2009-04-03 22:42:36 +07:00
|
|
|
/* General filesystem caching backing cache interface
|
|
|
|
*
|
|
|
|
* Copyright (C) 2004-2007 Red Hat, Inc. All Rights Reserved.
|
|
|
|
* Written by David Howells (dhowells@redhat.com)
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU General Public License
|
|
|
|
* as published by the Free Software Foundation; either version
|
|
|
|
* 2 of the License, or (at your option) any later version.
|
|
|
|
*
|
|
|
|
* NOTE!!! See:
|
|
|
|
*
|
|
|
|
* Documentation/filesystems/caching/backend-api.txt
|
|
|
|
*
|
|
|
|
* for a description of the cache backend interface declared here.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifndef _LINUX_FSCACHE_CACHE_H
|
|
|
|
#define _LINUX_FSCACHE_CACHE_H
|
|
|
|
|
|
|
|
#include <linux/fscache.h>
|
|
|
|
#include <linux/sched.h>
|
2010-07-21 03:09:01 +07:00
|
|
|
#include <linux/workqueue.h>
|
2009-04-03 22:42:36 +07:00
|
|
|
|
|
|
|
#define NR_MAXCACHES BITS_PER_LONG
|
|
|
|
|
|
|
|
struct fscache_cache;
|
|
|
|
struct fscache_cache_ops;
|
|
|
|
struct fscache_object;
|
|
|
|
struct fscache_operation;
|
|
|
|
|
2018-04-04 19:41:27 +07:00
|
|
|
enum fscache_obj_ref_trace {
|
|
|
|
fscache_obj_get_add_to_deps,
|
|
|
|
fscache_obj_get_queue,
|
|
|
|
fscache_obj_put_alloc_fail,
|
|
|
|
fscache_obj_put_attach_fail,
|
|
|
|
fscache_obj_put_drop_obj,
|
|
|
|
fscache_obj_put_enq_dep,
|
|
|
|
fscache_obj_put_queue,
|
|
|
|
fscache_obj_put_work,
|
|
|
|
fscache_obj_ref__nr_traces
|
|
|
|
};
|
|
|
|
|
2009-04-03 22:42:36 +07:00
|
|
|
/*
|
|
|
|
* cache tag definition
|
|
|
|
*/
|
|
|
|
struct fscache_cache_tag {
|
|
|
|
struct list_head link;
|
|
|
|
struct fscache_cache *cache; /* cache referred to by this tag */
|
|
|
|
unsigned long flags;
|
|
|
|
#define FSCACHE_TAG_RESERVED 0 /* T if tag is reserved for a cache */
|
|
|
|
atomic_t usage;
|
|
|
|
char name[0]; /* tag name */
|
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* cache definition
|
|
|
|
*/
|
|
|
|
struct fscache_cache {
|
|
|
|
const struct fscache_cache_ops *ops;
|
|
|
|
struct fscache_cache_tag *tag; /* tag representing this cache */
|
|
|
|
struct kobject *kobj; /* system representation of this cache */
|
|
|
|
struct list_head link; /* link in list of caches */
|
|
|
|
size_t max_index_size; /* maximum size of index data */
|
|
|
|
char identifier[36]; /* cache label */
|
|
|
|
|
|
|
|
/* node management */
|
|
|
|
struct work_struct op_gc; /* operation garbage collector */
|
|
|
|
struct list_head object_list; /* list of data/index objects */
|
|
|
|
struct list_head op_gc_list; /* list of ops to be deleted */
|
|
|
|
spinlock_t object_list_lock;
|
|
|
|
spinlock_t op_gc_list_lock;
|
|
|
|
atomic_t object_count; /* no. of live objects in this cache */
|
|
|
|
struct fscache_object *fsdef; /* object for the fsdef index */
|
|
|
|
unsigned long flags;
|
|
|
|
#define FSCACHE_IOERROR 0 /* cache stopped on I/O error */
|
|
|
|
#define FSCACHE_CACHE_WITHDRAWN 1 /* cache has been withdrawn */
|
|
|
|
};
|
|
|
|
|
|
|
|
extern wait_queue_head_t fscache_cache_cleared_wq;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* operation to be applied to a cache object
|
|
|
|
* - retrieval initiation operations are done in the context of the process
|
|
|
|
* that issued them, and not in an async thread pool
|
|
|
|
*/
|
|
|
|
typedef void (*fscache_operation_release_t)(struct fscache_operation *op);
|
|
|
|
typedef void (*fscache_operation_processor_t)(struct fscache_operation *op);
|
2015-02-24 17:05:29 +07:00
|
|
|
typedef void (*fscache_operation_cancel_t)(struct fscache_operation *op);
|
2009-04-03 22:42:36 +07:00
|
|
|
|
FS-Cache: Fix operation state management and accounting
Fix the state management of internal fscache operations and the accounting of
what operations are in what states.
This is done by:
(1) Give struct fscache_operation a enum variable that directly represents the
state it's currently in, rather than spreading this knowledge over a bunch
of flags, who's processing the operation at the moment and whether it is
queued or not.
This makes it easier to write assertions to check the state at various
points and to prevent invalid state transitions.
(2) Add an 'operation complete' state and supply a function to indicate the
completion of an operation (fscache_op_complete()) and make things call
it. The final call to fscache_put_operation() can then check that an op
in the appropriate state (complete or cancelled).
(3) Adjust the use of object->n_ops, ->n_in_progress, ->n_exclusive to better
govern the state of an object:
(a) The ->n_ops is now the number of extant operations on the object
and is now decremented by fscache_put_operation() only.
(b) The ->n_in_progress is simply the number of objects that have been
taken off of the object's pending queue for the purposes of being
run. This is decremented by fscache_op_complete() only.
(c) The ->n_exclusive is the number of exclusive ops that have been
submitted and queued or are in progress. It is decremented by
fscache_op_complete() and by fscache_cancel_op().
fscache_put_operation() and fscache_operation_gc() now no longer try to
clean up ->n_exclusive and ->n_in_progress. That was leading to double
decrements against fscache_cancel_op().
fscache_cancel_op() now no longer decrements ->n_ops. That was leading to
double decrements against fscache_put_operation().
fscache_submit_exclusive_op() now decides whether it has to queue an op
based on ->n_in_progress being > 0 rather than ->n_ops > 0 as the latter
will persist in being true even after all preceding operations have been
cancelled or completed. Furthermore, if an object is active and there are
runnable ops against it, there must be at least one op running.
(4) Add a remaining-pages counter (n_pages) to struct fscache_retrieval and
provide a function to record completion of the pages as they complete.
When n_pages reaches 0, the operation is deemed to be complete and
fscache_op_complete() is called.
Add calls to fscache_retrieval_complete() anywhere we've finished with a
page we've been given to read or allocate for. This includes places where
we just return pages to the netfs for reading from the server and where
accessing the cache fails and we discard the proposed netfs page.
The bugs in the unfixed state management manifest themselves as oopses like the
following where the operation completion gets out of sync with return of the
cookie by the netfs. This is possible because the cache unlocks and returns
all the netfs pages before recording its completion - which means that there's
nothing to stop the netfs discarding them and returning the cookie.
FS-Cache: Cookie 'NFS.fh' still has outstanding reads
------------[ cut here ]------------
kernel BUG at fs/fscache/cookie.c:519!
invalid opcode: 0000 [#1] SMP
CPU 1
Modules linked in: cachefiles nfs fscache auth_rpcgss nfs_acl lockd sunrpc
Pid: 400, comm: kswapd0 Not tainted 3.1.0-rc7-fsdevel+ #1090 /DG965RY
RIP: 0010:[<ffffffffa007050a>] [<ffffffffa007050a>] __fscache_relinquish_cookie+0x170/0x343 [fscache]
RSP: 0018:ffff8800368cfb00 EFLAGS: 00010282
RAX: 000000000000003c RBX: ffff880023cc8790 RCX: 0000000000000000
RDX: 0000000000002f2e RSI: 0000000000000001 RDI: ffffffff813ab86c
RBP: ffff8800368cfb50 R08: 0000000000000002 R09: 0000000000000000
R10: ffff88003a1b7890 R11: ffff88001df6e488 R12: ffff880023d8ed98
R13: ffff880023cc8798 R14: 0000000000000004 R15: ffff88003b8bf370
FS: 0000000000000000(0000) GS:ffff88003bd00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00000000008ba008 CR3: 0000000023d93000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process kswapd0 (pid: 400, threadinfo ffff8800368ce000, task ffff88003b8bf040)
Stack:
ffff88003b8bf040 ffff88001df6e528 ffff88001df6e528 ffffffffa00b46b0
ffff88003b8bf040 ffff88001df6e488 ffff88001df6e620 ffffffffa00b46b0
ffff88001ebd04c8 0000000000000004 ffff8800368cfb70 ffffffffa00b2c91
Call Trace:
[<ffffffffa00b2c91>] nfs_fscache_release_inode_cookie+0x3b/0x47 [nfs]
[<ffffffffa008f25f>] nfs_clear_inode+0x3c/0x41 [nfs]
[<ffffffffa0090df1>] nfs4_evict_inode+0x2f/0x33 [nfs]
[<ffffffff810d8d47>] evict+0xa1/0x15c
[<ffffffff810d8e2e>] dispose_list+0x2c/0x38
[<ffffffff810d9ebd>] prune_icache_sb+0x28c/0x29b
[<ffffffff810c56b7>] prune_super+0xd5/0x140
[<ffffffff8109b615>] shrink_slab+0x102/0x1ab
[<ffffffff8109d690>] balance_pgdat+0x2f2/0x595
[<ffffffff8103e009>] ? process_timeout+0xb/0xb
[<ffffffff8109dba3>] kswapd+0x270/0x289
[<ffffffff8104c5ea>] ? __init_waitqueue_head+0x46/0x46
[<ffffffff8109d933>] ? balance_pgdat+0x595/0x595
[<ffffffff8104bf7a>] kthread+0x7f/0x87
[<ffffffff813ad6b4>] kernel_thread_helper+0x4/0x10
[<ffffffff81026b98>] ? finish_task_switch+0x45/0xc0
[<ffffffff813abcdd>] ? retint_restore_args+0xe/0xe
[<ffffffff8104befb>] ? __init_kthread_worker+0x53/0x53
[<ffffffff813ad6b0>] ? gs_change+0xb/0xb
Signed-off-by: David Howells <dhowells@redhat.com>
2012-12-21 04:52:35 +07:00
|
|
|
enum fscache_operation_state {
|
|
|
|
FSCACHE_OP_ST_BLANK, /* Op is not yet submitted */
|
|
|
|
FSCACHE_OP_ST_INITIALISED, /* Op is initialised */
|
|
|
|
FSCACHE_OP_ST_PENDING, /* Op is blocked from running */
|
|
|
|
FSCACHE_OP_ST_IN_PROGRESS, /* Op is in progress */
|
|
|
|
FSCACHE_OP_ST_COMPLETE, /* Op is complete */
|
|
|
|
FSCACHE_OP_ST_CANCELLED, /* Op has been cancelled */
|
|
|
|
FSCACHE_OP_ST_DEAD /* Op is now dead */
|
|
|
|
};
|
|
|
|
|
2009-04-03 22:42:36 +07:00
|
|
|
struct fscache_operation {
|
2010-07-21 03:09:01 +07:00
|
|
|
struct work_struct work; /* record for async ops */
|
2009-04-03 22:42:36 +07:00
|
|
|
struct list_head pend_link; /* link in object->pending_ops */
|
|
|
|
struct fscache_object *object; /* object to be operated upon */
|
|
|
|
|
|
|
|
unsigned long flags;
|
|
|
|
#define FSCACHE_OP_TYPE 0x000f /* operation type */
|
2010-07-21 03:09:01 +07:00
|
|
|
#define FSCACHE_OP_ASYNC 0x0001 /* - async op, processor may sleep for disk */
|
|
|
|
#define FSCACHE_OP_MYTHREAD 0x0002 /* - processing is done be issuing thread, not pool */
|
2009-04-03 22:42:36 +07:00
|
|
|
#define FSCACHE_OP_WAITING 4 /* cleared when op is woken */
|
|
|
|
#define FSCACHE_OP_EXCLUSIVE 5 /* exclusive op, other ops must wait */
|
FS-Cache: Fix operation state management and accounting
Fix the state management of internal fscache operations and the accounting of
what operations are in what states.
This is done by:
(1) Give struct fscache_operation a enum variable that directly represents the
state it's currently in, rather than spreading this knowledge over a bunch
of flags, who's processing the operation at the moment and whether it is
queued or not.
This makes it easier to write assertions to check the state at various
points and to prevent invalid state transitions.
(2) Add an 'operation complete' state and supply a function to indicate the
completion of an operation (fscache_op_complete()) and make things call
it. The final call to fscache_put_operation() can then check that an op
in the appropriate state (complete or cancelled).
(3) Adjust the use of object->n_ops, ->n_in_progress, ->n_exclusive to better
govern the state of an object:
(a) The ->n_ops is now the number of extant operations on the object
and is now decremented by fscache_put_operation() only.
(b) The ->n_in_progress is simply the number of objects that have been
taken off of the object's pending queue for the purposes of being
run. This is decremented by fscache_op_complete() only.
(c) The ->n_exclusive is the number of exclusive ops that have been
submitted and queued or are in progress. It is decremented by
fscache_op_complete() and by fscache_cancel_op().
fscache_put_operation() and fscache_operation_gc() now no longer try to
clean up ->n_exclusive and ->n_in_progress. That was leading to double
decrements against fscache_cancel_op().
fscache_cancel_op() now no longer decrements ->n_ops. That was leading to
double decrements against fscache_put_operation().
fscache_submit_exclusive_op() now decides whether it has to queue an op
based on ->n_in_progress being > 0 rather than ->n_ops > 0 as the latter
will persist in being true even after all preceding operations have been
cancelled or completed. Furthermore, if an object is active and there are
runnable ops against it, there must be at least one op running.
(4) Add a remaining-pages counter (n_pages) to struct fscache_retrieval and
provide a function to record completion of the pages as they complete.
When n_pages reaches 0, the operation is deemed to be complete and
fscache_op_complete() is called.
Add calls to fscache_retrieval_complete() anywhere we've finished with a
page we've been given to read or allocate for. This includes places where
we just return pages to the netfs for reading from the server and where
accessing the cache fails and we discard the proposed netfs page.
The bugs in the unfixed state management manifest themselves as oopses like the
following where the operation completion gets out of sync with return of the
cookie by the netfs. This is possible because the cache unlocks and returns
all the netfs pages before recording its completion - which means that there's
nothing to stop the netfs discarding them and returning the cookie.
FS-Cache: Cookie 'NFS.fh' still has outstanding reads
------------[ cut here ]------------
kernel BUG at fs/fscache/cookie.c:519!
invalid opcode: 0000 [#1] SMP
CPU 1
Modules linked in: cachefiles nfs fscache auth_rpcgss nfs_acl lockd sunrpc
Pid: 400, comm: kswapd0 Not tainted 3.1.0-rc7-fsdevel+ #1090 /DG965RY
RIP: 0010:[<ffffffffa007050a>] [<ffffffffa007050a>] __fscache_relinquish_cookie+0x170/0x343 [fscache]
RSP: 0018:ffff8800368cfb00 EFLAGS: 00010282
RAX: 000000000000003c RBX: ffff880023cc8790 RCX: 0000000000000000
RDX: 0000000000002f2e RSI: 0000000000000001 RDI: ffffffff813ab86c
RBP: ffff8800368cfb50 R08: 0000000000000002 R09: 0000000000000000
R10: ffff88003a1b7890 R11: ffff88001df6e488 R12: ffff880023d8ed98
R13: ffff880023cc8798 R14: 0000000000000004 R15: ffff88003b8bf370
FS: 0000000000000000(0000) GS:ffff88003bd00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00000000008ba008 CR3: 0000000023d93000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process kswapd0 (pid: 400, threadinfo ffff8800368ce000, task ffff88003b8bf040)
Stack:
ffff88003b8bf040 ffff88001df6e528 ffff88001df6e528 ffffffffa00b46b0
ffff88003b8bf040 ffff88001df6e488 ffff88001df6e620 ffffffffa00b46b0
ffff88001ebd04c8 0000000000000004 ffff8800368cfb70 ffffffffa00b2c91
Call Trace:
[<ffffffffa00b2c91>] nfs_fscache_release_inode_cookie+0x3b/0x47 [nfs]
[<ffffffffa008f25f>] nfs_clear_inode+0x3c/0x41 [nfs]
[<ffffffffa0090df1>] nfs4_evict_inode+0x2f/0x33 [nfs]
[<ffffffff810d8d47>] evict+0xa1/0x15c
[<ffffffff810d8e2e>] dispose_list+0x2c/0x38
[<ffffffff810d9ebd>] prune_icache_sb+0x28c/0x29b
[<ffffffff810c56b7>] prune_super+0xd5/0x140
[<ffffffff8109b615>] shrink_slab+0x102/0x1ab
[<ffffffff8109d690>] balance_pgdat+0x2f2/0x595
[<ffffffff8103e009>] ? process_timeout+0xb/0xb
[<ffffffff8109dba3>] kswapd+0x270/0x289
[<ffffffff8104c5ea>] ? __init_waitqueue_head+0x46/0x46
[<ffffffff8109d933>] ? balance_pgdat+0x595/0x595
[<ffffffff8104bf7a>] kthread+0x7f/0x87
[<ffffffff813ad6b4>] kernel_thread_helper+0x4/0x10
[<ffffffff81026b98>] ? finish_task_switch+0x45/0xc0
[<ffffffff813abcdd>] ? retint_restore_args+0xe/0xe
[<ffffffff8104befb>] ? __init_kthread_worker+0x53/0x53
[<ffffffff813ad6b0>] ? gs_change+0xb/0xb
Signed-off-by: David Howells <dhowells@redhat.com>
2012-12-21 04:52:35 +07:00
|
|
|
#define FSCACHE_OP_DEC_READ_CNT 6 /* decrement object->n_reads on destruction */
|
FS-Cache: Simplify cookie retention for fscache_objects, fixing oops
Simplify the way fscache cache objects retain their cookie. The way I
implemented the cookie storage handling made synchronisation a pain (ie. the
object state machine can't rely on the cookie actually still being there).
