diff --git a/dlls/d3dx9_36/preshader.c b/dlls/d3dx9_36/preshader.c
index 3b06f66972c..25374a3d97c 100644
--- a/dlls/d3dx9_36/preshader.c
+++ b/dlls/d3dx9_36/preshader.c
@@ -269,6 +269,7 @@ struct const_upload_info
     unsigned int major_stride;
     unsigned int major_count;
     unsigned int count;
+    unsigned int minor_remainder;
 };
 
 static enum pres_value_type table_type_from_param_type(D3DXPARAMETER_TYPE type)
@@ -1038,10 +1039,22 @@ static void get_const_upload_info(struct d3dx_const_param_eval_output *const_set
         info->major = param->rows;
         info->minor = param->columns;
     }
-    info->major_stride = max(info->minor, get_reg_components(table));
-    info->major_count = min(info->major * info->major_stride,
-            get_offset_reg(table, const_set->register_count) + info->major_stride - 1) / info->major_stride;
-    info->count = info->major_count * info->minor;
+
+    if (get_reg_components(table) == 1)
+    {
+        unsigned int const_length = get_offset_reg(table, const_set->register_count);
+
+        info->major_stride = info->minor;
+        info->major_count = const_length / info->major_stride;
+        info->minor_remainder = const_length % info->major_stride;
+    }
+    else
+    {
+        info->major_stride = get_reg_components(table);
+        info->major_count = const_set->register_count;
+        info->minor_remainder = 0;
+    }
+    info->count = info->major_count * info->minor + info->minor_remainder;
 }
 
 static void pres_int_from_float(void *out, const void *in, unsigned int count)
@@ -1172,16 +1185,19 @@ static void set_constants(struct d3dx_regstore *rs, struct d3dx_const_tab *const
             {
                 for (j = 0; j < info.minor; ++j)
                 {
-                    unsigned int offset;
-
-                    offset = i * info.major_stride + j;
-                    if (get_reg_offset(table, offset) >= const_set->register_count)
-                        break;
                     if (info.transpose)
                         param_offset = i + j * info.major;
                     else
                         param_offset = i * info.minor + j;
-                    out[offset] = data[param_offset];
+                    out[i * info.major_stride + j] = data[param_offset];
+                }
+            }
+            if (info.transpose)
+            {
+                for (j = 0; j < info.minor_remainder; ++j)
+                {
+                    param_offset = i + j * info.major;
+                    out[i * info.major_stride + j] = data[param_offset];
                 }
             }
             start_offset += get_offset_reg(table, const_set->register_count);
@@ -1417,6 +1433,16 @@ static HRESULT init_set_constants_param(struct d3dx_const_tab *const_tab, ID3DXC
             && !info.transpose && info.minor == info.major_stride
             && info.count == get_offset_reg(const_set.table, const_set.register_count)
             && info.count * sizeof(unsigned int) <= param->bytes;
+    if (info.minor_remainder && !const_set.direct_copy && !info.transpose)
+        FIXME("Incomplete last row for not transposed matrix which cannot be directly copied, parameter %s.\n",
+                debugstr_a(param->name));
+
+    if (info.major_count > info.major
+            || (info.major_count == info.major && info.minor_remainder))
+    {
+        WARN("Constant dimensions exceed parameter size.\n");
+        return D3DERR_INVALIDCALL;
+    }
 
     if (FAILED(hr = append_const_set(const_tab, &const_set)))
         return hr;