[Mesa-dev] [PATCH] r600: fix bfe where src/dst are same.

[Dave Airlie]

From: Dave Airlie <airl...@redhat.com>

This fixes overlaps where src/dst are the same.

Fixes a bunch of the deqp bitfield tests.

Signed-off-by: Dave Airlie <airl...@redhat.com>
---
 src/gallium/drivers/r600/r600_shader.c | 29 ++++++++++++++++++++++++-----
 1 file changed, 24 insertions(+), 5 deletions(-)

diff --git a/src/gallium/drivers/r600/r600_shader.c 
b/src/gallium/drivers/r600/r600_shader.c
index 9345291..55504a8 100644
--- a/src/gallium/drivers/r600/r600_shader.c
+++ b/src/gallium/drivers/r600/r600_shader.c
@@ -6618,7 +6618,7 @@ static int tgsi_make_src_for_op3(struct r600_shader_ctx 
*ctx,
        return 0;
 }
 
-static int tgsi_op3(struct r600_shader_ctx *ctx)
+static int tgsi_op3_dst(struct r600_shader_ctx *ctx, int dst)
 {
        struct tgsi_full_instruction *inst = 
&ctx->parse.FullToken.FullInstruction;
        struct r600_bytecode_alu alu;
@@ -6648,7 +6648,11 @@ static int tgsi_op3(struct r600_shader_ctx *ctx)
                                return r;
                }
 
-               tgsi_dst(ctx, &inst->Dst[0], i, &alu.dst);
+               if (dst == -1) {
+                       tgsi_dst(ctx, &inst->Dst[0], i, &alu.dst);
+               } else {
+                       alu.dst.sel = dst;
+               }
                alu.dst.chan = i;
                alu.dst.write = 1;
                alu.is_op3 = 1;
@@ -6662,6 +6666,11 @@ static int tgsi_op3(struct r600_shader_ctx *ctx)
        return 0;
 }
 
+static int tgsi_op3(struct r600_shader_ctx *ctx)
+{
+       return tgsi_op3_dst(ctx, -1);
+}
+
 static int tgsi_dp(struct r600_shader_ctx *ctx)
 {
        struct tgsi_full_instruction *inst = 
&ctx->parse.FullToken.FullInstruction;
@@ -9581,8 +9590,15 @@ static int tgsi_bfe(struct r600_shader_ctx *ctx)
        struct r600_bytecode_alu alu;
        int lasti = tgsi_last_instruction(inst->Dst[0].Register.WriteMask);
        int r, i;
+       int dst = -1;
 
-       r = tgsi_op3(ctx);
+       if ((inst->Src[0].Register.File == inst->Dst[0].Register.File &&
+            inst->Src[0].Register.Index == inst->Dst[0].Register.Index) ||
+           (inst->Src[2].Register.File == inst->Dst[0].Register.File &&
+            inst->Src[2].Register.Index == inst->Dst[0].Register.Index))
+               dst = r600_get_temp(ctx);
+
+       r = tgsi_op3_dst(ctx, dst);
        if (r)
                return r;
 
@@ -9607,10 +9623,13 @@ static int tgsi_bfe(struct r600_shader_ctx *ctx)
                alu.op = ALU_OP3_CNDE_INT;
                alu.is_op3 = 1;
                alu.src[0].sel = ctx->temp_reg;
-               alu.src[1].chan = i;
+               alu.src[0].chan = i;
 
                tgsi_dst(ctx, &inst->Dst[0], i, &alu.dst);
-               alu.src[1].sel = alu.dst.sel;
+               if (dst != -1)
+                       alu.src[1].sel = dst;
+               else
+                       alu.src[1].sel = alu.dst.sel;
                alu.src[1].chan = i;
                r600_bytecode_src(&alu.src[2], &ctx->src[0], i);
                alu.dst.write = 1;
-- 
2.9.5

_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev