Merge pull request #863 from Deixx/detach-zero-superblock

Zero metadata superblock on detach

src/metadata/metadata_structs.h

@@ -1,5 +1,6 @@
 /*
  * Copyright(c) 2012-2022 Intel Corporation
+ * Copyright(c) 2025 Huawei Technologies
  * SPDX-License-Identifier: BSD-3-Clause
  */
 
@@ -21,9 +22,8 @@
  * @brief Metadata shutdown status
  */
 enum ocf_metadata_shutdown_status {
-	ocf_metadata_clean_shutdown = 1, /*!< OCF shutdown graceful*/
 	ocf_metadata_dirty_shutdown = 0, /*!< Dirty OCF shutdown*/
-	ocf_metadata_detached = 2, /*!< Cache device detached */
+	ocf_metadata_clean_shutdown = 1, /*!< OCF shutdown graceful*/
 };
 
 /**

src/mngt/ocf_mngt_cache.c

@@ -1,6 +1,6 @@
 /*
  * Copyright(c) 2012-2022 Intel Corporation
- * Copyright(c) 2023-2024 Huawei Technologies
+ * Copyright(c) 2023-2025 Huawei Technologies
  * SPDX-License-Identifier: BSD-3-Clause
  */
 
@@ -1760,7 +1760,7 @@ static void _ocf_mngt_init_promotion(ocf_pipeline_t pipeline,
 	ocf_pipeline_next(pipeline);
 }
 
-static void _ocf_mngt_zero_superblock_complete(void *priv, int error)
+static void _ocf_mngt_attach_zero_superblock_complete(void *priv, int error)
 {
 	struct ocf_cache_attach_context *context = priv;
 	ocf_cache_t cache = context->cache;
@@ -1774,14 +1774,14 @@ static void _ocf_mngt_zero_superblock_complete(void *priv, int error)
 	ocf_pipeline_next(context->pipeline);
 }
 
-static void _ocf_mngt_zero_superblock(ocf_pipeline_t pipeline,
+static void _ocf_mngt_attach_zero_superblock(ocf_pipeline_t pipeline,
 		void *priv, ocf_pipeline_arg_t arg)
 {
 	struct ocf_cache_attach_context *context = priv;
 	ocf_cache_t cache = context->cache;
 
 	ocf_metadata_zero_superblock(cache,
-			_ocf_mngt_zero_superblock_complete, context);
+			_ocf_mngt_attach_zero_superblock_complete, context);
 }
 
 static void _ocf_mngt_attach_flush_metadata_complete(void *priv, int error)
@@ -2011,7 +2011,7 @@ struct ocf_pipeline_properties _ocf_mngt_cache_attach_pipeline_properties = {
 		OCF_PL_STEP(_ocf_mngt_attach_init_metadata),
 		OCF_PL_STEP(_ocf_mngt_attach_populate_free),
 		OCF_PL_STEP(_ocf_mngt_attach_init_services),
-		OCF_PL_STEP(_ocf_mngt_zero_superblock),
+		OCF_PL_STEP(_ocf_mngt_attach_zero_superblock),
 		OCF_PL_STEP(_ocf_mngt_attach_flush_metadata),
 		OCF_PL_STEP(_ocf_mngt_attach_discard),
 		OCF_PL_STEP(_ocf_mngt_attach_flush),
@@ -2172,6 +2172,31 @@ static void ocf_mngt_cache_stop_unplug(ocf_pipeline_t pipeline,
 			ocf_mngt_cache_stop_unplug_complete, context);
 }
 
