Merge pull request #868 from mmichal10/fix-occupancy-accounting

Fix occupancy accounting
This commit is contained in:
Robert Baldyga 2025-03-24 20:50:36 +01:00 committed by GitHub
commit 541fafd492
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
6 changed files with 72 additions and 43 deletions

View File

@ -1,6 +1,6 @@
/*
* Copyright(c) 2012-2022 Intel Corporation
* Copyright(c) 2024 Huawei Technologies
* Copyright(c) 2024-2025 Huawei Technologies
* SPDX-License-Identifier: BSD-3-Clause
*/
@ -39,6 +39,23 @@ static inline void backfill_queue_inc_block(struct ocf_cache *cache)
env_atomic_set(&cache->pending_read_misses_list_blocked, 1);
}
static int _ocf_backfill_update_metadata(struct ocf_request *req)
{
ocf_cache_t cache = req->cache;
ocf_hb_req_prot_lock_wr(req);
ocf_set_valid_map_info(req);
ocf_hb_req_prot_unlock_wr(req);
ocf_req_unlock(ocf_cache_line_concurrency(cache), req);
ocf_req_put(req);
return 0;
}
static void _ocf_backfill_complete(struct ocf_request *req, int error)
{
struct ocf_cache *cache = req->cache;
@ -61,10 +78,8 @@ static void _ocf_backfill_complete(struct ocf_request *req, int error)
if (error) {
ocf_engine_invalidate(req);
} else {
ocf_req_unlock(ocf_cache_line_concurrency(cache), req);
/* put the request at the last point of the completion path */
ocf_req_put(req);
ocf_queue_push_req_cb(req, _ocf_backfill_update_metadata,
OCF_QUEUE_ALLOW_SYNC | OCF_QUEUE_PRIO_HIGH);
}
}

View File

@ -1,6 +1,6 @@
/*
* Copyright(c) 2012-2022 Intel Corporation
* Copyright(c) 2024 Huawei Technologies
* Copyright(c) 2024-2025 Huawei Technologies
* SPDX-License-Identifier: BSD-3-Clause
*/
@ -133,13 +133,6 @@ static int _ocf_read_generic_do(struct ocf_request *req)
return 0;
}
ocf_hb_req_prot_lock_wr(req);
/* Set valid status bits map */
ocf_set_valid_map_info(req);
ocf_hb_req_prot_unlock_wr(req);
}
if (ocf_engine_needs_repart(req)) {

View File

@ -1,5 +1,6 @@
/*
* Copyright(c) 2012-2021 Intel Corporation
* Copyright(c) 2025 Huawei Technologies
* SPDX-License-Identifier: BSD-3-Clause
*/
@ -316,25 +317,28 @@ static inline void metadata_set_valid_sec_one(struct ocf_cache *cache,
ocf_metadata_set_valid(cache, line, pos, pos);
}
/*
* Marks given cache line's bits as invalid
* Marks given cache line's sectors as invalid
*
* @return true if any of the cache line's bits was valid and the cache line
* became invalid (all bits invalid) after the operation
* @return false in other cases
* @return true if line was valid and became invalid (all sectors invalid)
* @return false if line was invalid and remains invalid or
* if line was valid and still has valid sectors
*/
static inline bool metadata_clear_valid_sec_changed(
struct ocf_cache *cache, ocf_cache_line_t line,
uint8_t start, uint8_t stop, bool *is_valid)
uint8_t start, uint8_t stop, bool *line_remains_valid)
{
bool was_any_valid;
bool line_was_valid, _line_remains_valid;
was_any_valid = ocf_metadata_test_valid(cache, line, 0,
line_was_valid = ocf_metadata_test_valid(cache, line, 0,
ocf_line_end_sector(cache), false);
*is_valid = ocf_metadata_clear_valid(cache, line,
_line_remains_valid = ocf_metadata_clear_valid(cache, line,
start, stop);
return was_any_valid && !*is_valid;
if (likely(line_remains_valid != NULL))
*line_remains_valid = _line_remains_valid;
return line_was_valid && !_line_remains_valid;
}
#endif /* METADATA_STATUS_H_ */