Instead of the the object being detached from the cookie and the cookie being
freed in __fscache_relinquish_cookie(), we defer both operations:
(*) The detachment of the object from the list in the cookie now takes place
in fscache_drop_object() and is thus governed by the object state machine
(fscache_detach_from_cookie() has been removed).
(*) The release of the cookie is now in fscache_object_destroy() - which is
called by the cache backend just before it frees the object.
This means that the fscache_cookie struct is now available to the cache all the
way through from ->alloc_object() to ->drop_object() and ->put_object() -
meaning that it's no longer necessary to take object->lock to guarantee access.
However, __fscache_relinquish_cookie() doesn't wait for the object to go all
the way through to destruction before letting the netfs proceed. That would
massively slow down the netfs. Since __fscache_relinquish_cookie() leaves the
cookie around, in must therefore break all attachments to the netfs - which
includes ->def, ->netfs_data and any outstanding page read/writes.
To handle this, struct fscache_cookie now has an n_active counter:
(1) This starts off initialised to 1.
(2) Any time the cache needs to get at the netfs data, it calls
fscache_use_cookie() to increment it - if it is not zero. If it was zero,
then access is not permitted.
(3) When the cache has finished with the data, it calls fscache_unuse_cookie()
to decrement it. This does a wake-up on it if it reaches 0.
(4) __fscache_relinquish_cookie() decrements n_active and then waits for it to
reach 0. The initialisation to 1 in step (1) ensures that we only get
wake ups when we're trying to get rid of the cookie.
This leaves __fscache_relinquish_cookie() a lot simpler.
***
This fixes a problem in the current code whereby if fscache_invalidate() is
followed sufficiently quickly by fscache_relinquish_cookie() then it is
possible for __fscache_relinquish_cookie() to have detached the cookie from the
object and cleared the pointer before a thread is dispatched to process the
invalidation state in the object state machine.
Since the pending write clearance was deferred to the invalidation state to
make it asynchronous, we need to either wait in relinquishment for the stores
tree to be cleared in the invalidation state or we need to handle the clearance
in relinquishment.
Further, if the relinquishment code does clear the tree, then the invalidation
state need to make the clearance contingent on still having the cookie to hand
(since that's where the tree is rooted) and we have to prevent the cookie from
disappearing for the duration.
This can lead to an oops like the following:
BUG: unable to handle kernel NULL pointer dereference at 000000000000000c
...
RIP: 0010:[<ffffffff8151023e>] _spin_lock+0xe/0x30
...
CR2: 000000000000000c ...
...
Process kslowd002 (...)
....
Call Trace:
[<ffffffffa01c3278>] fscache_invalidate_writes+0x38/0xd0 [fscache]
[<ffffffff810096f0>] ? __switch_to+0xd0/0x320
[<ffffffff8105e759>] ? find_busiest_queue+0x69/0x150
[<ffffffff8110ddd4>] ? slow_work_enqueue+0x104/0x180
[<ffffffffa01c1303>] fscache_object_slow_work_execute+0x5e3/0x9d0 [fscache]
[<ffffffff81096b67>] ? bit_waitqueue+0x17/0xd0
[<ffffffff8110e233>] slow_work_execute+0x233/0x310
[<ffffffff8110e515>] slow_work_thread+0x205/0x360
[<ffffffff81096ca0>] ? autoremove_wake_function+0x0/0x40
[<ffffffff8110e310>] ? slow_work_thread+0x0/0x360
[<ffffffff81096936>] kthread+0x96/0xa0
[<ffffffff8100c0ca>] child_rip+0xa/0x20
[<ffffffff810968a0>] ? kthread+0x0/0xa0
[<ffffffff8100c0c0>] ? child_rip+0x0/0x20
The parameter to fscache_invalidate_writes() was object->cookie which is NULL.
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
#define FSCACHE_OP_UNUSE_COOKIE 7 /* call fscache_unuse_cookie() on completion */
|
|
|
|
#define FSCACHE_OP_KEEP_FLAGS 0x00f0 /* flags to keep when repurposing an op */
|
2009-04-03 22:42:36 +07:00
|
|
|
|
FS-Cache: Fix operation state management and accounting
Fix the state management of internal fscache operations and the accounting of
what operations are in what states.
This is done by:
(1) Give struct fscache_operation a enum variable that directly represents the
state it's currently in, rather than spreading this knowledge over a bunch
of flags, who's processing the operation at the moment and whether it is
queued or not.
This makes it easier to write assertions to check the state at various
points and to prevent invalid state transitions.
(2) Add an 'operation complete' state and supply a function to indicate the
completion of an operation (fscache_op_complete()) and make things call
it. The final call to fscache_put_operation() can then check that an op
in the appropriate state (complete or cancelled).