+static void _ocf_mngt_detach_zero_superblock_complete(void *priv, int error)
+{
+	struct ocf_mngt_cache_unplug_context *context = priv;
+	ocf_cache_t cache = context->cache;
+
+	if (error) {
+		ocf_cache_log(cache, log_err,
+				"ERROR: Failed to clear the superblock on the detached device\n"
+				"The metadata on the device is in an invalid state"
+				" - manual superblock clearing is recommended.\n");
+	}
+
+	ocf_pipeline_next(context->pipeline);
+}
+
+static void _ocf_mngt_detach_zero_superblock(ocf_pipeline_t pipeline,
+		void *priv, ocf_pipeline_arg_t arg)
+{
+	struct ocf_mngt_cache_unplug_context *context = priv;
+	ocf_cache_t cache = context->cache;
+
+	ocf_metadata_zero_superblock(cache,
+			_ocf_mngt_detach_zero_superblock_complete, context);
+}
+
 static void _ocf_mngt_cache_put_io_queues(ocf_cache_t cache)
 {
 	ocf_queue_t queue, tmp_queue;
@@ -2443,7 +2468,7 @@ struct ocf_pipeline_properties _ocf_mngt_cache_standby_attach_pipeline_propertie
 		OCF_PL_STEP(_ocf_mngt_attach_populate_free),
 		OCF_PL_STEP(_ocf_mngt_standby_prepare_mempool),
 		OCF_PL_STEP(_ocf_mngt_standby_init_pio_concurrency),
-		OCF_PL_STEP(_ocf_mngt_zero_superblock),
+		OCF_PL_STEP(_ocf_mngt_attach_zero_superblock),
 		OCF_PL_STEP(_ocf_mngt_attach_flush_metadata),
 		OCF_PL_STEP(_ocf_mngt_attach_discard),
 		OCF_PL_STEP(_ocf_mngt_attach_flush),
@@ -3147,6 +3172,8 @@ static void _ocf_mngt_cache_unplug(ocf_cache_t cache, bool stop,
 		struct _ocf_mngt_cache_unplug_context *context,
 		_ocf_mngt_cache_unplug_end_t cmpl, void *priv)
 {
+	struct ocf_mngt_cache_unplug_context *ctx = priv;
+
 	context->cmpl = cmpl;
 	context->priv = priv;
 	context->cache = cache;
@@ -3157,14 +3184,13 @@ static void _ocf_mngt_cache_unplug(ocf_cache_t cache, bool stop,
 	__deinit_promotion_policy(cache);
 
 	if (!stop) {
-		/* Just set correct shutdown status */
+		/* Skip metadata update - will be zeroed later in the detach pipeline */
-		ocf_metadata_set_shutdown_status(cache, ocf_metadata_detached,
+		OCF_PL_NEXT_RET(ctx->pipeline);
-				_ocf_mngt_cache_unplug_complete, context);
-	} else {
-		/* Flush metadata */
-		ocf_metadata_flush_all(cache,
-				_ocf_mngt_cache_unplug_complete, context);
 	}
+
+	/* Flush metadata */
+	ocf_metadata_flush_all(cache,
+			_ocf_mngt_cache_unplug_complete, context);
 }
 
 static int _ocf_mngt_cache_load_core_log(ocf_core_t core, void *cntx)
@@ -3850,6 +3876,7 @@ struct ocf_pipeline_properties ocf_mngt_cache_detach_pipeline_properties = {
 		OCF_PL_STEP(ocf_mngt_cache_stop_check_dirty),
 		OCF_PL_STEP(ocf_mngt_cache_detach_update_metadata),
 		OCF_PL_STEP(ocf_mngt_cache_detach_unplug),
+		OCF_PL_STEP(_ocf_mngt_detach_zero_superblock),
 		OCF_PL_STEP(ocf_mngt_cache_close_cache_volume),
 		OCF_PL_STEP(ocf_mngt_cache_deinit_metadata),
 		OCF_PL_STEP(ocf_mngt_cache_deinit_cache_volume),

tests/functional/tests/management/test_attach_cache.py

@@ -1,16 +1,13 @@
 #
 # Copyright(c) 2019-2022 Intel Corporation
-# Copyright(c) 2024 Huawei Technologies
+# Copyright(c) 2024-2025 Huawei Technologies
 # SPDX-License-Identifier: BSD-3-Clause
 #
 
 import logging
-from ctypes import c_int, c_void_p, byref, c_uint32, memmove, cast
+from ctypes import c_void_p, memmove, cast
-from random import randrange
-from itertools import count
 
 import pytest
-
 from pyocf.types.cache import (
     Cache,
     CacheMode,
@@ -70,12 +67,32 @@ def test_detach_cache_twice(pyocf_ctx):
     cache.stop()
 
 
+def test_detach_cache_zero_superblock(pyocf_ctx):
+    """Check if superblock is zeroed after detach and the cache device can be reattached without
+    --force option.
+    """
+    cache_device = RamVolume(Size.from_MiB(50))
+    cache = Cache.start_on_device(cache_device)
+
+    cache.detach_device()
+
+    data = cache_device.get_bytes()
+
+    page_size = 4096
+    assert data[:page_size] == b'\x00'*page_size
+
+    cache.attach_device(cache_device, force=False)
+    cache.detach_device()
+
+    cache.stop()
+
+
 @pytest.mark.parametrize("cls", CacheLineSize)
 @pytest.mark.parametrize("mode", [CacheMode.WB, CacheMode.WT, CacheMode.WO])
 @pytest.mark.parametrize("new_cache_size", [80, 120])
 def test_attach_different_size(pyocf_ctx, new_cache_size, mode: CacheMode, cls: CacheLineSize):
     """Start cache and add partition with limited occupancy. Fill partition with data,
-    attach cache with different size and trigger IO. Verify if occupancy thresold is
+    attach cache with different size and trigger IO. Verify if occupancy threshold is
     respected with both original and new cache device.
     """
     cache_device = RamVolume(Size.from_MiB(100))