feat(esp_driver_dma): add blocking async memcpy API

Add a blocking wrapper for async memcpy so simple users can wait for one DMA copy without writing their own ISR callback and semaphore plumbing. Update functional tests and documentation to use the simpler API where async completion handling is not needed.

Co-authored-by: Cursor <cursoragent@cursor.com>

docs/en/api-reference/peripherals/async_memcpy.rst

@@ -67,6 +67,14 @@ The prototype of the callback function is :cpp:type:`async_memcpy_isr_cb_t`. The
     // Do something else here
     xSemaphoreTake(my_semaphore, portMAX_DELAY); // Wait until the buffer copy is done
 
+For simpler use cases where the task only needs to wait until one copy finishes, use :cpp:func:`esp_memcpy_blocking`. This API is built on top of the async request path and waits internally for the completion callback.
+
+.. code-block:: c
+
+    ESP_ERROR_CHECK(esp_memcpy_blocking(driver_handle, to, from, copy_len, -1));
+
+The blocking API must not be called from ISR context. Currently, it only supports ``timeout_ms = -1``, which means waiting indefinitely until the memory copy completes.
+
 
 Uninstall Driver
 ----------------