(3) Adjust the use of object->n_ops, ->n_in_progress, ->n_exclusive to better
govern the state of an object:
(a) The ->n_ops is now the number of extant operations on the object
and is now decremented by fscache_put_operation() only.
(b) The ->n_in_progress is simply the number of objects that have been
taken off of the object's pending queue for the purposes of being
run. This is decremented by fscache_op_complete() only.
(c) The ->n_exclusive is the number of exclusive ops that have been
submitted and queued or are in progress. It is decremented by
fscache_op_complete() and by fscache_cancel_op().
fscache_put_operation() and fscache_operation_gc() now no longer try to
clean up ->n_exclusive and ->n_in_progress. That was leading to double
decrements against fscache_cancel_op().
fscache_cancel_op() now no longer decrements ->n_ops. That was leading to
double decrements against fscache_put_operation().
fscache_submit_exclusive_op() now decides whether it has to queue an op
based on ->n_in_progress being > 0 rather than ->n_ops > 0 as the latter
will persist in being true even after all preceding operations have been
cancelled or completed. Furthermore, if an object is active and there are
runnable ops against it, there must be at least one op running.
(4) Add a remaining-pages counter (n_pages) to struct fscache_retrieval and
provide a function to record completion of the pages as they complete.
When n_pages reaches 0, the operation is deemed to be complete and
fscache_op_complete() is called.
Add calls to fscache_retrieval_complete() anywhere we've finished with a
page we've been given to read or allocate for. This includes places where
we just return pages to the netfs for reading from the server and where
accessing the cache fails and we discard the proposed netfs page.
The bugs in the unfixed state management manifest themselves as oopses like the
following where the operation completion gets out of sync with return of the
cookie by the netfs. This is possible because the cache unlocks and returns
all the netfs pages before recording its completion - which means that there's
nothing to stop the netfs discarding them and returning the cookie.
FS-Cache: Cookie 'NFS.fh' still has outstanding reads
------------[ cut here ]------------
kernel BUG at fs/fscache/cookie.c:519!
invalid opcode: 0000 [#1] SMP
CPU 1
Modules linked in: cachefiles nfs fscache auth_rpcgss nfs_acl lockd sunrpc
Pid: 400, comm: kswapd0 Not tainted 3.1.0-rc7-fsdevel+ #1090 /DG965RY
RIP: 0010:[<ffffffffa007050a>] [<ffffffffa007050a>] __fscache_relinquish_cookie+0x170/0x343 [fscache]
RSP: 0018:ffff8800368cfb00 EFLAGS: 00010282
RAX: 000000000000003c RBX: ffff880023cc8790 RCX: 0000000000000000
RDX: 0000000000002f2e RSI: 0000000000000001 RDI: ffffffff813ab86c
RBP: ffff8800368cfb50 R08: 0000000000000002 R09: 0000000000000000
R10: ffff88003a1b7890 R11: ffff88001df6e488 R12: ffff880023d8ed98
R13: ffff880023cc8798 R14: 0000000000000004 R15: ffff88003b8bf370
FS: 0000000000000000(0000) GS:ffff88003bd00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00000000008ba008 CR3: 0000000023d93000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process kswapd0 (pid: 400, threadinfo ffff8800368ce000, task ffff88003b8bf040)
Stack:
ffff88003b8bf040 ffff88001df6e528 ffff88001df6e528 ffffffffa00b46b0
ffff88003b8bf040 ffff88001df6e488 ffff88001df6e620 ffffffffa00b46b0
ffff88001ebd04c8 0000000000000004 ffff8800368cfb70 ffffffffa00b2c91
Call Trace:
[<ffffffffa00b2c91>] nfs_fscache_release_inode_cookie+0x3b/0x47 [nfs]
[<ffffffffa008f25f>] nfs_clear_inode+0x3c/0x41 [nfs]
[<ffffffffa0090df1>] nfs4_evict_inode+0x2f/0x33 [nfs]
[<ffffffff810d8d47>] evict+0xa1/0x15c
[<ffffffff810d8e2e>] dispose_list+0x2c/0x38
[<ffffffff810d9ebd>] prune_icache_sb+0x28c/0x29b
[<ffffffff810c56b7>] prune_super+0xd5/0x140
[<ffffffff8109b615>] shrink_slab+0x102/0x1ab
[<ffffffff8109d690>] balance_pgdat+0x2f2/0x595
[<ffffffff8103e009>] ? process_timeout+0xb/0xb
[<ffffffff8109dba3>] kswapd+0x270/0x289
[<ffffffff8104c5ea>] ? __init_waitqueue_head+0x46/0x46
[<ffffffff8109d933>] ? balance_pgdat+0x595/0x595
[<ffffffff8104bf7a>] kthread+0x7f/0x87
[<ffffffff813ad6b4>] kernel_thread_helper+0x4/0x10
[<ffffffff81026b98>] ? finish_task_switch+0x45/0xc0
[<ffffffff813abcdd>] ? retint_restore_args+0xe/0xe
[<ffffffff8104befb>] ? __init_kthread_worker+0x53/0x53
[<ffffffff813ad6b0>] ? gs_change+0xb/0xb
Signed-off-by: David Howells <dhowells@redhat.com>
2012-12-21 04:52:35 +07:00
|
|
|
enum fscache_operation_state state;
|
2009-04-03 22:42:36 +07:00
|
|
|
atomic_t usage;
|
|
|
|
unsigned debug_id; /* debugging ID */
|
|
|
|
|
|
|
|
/* operation processor callback
|
|
|
|
* - can be NULL if FSCACHE_OP_WAITING is going to be used to perform
|
|
|
|
* the op in a non-pool thread */
|
|
|
|
fscache_operation_processor_t processor;
|
|
|
|
|
2015-02-24 17:05:29 +07:00
|
|
|
/* Operation cancellation cleanup (optional) */
|
|
|
|
fscache_operation_cancel_t cancel;
|
|
|
|
|
2009-04-03 22:42:36 +07:00
|
|
|
/* operation releaser */
|
|
|
|
fscache_operation_release_t release;
|
|
|
|
};
|
|
|
|
|
|
|
|
extern atomic_t fscache_op_debug_id;
|
2010-07-21 03:09:01 +07:00
|
|
|
extern void fscache_op_work_func(struct work_struct *work);
