Get rid of req->io_if

Remove one callback indirection level. I/O never changes it's direction
so there is no point in storing both read and write callbacks for each
request.

Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>

src/utils/utils_cleaner.c

@@ -293,11 +293,6 @@ static int _ocf_cleaner_fire_flush_cache(struct ocf_request *req)
 	return 0;
 }
 
-static const struct ocf_io_if _io_if_flush_cache = {
-	.read = _ocf_cleaner_fire_flush_cache,
-	.write = _ocf_cleaner_fire_flush_cache,
-};
-
 static void _ocf_cleaner_metadata_io_end(struct ocf_request *req, int error)
 {
 	if (error) {
@@ -309,7 +304,7 @@ static void _ocf_cleaner_metadata_io_end(struct ocf_request *req, int error)
 
 	OCF_DEBUG_MSG(req->cache, "Metadata flush finished");
 
-	req->io_if = &_io_if_flush_cache;
+	req->engine_handler = _ocf_cleaner_fire_flush_cache;
 	ocf_engine_push_req_front(req, true);
 }
 
@@ -361,11 +356,6 @@ static int _ocf_cleaner_update_metadata(struct ocf_request *req)
 	return 0;
 }
 
-static const struct ocf_io_if _io_if_update_metadata = {
-		.read = _ocf_cleaner_update_metadata,
-		.write = _ocf_cleaner_update_metadata,
-};
-
 static void _ocf_cleaner_flush_cores_io_end(struct ocf_map_info *map,
 		struct ocf_request *req, int error)
 {
@@ -393,7 +383,7 @@ static void _ocf_cleaner_flush_cores_io_end(struct ocf_map_info *map,
 	/*
 	 * All core writes done, switch to post cleaning activities
 	 */
-	req->io_if = &_io_if_update_metadata;
+	req->engine_handler = _ocf_cleaner_update_metadata;
 	ocf_engine_push_req_front(req, true);
 }
 
@@ -454,11 +444,6 @@ static int _ocf_cleaner_fire_flush_cores(struct ocf_request *req)
 	return 0;
 }
 
-static const struct ocf_io_if _io_if_flush_cores = {
-	.read = _ocf_cleaner_fire_flush_cores,
-	.write = _ocf_cleaner_fire_flush_cores,
-};
-
 static void _ocf_cleaner_core_io_end(struct ocf_request *req)
 {
 	if (env_atomic_dec_return(&req->req_remaining))
@@ -470,7 +455,7 @@ static void _ocf_cleaner_core_io_end(struct ocf_request *req)
 	 * All cache read requests done, now we can submit writes to cores,
 	 * Move processing to thread, where IO will be (and can be) submitted
 	 */
-	req->io_if = &_io_if_flush_cores;
+	req->engine_handler = _ocf_cleaner_fire_flush_cores;
 	ocf_engine_push_req_front(req, true);
 }
 
@@ -619,11 +604,6 @@ static int _ocf_cleaner_fire_core(struct ocf_request *req)
 	return 0;
 }
 
-static const struct ocf_io_if _io_if_fire_core = {
-		.read = _ocf_cleaner_fire_core,
-		.write = _ocf_cleaner_fire_core,
-};
-
 static void _ocf_cleaner_cache_io_end(struct ocf_request *req)
 {
 	if (env_atomic_dec_return(&req->req_remaining))
@@ -633,7 +613,7 @@ static void _ocf_cleaner_cache_io_end(struct ocf_request *req)
 	 * All cache read requests done, now we can submit writes to cores,
 	 * Move processing to thread, where IO will be (and can be) submitted
 	 */
-	req->io_if = &_io_if_fire_core;
+	req->engine_handler = _ocf_cleaner_fire_core;
 	ocf_engine_push_req_front(req, true);
 
 	OCF_DEBUG_MSG(req->cache, "Cache reads finished");
@@ -723,16 +703,11 @@ static int _ocf_cleaner_fire_cache(struct ocf_request *req)
 	return 0;
 }
 
-static const struct ocf_io_if _io_if_fire_cache = {
-	.read = _ocf_cleaner_fire_cache,
-	.write = _ocf_cleaner_fire_cache,
-};
-
 static int _ocf_cleaner_fire(struct ocf_request *req)
 {
 	int result;
 
-	req->io_if = &_io_if_fire_cache;
+	req->engine_handler = _ocf_cleaner_fire_cache;
 
 	/* Handle cache lines locks */
 	result = _ocf_cleaner_cache_line_lock(req);

src/utils/utils_parallelize.c

@@ -1,5 +1,5 @@
 /*
- * Copyright(c) 2012-2021 Intel Corporation
+ * Copyright(c) 2012-2022 Intel Corporation
  * SPDX-License-Identifier: BSD-3-Clause
  */
 
@@ -46,11 +46,6 @@ static int _ocf_parallelize_hndl(struct ocf_request *req)
 	return 0;
 }
 
-static const struct ocf_io_if _io_if_parallelize = {
-	.read = _ocf_parallelize_hndl,
-	.write = _ocf_parallelize_hndl,
-};
-
 int ocf_parallelize_create(ocf_parallelize_t *parallelize,
 		ocf_cache_t cache, unsigned shards_cnt, uint32_t priv_size,
 		ocf_parallelize_handle_t handle,
@@ -97,7 +92,8 @@ int ocf_parallelize_create(ocf_parallelize_t *parallelize,
 				goto err_reqs;
 			}
 			tmp_parallelize->reqs[i]->info.internal = true;
-			tmp_parallelize->reqs[i]->io_if = &_io_if_parallelize;
+			tmp_parallelize->reqs[i]->engine_handler =
+					_ocf_parallelize_hndl;
 			tmp_parallelize->reqs[i]->byte_position = i;
 			tmp_parallelize->reqs[i]->priv = tmp_parallelize;
 			i++;

src/utils/utils_pipeline.c

@@ -67,11 +67,6 @@ static int _ocf_pipeline_run_step(struct ocf_request *req)
 	return 0;
 }
 
-static const struct ocf_io_if _io_if_pipeline = {
-	.read = _ocf_pipeline_run_step,
-	.write = _ocf_pipeline_run_step,
-};
-
 int ocf_pipeline_create(ocf_pipeline_t *pipeline, ocf_cache_t cache,
 		struct ocf_pipeline_properties *properties)
 {
@@ -101,7 +96,7 @@ int ocf_pipeline_create(ocf_pipeline_t *pipeline, ocf_cache_t cache,
 	tmp_pipeline->error = 0;
 
 	req->info.internal = true;
-	req->io_if = &_io_if_pipeline;
+	req->engine_handler = _ocf_pipeline_run_step;
 	req->priv = tmp_pipeline;
 
 	*pipeline = tmp_pipeline;