Commit d0dfbef6 authored by Jim Harris's avatar Jim Harris Committed by Konrad Sztyber
Browse files

test/unit/blob: introduce g_phys_blocklen 



Up until now, the unit tests were all just hardcoded for 4096B md
page size. Upcoming patches will enable larger md page sizes. So
introduce a suite parameter "g_phys_blocklen" which for now will
always be 4096. Future patches will emulate block devices with a
larger phys_blocklen, which is used as the md page size by default.

Signed-off-by: default avatarAtul Malakar <a.malakar@samsung.com>
Signed-off-by: default avatarJim Harris <jim.harris@samsung.com>
Change-Id: I133113474938d2b9490161a856a32ebfe659d674
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/25309


Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: default avatarSPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: default avatarBen Walker <ben@nvidia.com>
Community-CI: Community CI Samsung <spdk.community.ci.samsung@gmail.com>
Community-CI: Mellanox Build Bot
Reviewed-by: default avatarKonrad Sztyber <konrad.sztyber@intel.com>
parent 557f022f

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

+35 −21
Original line number Diff line number Diff line
@@ -2412,7 +2412,7 @@ bs_load(void) 


	/* Load should fail for device with an unsupported blocklen */

	dev = init_dev();

	dev->blocklen = BLOCKLEN * 2;

	dev->blocklen = g_phys_blocklen * 2;

	spdk_bs_load(dev, NULL, bs_op_with_handle_complete, NULL);

	poll_threads();

	CU_ASSERT(g_bserrno == -EINVAL);
@@ -2770,7 +2770,8 @@ bs_load_after_failed_grow(void) 
	super_block = (struct spdk_bs_super_block *)g_dev_buffer;

	CU_ASSERT(super_block->clean == 1);



	mask = (struct spdk_bs_md_mask *)(g_dev_buffer + super_block->used_cluster_mask_start * BLOCKLEN);

	mask = (struct spdk_bs_md_mask *)(g_dev_buffer + super_block->used_cluster_mask_start *

					  g_phys_blocklen);

	CU_ASSERT(mask->type == SPDK_MD_MASK_TYPE_USED_CLUSTERS);

	CU_ASSERT(mask->length == super_block->size / super_block->cluster_size);
@@ -2993,7 +2994,7 @@ bs_test_recover_cluster_count(void) 
	super_block.length = 0x1000;

	super_block.clean = 0;

	super_block.super_blob = 0xFFFFFFFFFFFFFFFF;

	super_block.cluster_size = BLOCKLEN;

	super_block.cluster_size = g_phys_blocklen;

	super_block.used_page_mask_start = 0x01;

	super_block.used_page_mask_len = 0x01;

	super_block.used_cluster_mask_start = 0x02;
@@ -3068,7 +3069,7 @@ bs_grow_live_size(uint64_t new_blockcnt) 
	CU_ASSERT(super_block.size == bdev_size);

	CU_ASSERT(super_block.clean == 0);

	/* The used_cluster mask is not written out until first spdk_bs_unload. */

	memcpy(&mask, g_dev_buffer + super_block.used_cluster_mask_start * BLOCKLEN,

	memcpy(&mask, g_dev_buffer + super_block.used_cluster_mask_start * g_phys_blocklen,

	       sizeof(struct spdk_bs_md_mask));

	CU_ASSERT(mask.type == 0);

	CU_ASSERT(mask.length == 0);
@@ -3082,7 +3083,7 @@ bs_grow_live_size(uint64_t new_blockcnt) 
	memcpy(&super_block, g_dev_buffer, sizeof(struct spdk_bs_super_block));

	CU_ASSERT(super_block.size == bdev_size);

	CU_ASSERT(super_block.clean == 1);

	memcpy(&mask, g_dev_buffer + super_block.used_cluster_mask_start * BLOCKLEN,

	memcpy(&mask, g_dev_buffer + super_block.used_cluster_mask_start * g_phys_blocklen,

	       sizeof(struct spdk_bs_md_mask));

	CU_ASSERT(mask.type == SPDK_MD_MASK_TYPE_USED_CLUSTERS);

	CU_ASSERT(mask.length == bdev_size / (1 * 1024 * 1024));
@@ -3197,7 +3198,7 @@ bs_grow_live_no_space(void) 
	memcpy(&super_block, g_dev_buffer, sizeof(struct spdk_bs_super_block));

	CU_ASSERT(super_block.size == bdev_size_init);

	CU_ASSERT(super_block.clean == 1);

	memcpy(&mask, g_dev_buffer + super_block.used_cluster_mask_start * BLOCKLEN,

	memcpy(&mask, g_dev_buffer + super_block.used_cluster_mask_start * g_phys_blocklen,

	       sizeof(struct spdk_bs_md_mask));

	CU_ASSERT(mask.type == SPDK_MD_MASK_TYPE_USED_CLUSTERS);

	CU_ASSERT(mask.length == bdev_size_init / (1 * 1024 * 1024));
@@ -3251,7 +3252,7 @@ bs_test_grow(void) 
	/*

	 * Make sure the used_cluster mask is correct.

