Daniel Madej
a15003d43e
Split count_pages and update on metadata deinit
...
This resets count_pages_variable on cache-detach, so during the following
cache-attach metadata size is calculated properly.
Signed-off-by: Daniel Madej <daniel.madej@huawei.com>
2025-02-06 11:54:53 +01:00
Robert Baldyga
ba4b81a94c
Merge pull request #863 from Deixx/detach-zero-superblock
...
Zero metadata superblock on detach
2025-02-06 11:22:55 +01:00
Daniel Madej
3263503de1
pyocf: Add test for re-attaching device w/o force
...
Signed-off-by: Daniel Madej <daniel.madej@huawei.com>
2025-02-06 09:24:40 +01:00
Daniel Madej
95c9c8987e
Zero metadata superblock on detach
...
Metadata from a detached cache device is not meant to be loaded.
Signed-off-by: Daniel Madej <daniel.madej@huawei.com>
2025-02-06 09:24:40 +01:00
Robert Baldyga
c200c24344
Merge pull request #851 from robertbaldyga/core-stat-fix-inv-waiter
...
Decrement core stats only if the cache line mapping is cleaned
2025-02-06 06:15:33 +01:00
Robert Baldyga
be068df400
Merge pull request #853 from mmichal10/repart
...
Repart
2025-02-04 16:39:49 +01:00
Robert Baldyga
08eb00665c
Merge pull request #854 from robertbaldyga/request-cleanup
...
A little cleanup between ocf_request and ocf_io
2025-02-04 15:20:44 +01:00
Robert Baldyga
c029f78e95
Make alock_rw a bit field
...
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2025-02-04 12:05:26 +01:00
Robert Baldyga
0e294fc75d
Merge pull request #859 from Deixx/detach-unfreeze-fix
...
ocf_cleaner_refcnt_unfreeze bug fix
2025-02-04 11:40:14 +01:00
Daniel Madej
8ce129de06
ocf_cleaner_refcnt_unfreeze bug fix
...
During core remove/detach ocf_cleaner_refcnt_freeze was called only
when cache was attached, but ocf_cleaner_refcnt_unfreeze was called
regardless of cache state.
Signed-off-by: Daniel Madej <daniel.madej@huawei.com>
2025-01-15 18:51:42 +01:00
Robert Baldyga
e630b811ff
Merge pull request #858 from robertbaldyga/attach-fix-race-condition
...
Fix race condition during cache attach
2024-11-25 19:37:03 +01:00
Robert Baldyga
b850727d17
tests: Fix d2c test
...
Cache attach operation is not supposed to complete unless all the d2c
requests are completed, thus need to handle it asynchronously.
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2024-11-21 21:29:23 +01:00
Robert Baldyga
0d06b3a597
Fix race condition during cache attach
...
After attaching new cache device handle all the IOs in Pass-Through mode
until all the d2c requests are completed.
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2024-11-21 21:26:00 +01:00
Kamil Gierszewski
42004058cc
github-actions: add checkpatch
...
Signed-off-by: Kamil Gierszewski <kamil.gierszewski@huawei.com>
2024-11-05 12:55:49 +01:00
Robert Baldyga
2c28f33e37
Merge pull request #857 from robertbaldyga/fix-ubasn-false-positive-2
...
Avoid UBSAN false positive - continuation
2024-10-30 15:22:55 +01:00
Robert Baldyga
2da05343e2
Avoid UBSAN false positive - continuation
...
Address issue in two more instances.
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2024-10-30 11:33:17 +01:00
Michal Mielewczyk
91ad6476e2
pyocf: Cache lines reclassification test
...
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-21 19:32:51 +02:00
Michal Mielewczyk
a18bee3d23
pyocf: ioclass statas api
...
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-21 19:32:51 +02:00
Michal Mielewczyk
6bfd2122d2
Repart all misclassified cache lines
...
Remapped cache lines my require repartitioning as well
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-16 18:37:42 +02:00
Michal Mielewczyk
b3fa1fc96a
Reset repart flag during refreshing request status
...
The flag isn't reset before retraversation so it might be true even if cache
line was reparted in the meantime
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-16 18:37:42 +02:00
Robert Baldyga
b3332793bb
Remove unused request field
...
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2024-10-15 16:15:02 +02:00
Robert Baldyga
0cf4e8124b
Remove io.ref_count
...
There is already refcounting on the request. No need for additional one.
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2024-10-15 09:55:56 +02:00
Robert Baldyga
85513332d7
Remove ocf_io_get()
...
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2024-10-15 09:55:56 +02:00
Robert Baldyga
a9718eeab1
Simplify fastpath handling
...
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2024-10-15 09:55:56 +02:00
Robert Baldyga
1ab882aa61
Merge pull request #852 from robertbaldyga/forward_io-fix-error-accounting
...
