summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* kclient: fix file layout; create in check_capshistoric/async_mdsSage Weil2009-02-242-2/+7
|
* objecter: copy modify bufferlistSage Weil2009-02-231-1/+1
| | | | Caller may not want it's bl to be clobbered.
* kclient: async creates now mostly workingSage Weil2009-02-235-16/+58
|
* kclient: simplify/fix ino prealloc queueSage Weil2009-02-232-55/+31
| | | | We just need to keep two extents at a time.
* mds: flush journal on create if dir filelock is unstableSage Weil2009-02-231-0/+3
| | | | This isn't particularly precise, but works for now.
* mds: change lock state when rdlock_try blocksSage Weil2009-02-232-17/+34
| | | | Previously we would wait but leave the lock state unchanges. Der.
* Makefile: link libcommon, lockdep lastSage Weil2009-02-233-20/+33
| | | | | | | | | | | Link libcommon.a last for everything. Include lockdep last in libcommon so to initialize first, shutdown last, so that g_lockdep is initialized to zero for all Mutexes. Put config.cc next to last so that the atomic_t thread counter is initialized before all threads are created, shutdown after all threads are destroyed.
* mds: fix up createSage Weil2009-02-202-4/+18
|
* mds: journal client prealloc inoSage Weil2009-02-203-4/+11
|
* kclient: initial pass at async create, writebackSage Weil2009-02-2012-47/+183
|
* mds: take rdlock during readdir; no ino alloc on createSage Weil2009-02-202-23/+37
|
* kclient: 'asyncmeta' mount optionSage Weil2009-02-194-5/+18
|
* kclient: do async preallocSage Weil2009-02-195-50/+85
|
* mds: CREATE op for async creation writebackSage Weil2009-02-193-0/+68
|
* kclient: async_createSage Weil2009-02-196-31/+129
| | | | | Need to eventually write these back, and take cap references to ensure that happens before, say, readdir.
* kclient: prealloc ino queueSage Weil2009-02-195-3/+138
|
* mds: masterinotable, client preallocSage Weil2009-02-1917-26/+342
|
* todoSage Weil2009-02-191-3/+0
|
* SafeTimer: getting rid of global timerYehuda Sadeh2009-02-192-11/+12
|
* kclient: only unregister after both acksSage Weil2009-02-192-22/+14
| | | | | | | | Partially reverts behavior of b732c0ddb8a990a97b98606a6d9a387e5a21e3e5 We need the request to stay in the request_tree so that when the safe arrives we can look it up (duh). Also, we need to keep a reference to the request during that period...
* mds: reverse rename argument orderSage Weil2009-02-191-2/+2
| | | | Order change introduced by 3dc43aa8e6606be15695d565715b6a7a1b2d38ea
* client: reverse rename argument orderSage Weil2009-02-191-2/+2
|
* kclient: add missing r_wait list_head initSage Weil2009-02-192-2/+2
|
* kclient: initialize request list_headsSage Weil2009-02-191-1/+2
|
* kclient: cast pathSage Weil2009-02-191-2/+2
| | | | | path[12] will either point to freshly allocated paths (that we later kfree) or a provided const path.
* kclient: unregister mds requests on unsafe replySage Weil2009-02-192-36/+45
| | | | | | | Fresh vs unsafe mds request replay is handled separately. So, unregister requests on first reply. This avoids a stray dentry reference if we unmount before getting a final safe (although eventually those should be explicitly flushed as well).
* kclient: clean up some mdsc lockingSage Weil2009-02-191-19/+4
| | | | Do not drop mds mutex in __do_request().
* kclient: fix path memory leakSage Weil2009-02-191-0/+4
|
* kclient: async mds requestion completion callbacksSage Weil2009-02-192-18/+47
|
* kclient: fix fh_to_dentry; move build_path to mds_client.cSage Weil2009-02-194-111/+121
|
* kclient: mds requests in terms of dentries, not pathsSage Weil2009-02-199-402/+253
| | | | | | Generate mds request paths when request is sent, so that it reflects the current state of mds capabilities. Eventually we can then make the relative paths based on the current capabilities for that mds..
* kclient: async mds requestsSage Weil2009-02-197-267/+294
| | | | | | Restructure the mds client to handle mds request asynchronously, so that we can handle requests that are not managed by a blocking thread.
* kclient: remove extra log outputYehuda Sadeh2009-02-191-1/+0
|
* kclient: sync writeout when not pdflushYehuda Sadeh2009-02-194-13/+3
|
* mds: only flush journal when we need toSage Weil2009-02-192-5/+25
| | | | Force a flush every tick just in case.
* kclient: set 'sync' flag for fsyncSage Weil2009-02-181-0/+3
| | | | | This will ensure that fsynced data is committed immediately by the OSDs.
* script: fix check_osd_request_latencySage Weil2009-02-181-2/+2
|
* script: check_osd_request_latencySage Weil2009-02-181-0/+46
|
* conf: can turn on debug logs in startup.confYehuda Sadeh2009-02-181-1/+4
|
* conf: add debug_ms optionYehuda Sadeh2009-02-171-0/+1
|
* mds: don't rdlock stat fields that client has EXCL cap forSage Weil2009-02-171-5/+19
|
* add ceph_init to distributionyehudasa2009-02-171-2/+2
|
* kclient: backport changes to kernel 2.6.25Yehuda Sadeh2009-02-171-1/+14
|
* kclient: fix mdsc->snap_empty cleanupYehuda Sadeh2009-02-131-1/+1
|
* kclient: remove client from client list before destoryingYehuda Sadeh2009-02-131-0/+4
|
* kclient: fix proc cleanupYehuda Sadeh2009-02-133-4/+13
|
* assert: some more assert_warnYehuda Sadeh2009-02-124-8/+24
|
* cosd: change some assert to assert_warn and clean exitYehuda Sadeh2009-02-123-4/+17
|
* kclient: show some meaningful info in client proc entryYehuda Sadeh2009-02-123-3/+112
|
* kclient: per client proc data fileYehuda Sadeh2009-02-121-4/+29
|
View Lorry Controller Status