Commit eebbd951 authored by Tomasz Zawadzki's avatar Tomasz Zawadzki
Browse files

lib/blob: pass Extent Page offset on cluster allocation 



Extent Pages claim and insertion can be asynchronous
when cluster allocation happens due to writing to a new cluster.

In such case lowest free cluster and lowest free md page
is claimed, and message is passed to md_thread.
Where inserting both into the arrays and md_sycn happens.

This patch adds parameters to pass the Extent Page offset
in such case.

Signed-off-by: default avatarTomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I46d8ace9cd5abc0bfe48174c2f2ec218145b9c75
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/479849


Tested-by: default avatarSPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: default avatarBen Walker <benjamin.walker@intel.com>
Reviewed-by: default avatarPaul Luse <paul.e.luse@intel.com>
Reviewed-by: default avatarJim Harris <james.r.harris@intel.com>
parent f60b4a7e

lib/blob/blobstore.c

+12 −7
Original line number Diff line number Diff line
@@ -54,7 +54,7 @@ static int spdk_bs_register_md_thread(struct spdk_blob_store *bs); 
static int spdk_bs_unregister_md_thread(struct spdk_blob_store *bs);

static void _spdk_blob_close_cpl(spdk_bs_sequence_t *seq, void *cb_arg, int bserrno);

static void _spdk_blob_insert_cluster_on_md_thread(struct spdk_blob *blob, uint32_t cluster_num,

		uint64_t cluster, spdk_blob_op_complete cb_fn, void *cb_arg);

		uint64_t cluster, uint32_t extent, spdk_blob_op_complete cb_fn, void *cb_arg);



static int _spdk_blob_set_xattr(struct spdk_blob *blob, const char *name, const void *value,

				uint16_t value_len, bool internal);
@@ -114,7 +114,7 @@ _spdk_blob_insert_cluster(struct spdk_blob *blob, uint32_t cluster_num, uint64_t 


static int

_spdk_bs_allocate_cluster(struct spdk_blob *blob, uint32_t cluster_num,

			  uint64_t *lowest_free_cluster, bool update_map)

			  uint64_t *lowest_free_cluster, uint32_t *lowest_free_md_page, bool update_map)

{

	pthread_mutex_lock(&blob->bs->used_clusters_mutex);

	*lowest_free_cluster = spdk_bit_array_find_first_clear(blob->bs->used_clusters,
@@ -1555,6 +1555,7 @@ _spdk_blob_resize(struct spdk_blob *blob, uint64_t sz) 
	uint64_t	i;

	uint64_t	*tmp;

	uint64_t	lfc; /* lowest free cluster */

	uint32_t	lfmd; /*  lowest free md page */

	uint64_t	num_clusters;

	struct spdk_blob_store *bs;
@@ -1614,9 +1615,11 @@ _spdk_blob_resize(struct spdk_blob *blob, uint64_t sz) 


	if (spdk_blob_is_thin_provisioned(blob) == false) {

		lfc = 0;

		lfmd = 0;

		for (i = num_clusters; i < sz; i++) {

			_spdk_bs_allocate_cluster(blob, i, &lfc, true);

			_spdk_bs_allocate_cluster(blob, i, &lfc, &lfmd, true);

			lfc++;

			lfmd++;

		}

	}
@@ -1781,6 +1784,7 @@ struct spdk_blob_copy_cluster_ctx { 
	uint8_t *buf;

	uint64_t page;

	uint64_t new_cluster;

	uint32_t new_extent_page;

	spdk_bs_sequence_t *seq;

};
@@ -1842,7 +1846,7 @@ _spdk_blob_write_copy_cpl(spdk_bs_sequence_t *seq, void *cb_arg, int bserrno) 
	cluster_number = _spdk_bs_page_to_cluster(ctx->blob->bs, ctx->page);



	_spdk_blob_insert_cluster_on_md_thread(ctx->blob, cluster_number, ctx->new_cluster,

					       _spdk_blob_insert_cluster_cpl, ctx);

					       ctx->new_extent_page, _spdk_blob_insert_cluster_cpl, ctx);

}



static void
@@ -1915,7 +1919,8 @@ _spdk_bs_allocate_and_copy_cluster(struct spdk_blob *blob, 
		}

	}



	rc = _spdk_bs_allocate_cluster(blob, cluster_number, &ctx->new_cluster, false);

	rc = _spdk_bs_allocate_cluster(blob, cluster_number, &ctx->new_cluster, &ctx->new_extent_page,

				       false);

	if (rc != 0) {

		spdk_free(ctx->buf);

		free(ctx);
@@ -1947,7 +1952,7 @@ _spdk_bs_allocate_and_copy_cluster(struct spdk_blob *blob, 
					     _spdk_blob_write_copy, ctx);

	} else {

		_spdk_blob_insert_cluster_on_md_thread(ctx->blob, cluster_number, ctx->new_cluster,

						       _spdk_blob_insert_cluster_cpl, ctx);

						       ctx->new_extent_page, _spdk_blob_insert_cluster_cpl, ctx);

	}

}
@@ -6150,7 +6155,7 @@ _spdk_blob_insert_cluster_msg(void *arg) 


static void

_spdk_blob_insert_cluster_on_md_thread(struct spdk_blob *blob, uint32_t cluster_num,

				       uint64_t cluster, spdk_blob_op_complete cb_fn, void *cb_arg)

				       uint64_t cluster, uint32_t extent_page, spdk_blob_op_complete cb_fn, void *cb_arg)

{

	struct spdk_blob_insert_cluster_ctx *ctx;

test/unit/lib/blob/blob.c/blob_ut.c

+4 −2
Original line number Diff line number Diff line
@@ -4521,6 +4521,7 @@ blob_insert_cluster_msg(void) 
	uint64_t free_clusters;

	uint64_t new_cluster = 0;

	uint32_t cluster_num = 3;

	uint32_t extent_page = 0;



	dev = init_dev();
@@ -4556,10 +4557,11 @@ blob_insert_cluster_msg(void) 
	/* Specify cluster_num to allocate and new_cluster will be returned to insert on md_thread.

	 * This is to simulate behaviour when cluster is allocated after blob creation.

	 * Such as _spdk_bs_allocate_and_copy_cluster(). */

	_spdk_bs_allocate_cluster(blob, cluster_num, &new_cluster, false);

	_spdk_bs_allocate_cluster(blob, cluster_num, &new_cluster, &extent_page, false);

	CU_ASSERT(blob->active.clusters[cluster_num] == 0);



	_spdk_blob_insert_cluster_on_md_thread(blob, cluster_num, new_cluster, blob_op_complete, NULL);

	_spdk_blob_insert_cluster_on_md_thread(blob, cluster_num, new_cluster, extent_page,

					       blob_op_complete, NULL);

	poll_threads();



	CU_ASSERT(blob->active.clusters[cluster_num] != 0);