	 */

	memcpy(&mask, g_dev_buffer + super_block.used_cluster_mask_start * BLOCKLEN,

	memcpy(&mask, g_dev_buffer + super_block.used_cluster_mask_start * g_phys_blocklen,

	       sizeof(struct spdk_bs_md_mask));

	CU_ASSERT(mask.type == SPDK_MD_MASK_TYPE_USED_CLUSTERS);

	CU_ASSERT(mask.length == bdev_size / (1 * 1024 * 1024));
@@ -3284,7 +3285,7 @@ bs_test_grow(void) 
	/*

	 * Make sure the used_cluster mask has been updated according to the bdev size

	 */

	memcpy(&mask, g_dev_buffer + super_block.used_cluster_mask_start * BLOCKLEN,

	memcpy(&mask, g_dev_buffer + super_block.used_cluster_mask_start * g_phys_blocklen,

	       sizeof(struct spdk_bs_md_mask));

	CU_ASSERT(mask.type == SPDK_MD_MASK_TYPE_USED_CLUSTERS);

	CU_ASSERT(mask.length == bdev_size / (1 * 1024 * 1024));
@@ -3350,7 +3351,7 @@ bs_cluster_sz(void) 
	 */

	dev = init_dev();

	spdk_bs_opts_init(&opts, sizeof(opts));

	opts.cluster_sz = SPDK_BS_PAGE_SIZE;

	opts.cluster_sz = g_phys_blocklen;



	/* Initialize a new blob store */

	spdk_bs_init(dev, &opts, bs_op_with_handle_complete, NULL);
@@ -3364,7 +3365,7 @@ bs_cluster_sz(void) 
	 */

	dev = init_dev();

	spdk_bs_opts_init(&opts, sizeof(opts));

	opts.cluster_sz = SPDK_BS_PAGE_SIZE - 1;

	opts.cluster_sz = g_phys_blocklen - 1;



	/* Initialize a new blob store */

	spdk_bs_init(dev, &opts, bs_op_with_handle_complete, NULL);
@@ -3463,7 +3464,7 @@ bs_resize_md(void) 


	dev = init_dev();

	spdk_bs_opts_init(&opts, sizeof(opts));

	opts.cluster_sz = CLUSTER_PAGE_COUNT * BLOCKLEN;

	opts.cluster_sz = CLUSTER_PAGE_COUNT * g_phys_blocklen;

	cluster_sz = opts.cluster_sz;



	/* Initialize a new blob store */
@@ -3648,7 +3649,7 @@ blob_crc(void) 
	CU_ASSERT(g_bserrno == 0);



	page_num = bs_blobid_to_page(blobid);

	index = DEV_BUFFER_BLOCKLEN * (bs->md_start + page_num);

	index = g_phys_blocklen * (bs->md_start + page_num);

	page = (struct spdk_blob_md_page *)&g_dev_buffer[index];

	page->crc = 0;
@@ -3949,7 +3950,7 @@ blob_dirty_shutdown(void) 
	/* Mark second blob as invalid */

	page_num = bs_blobid_to_page(blobid2);



	index = DEV_BUFFER_BLOCKLEN * (bs->md_start + page_num);

	index = g_phys_blocklen * (bs->md_start + page_num);

	page = (struct spdk_blob_md_page *)&g_dev_buffer[index];

	page->sequence_num = 1;

	page->crc = blob_md_page_calc_crc(page);
@@ -4357,7 +4358,16 @@ blob_insert_cluster_msg_test(void) 
	struct spdk_blob_store *bs = g_bs;

	struct spdk_blob *blob;

	struct spdk_blob_opts opts;

	struct spdk_blob_md_page page = {};

	/* For now, even if md_page_size is > 4KB, we still only use the first

	 * 4KB of it. The rest is left unused. Future changes may allow using the

	 * rest of the md_page, but that will require more extensive changes since

	 * then the struct spdk_blob_md_page cannot be used directly (since some

	 * fields such as crc would have variable placement in the struct).

	 */

	struct {

		struct spdk_blob_md_page page;

		uint8_t pad[DEV_MAX_PHYS_BLOCKLEN - sizeof(struct spdk_blob_md_page)];

	} md = {};

	spdk_blob_id blobid;

	uint64_t free_clusters;

	uint64_t new_cluster = 0;
@@ -4387,7 +4397,7 @@ blob_insert_cluster_msg_test(void) 
	CU_ASSERT(blob->active.clusters[cluster_num] == 0);

	spdk_spin_unlock(&bs->used_lock);



	blob_insert_cluster_on_md_thread(blob, cluster_num, new_cluster, extent_page, &page,

	blob_insert_cluster_on_md_thread(blob, cluster_num, new_cluster, extent_page, &md.page,

					 blob_op_complete, NULL);

	poll_threads();
@@ -4528,7 +4538,7 @@ blob_thin_prov_write_count_io(void) 
	uint8_t payload_write[BLOCKLEN];

	uint64_t write_bytes;

	uint64_t read_bytes;

	const uint32_t CLUSTER_SZ = 16384;

	const uint32_t CLUSTER_SZ = g_phys_blocklen * 4;

	uint32_t pages_per_cluster;

	uint32_t pages_per_extent_page;

	uint32_t i;
@@ -7580,7 +7590,7 @@ blob_io_unit_compatibility(void) 
	/* Create dev with 512 bytes io unit size */



	spdk_bs_opts_init(&bsopts, sizeof(bsopts));

	bsopts.cluster_sz = SPDK_BS_PAGE_SIZE * 4;	/* 8 * 4 = 32 io_unit */

	bsopts.cluster_sz = g_phys_blocklen * 4;

	snprintf(bsopts.bstype.bstype, sizeof(bsopts.bstype.bstype), "TESTTYPE");



	/* Try to initialize a new blob store with unsupported io_unit */
@@ -8365,7 +8375,7 @@ blob_esnap_io_size(uint32_t bs_blksz, uint32_t esnap_blksz) 
	struct spdk_blob_opts	opts;

	struct ut_esnap_opts	esnap_opts;

	struct spdk_blob	*blob;

	const uint32_t		cluster_sz = 16 * 1024;

	const uint32_t		cluster_sz = 4 * g_phys_blocklen;

	const uint64_t		esnap_num_clusters = 4;

	const uint32_t		esnap_sz = cluster_sz * esnap_num_clusters;

	const uint64_t		esnap_num_blocks = esnap_sz / esnap_blksz;
@@ -8503,7 +8513,7 @@ blob_esnap_io_512_4096(void) 
	struct spdk_bs_opts	bs_opts;

	struct spdk_blob_opts	blob_opts;

	struct ut_esnap_opts	esnap_opts;

	uint64_t		cluster_sz = 16 * 1024;

	uint64_t		cluster_sz = 4 * g_phys_blocklen;

	uint32_t		bs_blksz = 512;

	uint32_t		esnap_blksz = BLOCKLEN;

	uint64_t		esnap_num_blocks = 64;
@@ -9315,7 +9325,7 @@ blob_esnap_clone_resize(void) 
	struct ut_esnap_opts esnap_opts;

	struct spdk_blob *blob;

	uint32_t block, esnap_blksz = 512, bs_blksz = 512;

	const uint32_t cluster_sz = 16 * 1024;

	const uint32_t cluster_sz = 4 * g_phys_blocklen;

	const uint64_t esnap_num_clusters = 4;

	const uint32_t esnap_sz = cluster_sz * esnap_num_clusters;

	const uint64_t esnap_num_blocks = esnap_sz / esnap_blksz;
@@ -9895,7 +9905,7 @@ suite_esnap_bs_setup(void) 
	dev = init_dev();

	memset(g_dev_buffer, 0, DEV_BUFFER_SIZE);

	spdk_bs_opts_init(&bs_opts, sizeof(bs_opts));

	bs_opts.cluster_sz = 16 * 1024;

	bs_opts.cluster_sz = 4 * g_phys_blocklen;

	bs_opts.esnap_bs_dev_create = ut_esnap_create;

	spdk_bs_init(dev, &bs_opts, bs_op_with_handle_complete, NULL);

	poll_threads();
@@ -9987,6 +9997,7 @@ ut_setup_config_nocopy_noextent(void) 
{

	g_dev_copy_enabled = false;

	g_use_extent_table = false;

	g_phys_blocklen = 4096;



	return 0;

}
@@ -9996,6 +10007,7 @@ ut_setup_config_nocopy_extent(void) 
{

	g_dev_copy_enabled = false;

	g_use_extent_table = true;

	g_phys_blocklen = 4096;



	return 0;

}
@@ -10005,6 +10017,7 @@ ut_setup_config_copy_noextent(void) 
{

	g_dev_copy_enabled = true;

	g_use_extent_table = false;

	g_phys_blocklen = 4096;



	return 0;

}
@@ -10014,6 +10027,7 @@ ut_setup_config_copy_extent(void) 
{

	g_dev_copy_enabled = true;

	g_use_extent_table = true;

	g_phys_blocklen = 4096;



	return 0;

}

test/unit/lib/blob/bs_dev_common.c

+2 −0
Original line number Diff line number Diff line
@@ -11,6 +11,7 @@ 
#define DEV_BUFFER_SIZE (64 * 1024 * 1024)

#define DEV_BUFFER_BLOCKLEN (4096)

#define DEV_BUFFER_BLOCKCNT (DEV_BUFFER_SIZE / DEV_BUFFER_BLOCKLEN)

#define DEV_MAX_PHYS_BLOCKLEN (16384)

uint8_t *g_dev_buffer;

uint64_t g_dev_write_bytes;

uint64_t g_dev_read_bytes;
@@ -19,6 +20,7 @@ bool g_dev_writev_ext_called; 
bool g_dev_readv_ext_called;

bool g_dev_copy_enabled;

struct spdk_blob_ext_io_opts g_blob_ext_io_opts;

uint32_t g_phys_blocklen;



struct spdk_power_failure_counters {

	uint64_t general_counter;