Add driver async API

Add an asynchronous API to the driver. The current synchronous API
is now using the new async API internally. The main new functions
are ethosu_invoke_async() and ethosu_wait(). Every successfull call
to ethosu_invoke_async() must be followed by a call to ethosu_wait()
to get the status of the inference. The wait function can be called
in a blocking or non-blocking mode by specifying the `block`
argument to true/false. The regular synchronous invoke function is
implemented as a invoke_async followed by a wait(block=true) call.

Short sommary of changes:
 - Add an internal ethosu_job struct to keep track of
   inference data and job state.
 - Use async API in blocking mode for normal flow
 - Change default semaphore implementation to binary type
 - Move error prints out of interrupt context
 - Move ethosu_inference_begin() callback to right before HW invoke
 - Always call ethosu_inference_end() callback, even in case of errors
 - On NPU error, do not keep NPU powered after being reset

Change-Id: If4c3c46e3c6732a669e17251bd848dea5765a490

1 files changed, 11 insertions, 7 deletions

diff --git a/src/ethosu_device.h b/src/ethosu_device.h
index cc9ea8b..0b45fd6 100644
--- a/src/ethosu_device.h
+++ b/src/ethosu_device.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2021 Arm Limited. All rights reserved.
+ * Copyright (c) 2019-2022 Arm Limited. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -81,13 +81,17 @@ enum ethosu_error_codes ethosu_dev_axi_init(struct ethosu_device *dev);
  *                             - All input tensors
  *                             - All output tensors
  * \param[in] num_base_addr    Number of base addresses.
- * \return                     \ref ethosu_error_codes
  */
-enum ethosu_error_codes ethosu_dev_run_command_stream(struct ethosu_device *dev,
-                                                      const uint8_t *cmd_stream_ptr,
-                                                      uint32_t cms_length,
-                                                      const uint64_t *base_addr,
-                                                      int num_base_addr);
+void ethosu_dev_run_command_stream(struct ethosu_device *dev,
+                                   const uint8_t *cmd_stream_ptr,
+                                   uint32_t cms_length,
+                                   const uint64_t *base_addr,
+                                   int num_base_addr);
+
+/**
+ * Print information on NPU error status
+ */
+void ethosu_dev_print_err_status(struct ethosu_device *dev);
 
 /**
  *  Interrupt handler on device layer