|
2009-04-03 22:42:36 +07:00
|
|
|
|
|
|
|
extern void fscache_enqueue_operation(struct fscache_operation *);
|
2012-12-14 03:03:13 +07:00
|
|
|
extern void fscache_op_complete(struct fscache_operation *, bool);
|
2009-04-03 22:42:36 +07:00
|
|
|
extern void fscache_put_operation(struct fscache_operation *);
|
2018-04-04 19:41:27 +07:00
|
|
|
extern void fscache_operation_init(struct fscache_cookie *,
|
|
|
|
struct fscache_operation *,
|
2015-02-25 20:26:39 +07:00
|
|
|
fscache_operation_processor_t,
|
2015-02-24 17:05:29 +07:00
|
|
|
fscache_operation_cancel_t,
|
2015-02-25 20:26:39 +07:00
|
|
|
fscache_operation_release_t);
|
2009-04-03 22:42:36 +07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* data read operation
|
|
|
|
*/
|
|
|
|
struct fscache_retrieval {
|
|
|
|
struct fscache_operation op;
|
2015-02-24 17:05:29 +07:00
|
|
|
struct fscache_cookie *cookie; /* The netfs cookie */
|
2009-04-03 22:42:36 +07:00
|
|
|
struct address_space *mapping; /* netfs pages */
|
|
|
|
fscache_rw_complete_t end_io_func; /* function to call on I/O completion */
|
|
|
|
void *context; /* netfs read context (pinned) */
|
|
|
|
struct list_head to_do; /* list of things to be done by the backend */
|
|
|
|
unsigned long start_time; /* time at which retrieval started */
|
2013-05-21 19:44:15 +07:00
|
|
|
atomic_t n_pages; /* number of pages to be retrieved */
|
2009-04-03 22:42:36 +07:00
|
|
|
};
|
|
|
|
|
|
|
|
typedef int (*fscache_page_retrieval_func_t)(struct fscache_retrieval *op,
|
|
|
|
struct page *page,
|
|
|
|
gfp_t gfp);
|
|
|
|
|
|
|
|
typedef int (*fscache_pages_retrieval_func_t)(struct fscache_retrieval *op,
|
|
|
|
struct list_head *pages,
|
|
|
|
unsigned *nr_pages,
|
|
|
|
gfp_t gfp);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* fscache_get_retrieval - Get an extra reference on a retrieval operation
|
|
|
|
* @op: The retrieval operation to get a reference on
|
|
|
|
*
|
|
|
|
* Get an extra reference on a retrieval operation.
|
|
|
|
*/
|
|
|
|
static inline
|
|
|
|
struct fscache_retrieval *fscache_get_retrieval(struct fscache_retrieval *op)
|
|
|
|
{
|
|
|
|
atomic_inc(&op->op.usage);
|
|
|
|
return op;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* fscache_enqueue_retrieval - Enqueue a retrieval operation for processing
|
|
|
|
* @op: The retrieval operation affected
|
|
|
|
*
|
|
|
|
* Enqueue a retrieval operation for processing by the FS-Cache thread pool.
|
|
|
|
*/
|
|
|
|
static inline void fscache_enqueue_retrieval(struct fscache_retrieval *op)
|
|
|
|
{
|
|
|
|
fscache_enqueue_operation(&op->op);
|
|
|
|
}
|
|
|
|
|
FS-Cache: Fix operation state management and accounting
Fix the state management of internal fscache operations and the accounting of
what operations are in what states.
This is done by:
(1) Give struct fscache_operation a enum variable that directly represents the
state it's currently in, rather than spreading this knowledge over a bunch
of flags, who's processing the operation at the moment and whether it is
queued or not.
This makes it easier to write assertions to check the state at various
points and to prevent invalid state transitions.
(2) Add an 'operation complete' state and supply a function to indicate the
completion of an operation (fscache_op_complete()) and make things call
it. The final call to fscache_put_operation() can then check that an op
in the appropriate state (complete or cancelled).
(3) Adjust the use of object->n_ops, ->n_in_progress, ->n_exclusive to better
govern the state of an object:
(a) The ->n_ops is now the number of extant operations on the object
and is now decremented by fscache_put_operation() only.
(b) The ->n_in_progress is simply the number of objects that have been
taken off of the object's pending queue for the purposes of being
run. This is decremented by fscache_op_complete() only.
(c) The ->n_exclusive is the number of exclusive ops that have been
submitted and queued or are in progress. It is decremented by
fscache_op_complete() and by fscache_cancel_op().
fscache_put_operation() and fscache_operation_gc() now no longer try to
clean up ->n_exclusive and ->n_in_progress. That was leading to double
decrements against fscache_cancel_op().
fscache_cancel_op() now no longer decrements ->n_ops. That was leading to
double decrements against fscache_put_operation().
fscache_submit_exclusive_op() now decides whether it has to queue an op
based on ->n_in_progress being > 0 rather than ->n_ops > 0 as the latter
will persist in being true even after all preceding operations have been
cancelled or completed. Furthermore, if an object is active and there are
runnable ops against it, there must be at least one op running.
(4) Add a remaining-pages counter (n_pages) to struct fscache_retrieval and
provide a function to record completion of the pages as they complete.
When n_pages reaches 0, the operation is deemed to be complete and
fscache_op_complete() is called.
Add calls to fscache_retrieval_complete() anywhere we've finished with a
page we've been given to read or allocate for. This includes places where
we just return pages to the netfs for reading from the server and where
accessing the cache fails and we discard the proposed netfs page.