View File

@ -1,6 +1,6 @@
/*
* Copyright(c) 2012-2021 Intel Corporation
* Copyright(c) 2024 Huawei Technologies
* Copyright(c) 2024-2025 Huawei Technologies
* SPDX-License-Identifier: BSD-3-Clause
*/
@ -12,28 +12,34 @@ static void __set_cache_line_invalid(struct ocf_cache *cache, uint8_t start_bit,
ocf_core_id_t core_id, ocf_part_id_t part_id)
{
ocf_core_t core;
bool is_valid, changed;
bool line_remains_valid, line_became_invalid;
ENV_BUG_ON(core_id >= OCF_CORE_MAX);
core = ocf_cache_get_core(cache, core_id);
changed = metadata_clear_valid_sec_changed(cache, line,
start_bit, end_bit, &is_valid);
line_became_invalid = metadata_clear_valid_sec_changed(cache, line,
start_bit, end_bit, &line_remains_valid);
/* If we have waiters, do not remove cache line
* for this cache line which will use one, clear
* only valid bits
*/
if (!is_valid && !ocf_cache_line_are_waiters(
ocf_cache_line_concurrency(cache), line)) {
if (changed) {
env_atomic_dec(&core->runtime_meta->cached_clines);
env_atomic_dec(&core->runtime_meta->
part_counters[part_id].cached_clines);
}
ocf_lru_rm_cline(cache, line);
ocf_metadata_remove_cache_line(cache, line);
if (line_remains_valid) {
ENV_BUG_ON(line_became_invalid);
return;
}
/* If there are any waiters, don't transfer the line to the freelist.
* The new owner will take care about the repart anyways
*/
if (ocf_cache_line_are_waiters(ocf_cache_line_concurrency(cache), line))
return;
ocf_lru_rm_cline(cache, line);
ocf_metadata_remove_cache_line(cache, line);
if (!line_became_invalid)
return;
env_atomic_dec(&core->runtime_meta->cached_clines);
env_atomic_dec(&core->runtime_meta->
part_counters[part_id].cached_clines);
}
void set_cache_line_invalid(struct ocf_cache *cache, uint8_t start_bit,

View File

@ -1,10 +1,11 @@
#
# 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
import time
from math import ceil, isclose
from ctypes import c_int
@ -45,6 +46,8 @@ def test_eviction_two_cores(pyocf_ctx, mode: CacheMode, cls: CacheLineSize):
send_io(vol1, test_data)
send_io(vol2, test_data)
cache.settle()
stats1 = core1.get_stats()
stats2 = core2.get_stats()
# IO to the second core should evict all the data from the first core
@ -285,6 +288,7 @@ def test_eviction_freelist(pyocf_ctx, cls: CacheLineSize, cache_mode: CacheMode,
for j in range(cache_lines_written):
addr = (cache_lines_written * i + j) * data.size
send_io(vol, data, addr, ioclass, io_dir)
cache.settle()
assert (
get_ioclass_occupancy(cache, ioclass) == expected_occpancy_4k
), f"Doesn't match for ioclass {ioclass}"
@ -298,6 +302,9 @@ def test_eviction_freelist(pyocf_ctx, cls: CacheLineSize, cache_mode: CacheMode,
addr += data.size
send_io(vol, data, addr, high_prio_ioclass, io_dir)
cache.settle()
time.sleep(1)
assert cache.get_stats()["usage"]["occupancy"]["value"] == cache_size_4k
for ioclass in low_prio_ioclasses:

View File

@ -1,6 +1,6 @@
#
# Copyright(c) 2019-2022 Intel Corporation
# Copyright(c) 2024 Huawei Technologies
# Copyright(c) 2024-2025 Huawei Technologies
# SPDX-License-Identifier: BSD-3-Clause
#
@ -212,7 +212,9 @@ def test_stop(pyocf_ctx, mode: CacheMode, cls: CacheLineSize, with_flush: bool):
cls_no = 10
run_io_and_cache_data_if_possible(front_vol, mode, cls, cls_no)
run_io_and_cache_data_if_possible(cache, front_vol, mode, cls, cls_no)
cache.settle()
stats = cache.get_stats()
assert int(stats["conf"]["dirty"]) == (
@ -494,7 +496,7 @@ def test_start_stop_noqueue(pyocf_ctx):
assert not c.results["error"], "Failed to stop cache: {}".format(c.results["error"])
def run_io_and_cache_data_if_possible(vol, mode, cls, cls_no):
def run_io_and_cache_data_if_possible(cache, vol, mode, cls, cls_no):
queue = vol.parent.get_default_queue()
test_data = Data(cls_no * cls)
@ -508,6 +510,8 @@ def run_io_and_cache_data_if_possible(vol, mode, cls, cls_no):
logger.info("[STAGE] Write to exported object")
io_to_core(vol, queue, test_data, 0)
cache.settle()
stats = vol.parent.cache.get_stats()
assert stats["usage"]["occupancy"]["value"] == (
(cls_no * cls / CacheLineSize.LINE_4KiB) if mode != CacheMode.PT else 0