Fix error accounting in forward_io
2024-10-15 08:44:47 +02:00
Robert Baldyga
e5a2042e0d
Merge pull request #847 from mmichal10/pre-refcnt-fixes
...
Refactors&fixes
2024-10-14 17:31:54 +02:00
Robert Baldyga
a1af1809d8
Fix error accounting in forward_io
...
Resetting cache_error/core_error in ocf_req_forward_* functions may lead
to overwriting already reported error if the forward is being done in the
loop.
To avoid this potential problem, introduce set of forward init functions
intended to be called before the entire forward operation, which resets
the error code and sets a forward callback.
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2024-10-14 16:54:51 +02:00
Robert Baldyga
3959f58f1c
Decrement core stats only if the cache line mapping is cleaned
...
We skip cleaning the mapping and moving the cache line to the free list
in a situation when there are waiters on the cache line lock.
Refrain from decrementing core stats in that situation as well, and let
repart of the waiter request do the updates as needed.
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2024-10-14 13:28:56 +02:00
Michal Mielewczyk
1b2a9e03c3
Add missing cache unlock in init rollback
...
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-10 07:32:38 +02:00
Michal Mielewczyk
c82fd173c6
Remove redundant list_del(ctx->caches) during init
...
New caches are added to the list at the point where they are already
initialized and no errors are possible at this point, hence list_del() in error
handling is redundant.
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-10 07:32:38 +02:00
Michal Mielewczyk
d22885ef7d
example: Introduce error handling
...
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-10 07:32:38 +02:00
Michal Mielewczyk
e8e7a1600c
Log errors on cache init
...
The more (logging) the merrier
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-10 07:32:38 +02:00
Michal Mielewczyk
f6bdd354d0
Don't bug on cache init
...
Even if locking the new cache should never fail it's not an unrecoverable
state.
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-08 11:14:20 +02:00
Roel Apfelbaum
1d1561649c
Remove redundant fallback-PT counter accesses
...
The fewer (atomic variable accesses on IO path) the better fare
Signed-off-by: Roel Apfelbaum <roel.apfelbaum@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-08 11:14:20 +02:00
Robert Baldyga
6f02a625ad
pipeline: Introduce debug logging
...
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-08 11:14:20 +02:00
Michal Mielewczyk
40a850cd9c
Remove dead code
...
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-08 11:14:20 +02:00
Michal Mielewczyk
0bb2621c50
Increment ctx.refcnt before creating a new cache
...
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-08 11:14:20 +02:00
Michal Mielewczyk
fae30462b1
Decrement cache.refcnt if locking cache failed
...
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-08 11:14:20 +02:00
Robert Baldyga
6ad1007e6f
Merge pull request #848 from robertbaldyga/maintainer-email
...
Update maintainer's email
2024-10-07 13:30:13 +02:00
Robert Baldyga
89d48a7d39
Update maintainer's email
...
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2024-10-07 09:37:34 +02:00
Robert Baldyga
d166bb7938
Merge pull request #846 from mmichal10/flush-tests
...
Flush tests
2024-10-04 08:50:33 +02:00
Robert Baldyga
b16b49c84d
tests: Introduce cache/core flush test
...
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-04 07:54:46 +02:00
Roel Apfelbaum
b8ea10f30c
pyocf: Fix core.reset_stats()
...
Signed-off-by: Roel Apfelbaum <roel.apfelbaum@huawei.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-04 07:52:14 +02:00
Robert Baldyga
f729887938
Merge pull request #845 from robertbaldyga/fix-ubsan-false-positive
...
Avoid UBSAN false positive
2024-10-02 16:58:42 +02:00
Robert Baldyga
75fb6a59e0
Avoid UBSAN false positive
...
UBSAN: array-index-out-of-bounds in src/ocf_request.c:230:44
index 1 is out of range for type 'ocf_map_info [*]'
Signed-off-by: Robert Baldyga <robert.baldyga@huawei.com>
2024-10-02 16:47:17 +02:00
Robert Baldyga
630748a1e7
Merge pull request #844 from mmichal10/engine-errors-test
...
Engine errors test
2024-10-02 16:26:00 +02:00
Michal Mielewczyk
a681f7ec7e
pyocf: Test for reporting and accounting IO errors
...
Test for bug fixed in 237f6c708a
2024-10-02 16:21:10 +02:00
Michal Mielewczyk
6ec4f6cceb
pyocf: rio: reads and writes accounting
...
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-02 15:55:28 +02:00
Michal Mielewczyk
818dacfe89
pyocf: Disarm error volumes in teardown
...
This prevents obfuscating the actual test errors
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@huawei.com>
2024-10-02 15:52:21 +02:00
Robert Baldyga
575b1abc50
Merge pull request #840 from mmichal10/d2c-dc
...
Test for IO in d2c after attaching cache
2024-10-02 15:50:12 +02:00