The bugs in the unfixed state management manifest themselves as oopses like the
following where the operation completion gets out of sync with return of the
cookie by the netfs. This is possible because the cache unlocks and returns
all the netfs pages before recording its completion - which means that there's
nothing to stop the netfs discarding them and returning the cookie.
FS-Cache: Cookie 'NFS.fh' still has outstanding reads
------------[ cut here ]------------
kernel BUG at fs/fscache/cookie.c:519!
invalid opcode: 0000 [#1] SMP
CPU 1
Modules linked in: cachefiles nfs fscache auth_rpcgss nfs_acl lockd sunrpc
Pid: 400, comm: kswapd0 Not tainted 3.1.0-rc7-fsdevel+ #1090 /DG965RY
RIP: 0010:[<ffffffffa007050a>] [<ffffffffa007050a>] __fscache_relinquish_cookie+0x170/0x343 [fscache]
RSP: 0018:ffff8800368cfb00 EFLAGS: 00010282
RAX: 000000000000003c RBX: ffff880023cc8790 RCX: 0000000000000000
RDX: 0000000000002f2e RSI: 0000000000000001 RDI: ffffffff813ab86c
RBP: ffff8800368cfb50 R08: 0000000000000002 R09: 0000000000000000
R10: ffff88003a1b7890 R11: ffff88001df6e488 R12: ffff880023d8ed98
R13: ffff880023cc8798 R14: 0000000000000004 R15: ffff88003b8bf370
FS: 0000000000000000(0000) GS:ffff88003bd00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00000000008ba008 CR3: 0000000023d93000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process kswapd0 (pid: 400, threadinfo ffff8800368ce000, task ffff88003b8bf040)
Stack:
ffff88003b8bf040 ffff88001df6e528 ffff88001df6e528 ffffffffa00b46b0
ffff88003b8bf040 ffff88001df6e488 ffff88001df6e620 ffffffffa00b46b0
ffff88001ebd04c8 0000000000000004 ffff8800368cfb70 ffffffffa00b2c91
Call Trace:
[<ffffffffa00b2c91>] nfs_fscache_release_inode_cookie+0x3b/0x47 [nfs]
[<ffffffffa008f25f>] nfs_clear_inode+0x3c/0x41 [nfs]
[<ffffffffa0090df1>] nfs4_evict_inode+0x2f/0x33 [nfs]
[<ffffffff810d8d47>] evict+0xa1/0x15c
[<ffffffff810d8e2e>] dispose_list+0x2c/0x38
[<ffffffff810d9ebd>] prune_icache_sb+0x28c/0x29b
[<ffffffff810c56b7>] prune_super+0xd5/0x140
[<ffffffff8109b615>] shrink_slab+0x102/0x1ab
[<ffffffff8109d690>] balance_pgdat+0x2f2/0x595
[<ffffffff8103e009>] ? process_timeout+0xb/0xb
[<ffffffff8109dba3>] kswapd+0x270/0x289
[<ffffffff8104c5ea>] ? __init_waitqueue_head+0x46/0x46
[<ffffffff8109d933>] ? balance_pgdat+0x595/0x595
[<ffffffff8104bf7a>] kthread+0x7f/0x87
[<ffffffff813ad6b4>] kernel_thread_helper+0x4/0x10
[<ffffffff81026b98>] ? finish_task_switch+0x45/0xc0
[<ffffffff813abcdd>] ? retint_restore_args+0xe/0xe
[<ffffffff8104befb>] ? __init_kthread_worker+0x53/0x53
[<ffffffff813ad6b0>] ? gs_change+0xb/0xb
Signed-off-by: David Howells <dhowells@redhat.com>
2012-12-21 04:52:35 +07:00
|
|
|
/**
|
|
|
|
* fscache_retrieval_complete - Record (partial) completion of a retrieval
|
|
|
|
* @op: The retrieval operation affected
|
|
|
|
* @n_pages: The number of pages to account for
|
|
|
|
*/
|
|
|
|
static inline void fscache_retrieval_complete(struct fscache_retrieval *op,
|
|
|
|
int n_pages)
|
|
|
|
{
|
2013-05-21 19:44:15 +07:00
|
|
|
atomic_sub(n_pages, &op->n_pages);
|
|
|
|
if (atomic_read(&op->n_pages) <= 0)
|
2018-04-04 19:41:26 +07:00
|
|
|
fscache_op_complete(&op->op, false);
|
FS-Cache: Fix operation state management and accounting
Fix the state management of internal fscache operations and the accounting of
what operations are in what states.
This is done by:
(1) Give struct fscache_operation a enum variable that directly represents the
state it's currently in, rather than spreading this knowledge over a bunch
of flags, who's processing the operation at the moment and whether it is
queued or not.
This makes it easier to write assertions to check the state at various
points and to prevent invalid state transitions.
(2) Add an 'operation complete' state and supply a function to indicate the
completion of an operation (fscache_op_complete()) and make things call
it. The final call to fscache_put_operation() can then check that an op
in the appropriate state (complete or cancelled).
(3) Adjust the use of object->n_ops, ->n_in_progress, ->n_exclusive to better
govern the state of an object:
(a) The ->n_ops is now the number of extant operations on the object
and is now decremented by fscache_put_operation() only.
(b) The ->n_in_progress is simply the number of objects that have been
taken off of the object's pending queue for the purposes of being
run. This is decremented by fscache_op_complete() only.
(c) The ->n_exclusive is the number of exclusive ops that have been
submitted and queued or are in progress. It is decremented by
fscache_op_complete() and by fscache_cancel_op().
fscache_put_operation() and fscache_operation_gc() now no longer try to
clean up ->n_exclusive and ->n_in_progress. That was leading to double
decrements against fscache_cancel_op().
fscache_cancel_op() now no longer decrements ->n_ops. That was leading to
double decrements against fscache_put_operation().
fscache_submit_exclusive_op() now decides whether it has to queue an op
based on ->n_in_progress being > 0 rather than ->n_ops > 0 as the latter
will persist in being true even after all preceding operations have been
cancelled or completed. Furthermore, if an object is active and there are
runnable ops against it, there must be at least one op running.
