github/ocf
4
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Allocate requests for management path separately

Browse Source 
Management path does not benefit much from mpools, as number of requests
allocated is very small. It's less restrictive (mngt_queue does not have
single-CPU affinity) thus avoiding mpool usage in management path allows
to introduce additional restrictions on mpool, leading to I/O performance
improvement.

Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
This commit is contained in:
Robert Baldyga
2023-09-25 15:07:30 +02:00
committed by Michal Mielewczyk
parent 6cd5a27ea9
commit 460cd461d3
9 changed files with 138 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
src/utils/utils_cleaner.c
View File

@@ -40,8 +40,7 @@
static struct ocf_request *_ocf_cleaner_alloc_req(struct ocf_cache *cache,
uint32_t count, const struct ocf_cleaner_attribs *attribs)
{
struct ocf_request *req = ocf_req_new_extended(attribs->io_queue, NULL,
0, count * ocf_line_size(cache), OCF_READ);
struct ocf_request *req = ocf_req_new_cleaner(attribs->io_queue, count);
int ret;
if (!req)

3
src/utils/utils_parallelize.c
View File

@@ -105,8 +105,7 @@ int ocf_parallelize_create(ocf_parallelize_t *parallelize,
} else {
queue = cache->mngt_queue;
}
tmp_parallelize->reqs[i] = ocf_req_new(queue,
NULL, 0, 0, 0);
tmp_parallelize->reqs[i] = ocf_req_new_mngt(queue);
if (!tmp_parallelize->reqs[i]) {
result = -OCF_ERR_NO_MEM;
goto err_reqs;

2
src/utils/utils_pipeline.c
View File

@@ -87,7 +87,7 @@ int ocf_pipeline_create(ocf_pipeline_t *pipeline, ocf_cache_t cache,
tmp_pipeline->priv = (void *)priv;
}
req = ocf_req_new(cache->mngt_queue, NULL, 0, 0, 0);
req = ocf_req_new_mngt(cache->mngt_queue);
if (!req) {
env_vfree(tmp_pipeline);
return -OCF_ERR_NO_MEM;