Core volume I/O must not be queued on management queue - this would
break I/O accounting code, resulting in use-after-free type of errors
after cache detach, core remove etc.
Signed-off-by: Adam Rutkowski <adam.j.rutkowski@intel.com>
There is a risk that without uuid.data pointer denitialization it will
be freed during original volume deinit, zeroing uuid.data pointer
prevents that.
Signed-off-by: Michal Rakowski <michal.rakowski@intel.com>
This pointer is used to provide optional volume specific data
from the user down to bottom volume open callback. volume_params
is provided to OCF in ocf_mngt_cache_device_config.volume_params.
Signed-off-by: Adam Rutkowski <adam.j.rutkowski@intel.com>
In order to synchronize management operations with I/O OCF
maintains in-flight request counters. For example such ref
counters are used during ocf_mngt_detach to drain requests
accessing cache metadata (cache requests counter) and in
ocf_mngt_flush where we wait for outstanding requests sent
in write back mode (dirty requests counter).
Typically I/O threads increment cache/dirty counter when
creating request and decrement counter on request completion.
Management thread sets atomic variable to signal the start of
management operation. I/O threads react to this by changing
I/O requests mode so that the cache/dirty reference counter
is not incremented. As a result reference counter keeps getting
decremented. Management thread waits for the counter to drop to 0
and proceeds with management operation with assumption that no
cache/dirty requests are in progress.
This patch introduces a handy utility for requests reference
counting logic. ocf_refcnt_inc / dec are used to increment/
decrement counter. ocf_refcnt_freeze() makes subsequent
ocf_refcnt_inc() calls to return false, indicating that counter
cannot be incremented at this moment. ocf_refcnt_register_zero_cb
can be used to asynchronously wait for counter to drop to 0.
Signed-off-by: Adam Rutkowski <adam.j.rutkowski@intel.com>
Cache is used during pipeline_destroy which means that
doing put_cache before destroying pipeline may result in
accessing memory that was freed.
Signed-off-by: Vitaliy Mysak <vitaliy.mysak@intel.com>
For flush/purge entry points to be fully asynchronous we still
need to rework flush mutex and waiting for outstanding dirty
requests.
Signed-off-by: Adam Rutkowski <adam.j.rutkowski@intel.com>
Send cleaner IOs with appropriate queue set
This solves the issue of bottom adapter getting NULL in io->io_queue
Signed-off-by: Vitaliy Mysak <vitaliy.mysak@intel.com>
This allows to reuse same step functions giving them different parameters
on each step.
Additionally move pipeline to utils, to make it accessible to other
subsystems of OCF (e.g. metadata).
Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
NOTE: This is still not the real asynchronism. Metadata interfaces
are still not fully asynchronous.
Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
NOTE: This patch only changes API that pretends to be asynchronous.
Most of management operations are still performed synchronously.
The real asynchronism will be introduced in the next patches.
Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
Instead of calling flush separatly for each IO class, it is called after
collecting number of dirty cache lines defined by user or after iterating
through all IO classes.
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
Unlocking cache and putting queue are perormed in cleaning completion, so all
cleaning policies has to call completion.
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
PyOCF is a tool written with testing OCF functionality in mind.
It is a Python3 (3.6 version required) package which wraps OCF
by providing Python objects in place of OCF objects (volumes, queues,
etc). Thin layer of translation between OCF objects and PyOCF objects
enables using customized behaviors for OCF primitives by subclassing
PyOCF classes.
This initial version implements only WT and WI modes and single,
synchronously operating Queue.
TO DO:
- Queues/Cleaner/MetadataUpdater implemented as Python threads
- Loading of caches from PyOCF Volumes (fix bugs in OCF)
- Make sure it works multi-threaded for more sophisticated tests
Co-authored-by: Jan Musial <jan.musial@intel.com>
Signed-off-by: Michal Rakowski <michal.rakowski@intel.com>
Signed-off-by: Jan Musial <jan.musial@intel.com>
- Add cache trylock and read trylock functions.
- Introduce new error code -OCF_ERR_NO_LOCK.
- Change trylock functions in env to return this code in case of
lock contention.
[ENV CHANGES REQUIRED]
Following functions should return 0 on success or -OCF_ERR_NO_LOCK
in case of lock contention:
- env_mutex_trylock()
- env_rwsem_up_read_trylock()
- env_rwsem_up_write_trylock()
Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
- Queue allocation is now separated from starting cache.
- Queue can be created and destroyed in runtime.
- All queue ops accept queue handle instead of queue id.
- Cache stores queues as list instead of array.
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
Signed-off-by: Robert Baldyga <robert.baldyga@intel.com>
Instead of switching write policy to pass-through, barrier is rised
by incrementing counter in ocf_cache_t structure.
Signed-off-by: Michal Mielewczyk <michal.mielewczyk@intel.com>
