OP_DP4, cdst, mask, htmp, ceqn, none);
+	}
+
 	vp->insns[vp->nr_insns - 1].data[3] |= NV40_VP_INST_LAST;
 	vp->translated = TRUE;
 out_err:
@@ -883,6 +860,12 @@ nv40_vertprog_validate(struct nv40_context *nv40)
 	if (nv40->render_mode == HW) {
 		vp = nv40->vertprog;
 		constbuf = nv40->constbuf[PIPE_SHADER_VERTEX];
+
+		if ((nv40->dirty & NV40_NEW_UCP) ||
+		    memcmp(&nv40->clip, &vp->ucp, sizeof(vp->ucp))) {
+			nv40_vertprog_destroy(nv40, vp);
+			memcpy(&vp->ucp, &nv40->clip, sizeof(vp->ucp));
+		}
 	} else {
 		vp = nv40->swtnl.vertprog;
 		constbuf = NULL;
@@ -1045,16 +1028,36 @@ check_gpu_resources:
 void
 nv40_vertprog_destroy(struct nv40_context *nv40, struct nv40_vertex_program *vp)
 {
-	if (vp->nr_consts)
-		FREE(vp->consts);
-	if (vp->nr_insns)
+	struct nouveau_winsys *nvws = nv40->screen->nvws;
+
+	vp->translated = FALSE;
+
+	if (vp->nr_insns) {
 		FREE(vp->insns);
+		vp->insns = NULL;
+		vp->nr_insns = 0;
+	}
+
+	if (vp->nr_consts) {
+		FREE(vp->consts);
+		vp->consts = NULL;
+		vp->nr_consts = 0;
+	}
+
+	nvws->res_free(&vp->exec);
+	vp->exec_start = 0;
+	nvws->res_free(&vp->data);
+	vp->data_start = 0;
+	vp->data_start_min = 0;
+
+	vp->ir = vp->or = vp->clip_ctrl = 0;
+	so_ref(NULL, &vp->so);
 }
 
 struct nv40_state_entry nv40_state_vertprog = {
 	.validate = nv40_vertprog_validate,
 	.dirty = {
-		.pipe = NV40_NEW_VERTPROG,
+		.pipe = NV40_NEW_VERTPROG | NV40_NEW_UCP,
 		.hw = NV40_STATE_VERTPROG,
 	}
 };