diff --git a/libs/vkd3d-shader/hlsl_codegen.c b/libs/vkd3d-shader/hlsl_codegen.c
index 16b8df93..4bca693c 100644
--- a/libs/vkd3d-shader/hlsl_codegen.c
+++ b/libs/vkd3d-shader/hlsl_codegen.c
@@ -8250,6 +8250,22 @@ static bool sm4_generate_vsir_instr_expr(struct hlsl_ctx *ctx,
             generate_vsir_instr_expr_single_instr_op(ctx, program, expr, VKD3DSIH_XOR, 0, 0, true);
             return true;
 
+        case HLSL_OP2_DIV:
+            switch (dst_type->e.numeric.type)
+            {
+                case HLSL_TYPE_FLOAT:
+                    generate_vsir_instr_expr_single_instr_op(ctx, program, expr, VKD3DSIH_DIV, 0, 0, true);
+                    return true;
+
+                case HLSL_TYPE_UINT:
+                    sm4_generate_vsir_expr_with_two_destinations(ctx, program, VKD3DSIH_UDIV, expr, 0);
+                    return true;
+
+                default:
+                    hlsl_fixme(ctx, &expr->node.loc, "SM4 %s division expression.", dst_type_name);
+                    return false;
+            }
+
         case HLSL_OP2_EQUAL:
             VKD3D_ASSERT(dst_type->e.numeric.type == HLSL_TYPE_BOOL);
 
diff --git a/libs/vkd3d-shader/tpf.c b/libs/vkd3d-shader/tpf.c
index 31e196dc..b61db5d7 100644
--- a/libs/vkd3d-shader/tpf.c
+++ b/libs/vkd3d-shader/tpf.c
@@ -5032,24 +5032,6 @@ static void write_sm4_ret(const struct tpf_compiler *tpf)
     write_sm4_instruction(tpf, &instr);
 }
 
-static void write_sm4_binary_op(const struct tpf_compiler *tpf, enum vkd3d_sm4_opcode opcode,
-        const struct hlsl_ir_node *dst, const struct hlsl_ir_node *src1, const struct hlsl_ir_node *src2)
-{
-    struct sm4_instruction instr;
-
-    memset(&instr, 0, sizeof(instr));
-    instr.opcode = opcode;
-
-    sm4_dst_from_node(&instr.dsts[0], dst);
-    instr.dst_count = 1;
-
-    sm4_src_from_node(tpf, &instr.srcs[0], src1, instr.dsts[0].write_mask);
-    sm4_src_from_node(tpf, &instr.srcs[1], src2, instr.dsts[0].write_mask);
-    instr.src_count = 2;
-
-    write_sm4_instruction(tpf, &instr);
-}
-
 /* dp# instructions don't map the swizzle. */
 static void write_sm4_binary_op_dot(const struct tpf_compiler *tpf, enum vkd3d_sm4_opcode opcode,
         const struct hlsl_ir_node *dst, const struct hlsl_ir_node *src1, const struct hlsl_ir_node *src2)
@@ -5320,22 +5302,6 @@ static void write_sm4_expr(const struct tpf_compiler *tpf, const struct hlsl_ir_
 
     switch (expr->op)
     {
-        case HLSL_OP2_DIV:
-            switch (dst_type->e.numeric.type)
-            {
-                case HLSL_TYPE_FLOAT:
-                    write_sm4_binary_op(tpf, VKD3D_SM4_OP_DIV, &expr->node, arg1, arg2);
-                    break;
-
-                case HLSL_TYPE_UINT:
-                    write_sm4_binary_op_with_two_destinations(tpf, VKD3D_SM4_OP_UDIV, &expr->node, 0, arg1, arg2);
-                    break;
-
-                default:
-                    hlsl_fixme(tpf->ctx, &expr->node.loc, "SM4 %s division expression.", dst_type_string->buffer);
-            }
-            break;
-
         case HLSL_OP2_DOT:
             switch (dst_type->e.numeric.type)
             {
@@ -5890,6 +5856,7 @@ static void tpf_handle_instruction(struct tpf_compiler *tpf, const struct vkd3d_
         case VKD3DSIH_SAMPLE_INFO:
         case VKD3DSIH_SINCOS:
         case VKD3DSIH_SQRT:
+        case VKD3DSIH_UDIV:
         case VKD3DSIH_UGE:
         case VKD3DSIH_ULT:
         case VKD3DSIH_UMAX: