ivampiresp/Spdk
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

lib/blob: change replay of extent pages to batching

Browse Source 
Before this patch reading out the extent pages during
blobstore replay was serialized. Only issuing reads for next
extent page when previous operation finished.

This was done by continously calling _spdk_bs_load_replay_extent_page_cpl()
and decreasing ctx->num_extent_pages.

This patch changes spdk_bs_sequence_* to spdk_bs_batch_*.
All the reads are submitted at once, and only when all of them
finish we proceed to next valid md chain.

Goal of this change is improving efficiency and readability.

Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Change-Id: I807cdb98166e04706fedb494363f5776e3151827
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1540
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This commit is contained in:
Tomasz Zawadzki 2020-03-27 07:22:24 -04:00
parent 22dc6254d6
commit 955c58fa68
1 changed files with 35 additions and 33 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

68
lib/blob/blobstore.c
View File

@ -3917,13 +3917,12 @@ _spdk_bs_load_replay_md_chain_cpl(struct spdk_bs_load_ctx *ctx)
}
}
static void _spdk_bs_load_replay_extent_page(spdk_bs_sequence_t *seq, uint32_t page, void *cb_arg);
static void
_spdk_bs_load_replay_extent_page_cpl(spdk_bs_sequence_t *seq, void *cb_arg, int bserrno)
{
struct spdk_bs_load_ctx *ctx = cb_arg;
uint32_t page_num;
uint64_t i;
if (bserrno != 0) {
spdk_free(ctx->extent_pages);
@ -3931,53 +3930,58 @@ _spdk_bs_load_replay_extent_page_cpl(spdk_bs_sequence_t *seq, void *cb_arg, int
return;
}
/* Extent pages are only read when present within in chain md.
* Integrity of md is not right if that page was not a valid extent page. */
if (_spdk_bs_load_cur_extent_page_valid(&ctx->extent_pages[0]) != true) {
spdk_free(ctx->extent_pages);
_spdk_bs_load_ctx_fail(ctx, -EILSEQ);
return;
}
for (i = 0; i < ctx->num_extent_pages; i++) {
/* Extent pages are only read when present within in chain md.
* Integrity of md is not right if that page was not a valid extent page. */
if (_spdk_bs_load_cur_extent_page_valid(&ctx->extent_pages[i]) != true) {
spdk_free(ctx->extent_pages);
_spdk_bs_load_ctx_fail(ctx, -EILSEQ);
return;
}
page_num = ctx->extent_page_num[ctx->num_extent_pages - 1];
spdk_bit_array_set(ctx->bs->used_md_pages, page_num);
if (_spdk_bs_load_replay_md_parse_page(ctx, &ctx->extent_pages[0])) {
spdk_free(ctx->extent_pages);
_spdk_bs_load_ctx_fail(ctx, -EILSEQ);
return;
}
ctx->num_extent_pages--;
if (ctx->num_extent_pages > 0) {
spdk_free(ctx->extent_pages);
_spdk_bs_load_replay_extent_page(seq, ctx->extent_page_num[ctx->num_extent_pages - 1], ctx);
return;
page_num = ctx->extent_page_num[i];
spdk_bit_array_set(ctx->bs->used_md_pages, page_num);
if (_spdk_bs_load_replay_md_parse_page(ctx, &ctx->extent_pages[i])) {
spdk_free(ctx->extent_pages);
_spdk_bs_load_ctx_fail(ctx, -EILSEQ);
return;
}
}
spdk_free(ctx->extent_pages);
free(ctx->extent_page_num);
ctx->extent_page_num = NULL;
ctx->num_extent_pages = 0;
_spdk_bs_load_replay_md_chain_cpl(ctx);
}
static void
_spdk_bs_load_replay_extent_page(spdk_bs_sequence_t *seq, uint32_t page, void *cb_arg)
_spdk_bs_load_replay_extent_pages(struct spdk_bs_load_ctx *ctx)
{
struct spdk_bs_load_ctx *ctx = cb_arg;
spdk_bs_batch_t *batch;
uint32_t page;
uint64_t lba;
uint64_t i;
ctx->extent_pages = spdk_zmalloc(SPDK_BS_PAGE_SIZE, SPDK_BS_PAGE_SIZE,
ctx->extent_pages = spdk_zmalloc(SPDK_BS_PAGE_SIZE * ctx->num_extent_pages, SPDK_BS_PAGE_SIZE,
NULL, SPDK_ENV_SOCKET_ID_ANY, SPDK_MALLOC_DMA);
if (!ctx->extent_pages) {
_spdk_bs_load_ctx_fail(ctx, -ENOMEM);
return;
}
assert(page < ctx->super->md_len);
lba = _spdk_bs_md_page_to_lba(ctx->bs, page);
spdk_bs_sequence_read_dev(seq, &ctx->extent_pages[0], lba,
_spdk_bs_byte_to_lba(ctx->bs, SPDK_BS_PAGE_SIZE),
_spdk_bs_load_replay_extent_page_cpl, ctx);
batch = spdk_bs_sequence_to_batch(ctx->seq, _spdk_bs_load_replay_extent_page_cpl, ctx);
for (i = 0; i < ctx->num_extent_pages; i++) {
page = ctx->extent_page_num[i];
assert(page < ctx->super->md_len);
lba = _spdk_bs_md_page_to_lba(ctx->bs, page);
spdk_bs_batch_read_dev(batch, &ctx->extent_pages[i], lba,
_spdk_bs_byte_to_lba(ctx->bs, SPDK_BS_PAGE_SIZE));
}
spdk_bs_batch_close(batch);
}
static void
@ -4011,9 +4015,7 @@ _spdk_bs_load_replay_md_cpl(spdk_bs_sequence_t *seq, void *cb_arg, int bserrno)
return;
}
if (ctx->num_extent_pages != 0) {
/* Extent pages are read from last to first,
* decreasing the num_extent_pages as they are read. */
_spdk_bs_load_replay_extent_page(seq, ctx->extent_page_num[ctx->num_extent_pages - 1], ctx);
_spdk_bs_load_replay_extent_pages(ctx);
return;
}
}