(4) Add a remaining-pages counter (n_pages) to struct fscache_retrieval and
provide a function to record completion of the pages as they complete.
When n_pages reaches 0, the operation is deemed to be complete and
fscache_op_complete() is called.
Add calls to fscache_retrieval_complete() anywhere we've finished with a
page we've been given to read or allocate for. This includes places where
we just return pages to the netfs for reading from the server and where
accessing the cache fails and we discard the proposed netfs page.
The bugs in the unfixed state management manifest themselves as oopses like the
following where the operation completion gets out of sync with return of the
cookie by the netfs. This is possible because the cache unlocks and returns
all the netfs pages before recording its completion - which means that there's
nothing to stop the netfs discarding them and returning the cookie.
FS-Cache: Cookie 'NFS.fh' still has outstanding reads
------------[ cut here ]------------
kernel BUG at fs/fscache/cookie.c:519!
invalid opcode: 0000 [#1] SMP
CPU 1
Modules linked in: cachefiles nfs fscache auth_rpcgss nfs_acl lockd sunrpc
Pid: 400, comm: kswapd0 Not tainted 3.1.0-rc7-fsdevel+ #1090 /DG965RY
RIP: 0010:[<ffffffffa007050a>] [<ffffffffa007050a>] __fscache_relinquish_cookie+0x170/0x343 [fscache]
RSP: 0018:ffff8800368cfb00 EFLAGS: 00010282
RAX: 000000000000003c RBX: ffff880023cc8790 RCX: 0000000000000000
RDX: 0000000000002f2e RSI: 0000000000000001 RDI: ffffffff813ab86c
RBP: ffff8800368cfb50 R08: 0000000000000002 R09: 0000000000000000
R10: ffff88003a1b7890 R11: ffff88001df6e488 R12: ffff880023d8ed98
R13: ffff880023cc8798 R14: 0000000000000004 R15: ffff88003b8bf370
FS: 0000000000000000(0000) GS:ffff88003bd00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00000000008ba008 CR3: 0000000023d93000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process kswapd0 (pid: 400, threadinfo ffff8800368ce000, task ffff88003b8bf040)
Stack:
ffff88003b8bf040 ffff88001df6e528 ffff88001df6e528 ffffffffa00b46b0
ffff88003b8bf040 ffff88001df6e488 ffff88001df6e620 ffffffffa00b46b0
ffff88001ebd04c8 0000000000000004 ffff8800368cfb70 ffffffffa00b2c91
Call Trace:
[<ffffffffa00b2c91>] nfs_fscache_release_inode_cookie+0x3b/0x47 [nfs]
[<ffffffffa008f25f>] nfs_clear_inode+0x3c/0x41 [nfs]
[<ffffffffa0090df1>] nfs4_evict_inode+0x2f/0x33 [nfs]
[<ffffffff810d8d47>] evict+0xa1/0x15c
[<ffffffff810d8e2e>] dispose_list+0x2c/0x38
[<ffffffff810d9ebd>] prune_icache_sb+0x28c/0x29b
[<ffffffff810c56b7>] prune_super+0xd5/0x140
[<ffffffff8109b615>] shrink_slab+0x102/0x1ab
[<ffffffff8109d690>] balance_pgdat+0x2f2/0x595
[<ffffffff8103e009>] ? process_timeout+0xb/0xb
[<ffffffff8109dba3>] kswapd+0x270/0x289
[<ffffffff8104c5ea>] ? __init_waitqueue_head+0x46/0x46
[<ffffffff8109d933>] ? balance_pgdat+0x595/0x595
[<ffffffff8104bf7a>] kthread+0x7f/0x87
[<ffffffff813ad6b4>] kernel_thread_helper+0x4/0x10
[<ffffffff81026b98>] ? finish_task_switch+0x45/0xc0
[<ffffffff813abcdd>] ? retint_restore_args+0xe/0xe
[<ffffffff8104befb>] ? __init_kthread_worker+0x53/0x53
[<ffffffff813ad6b0>] ? gs_change+0xb/0xb
Signed-off-by: David Howells <dhowells@redhat.com>
2012-12-21 04:52:35 +07:00
|
|
|
}
|
|
|
|
|
2009-04-03 22:42:36 +07:00
|
|
|
/**
|
|
|
|
* fscache_put_retrieval - Drop a reference to a retrieval operation
|
|
|
|
* @op: The retrieval operation affected
|
|
|
|
*
|
|
|
|
* Drop a reference to a retrieval operation.
|
|
|
|
*/
|
|
|
|
static inline void fscache_put_retrieval(struct fscache_retrieval *op)
|
|
|
|
{
|
|
|
|
fscache_put_operation(&op->op);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* cached page storage work item
|
|
|
|
* - used to do three things:
|
|
|
|
* - batch writes to the cache
|
|
|
|
* - do cache writes asynchronously
|
|
|
|
* - defer writes until cache object lookup completion
|
|
|
|
*/
|
|
|
|
struct fscache_storage {
|
|
|
|
struct fscache_operation op;
|
|
|
|
pgoff_t store_limit; /* don't write more than this */
|
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* cache operations
|
|
|
|
*/
|
|
|
|
struct fscache_cache_ops {
|
|
|
|
/* name of cache provider */
|
|
|
|
const char *name;
|
|
|
|
|
|
|
|
/* allocate an object record for a cookie */
|
|
|
|
struct fscache_object *(*alloc_object)(struct fscache_cache *cache,
|
|
|
|
struct fscache_cookie *cookie);
|
|
|
|
|
2009-11-20 01:12:05 +07:00
|
|
|
/* look up the object for a cookie
|
|
|
|
* - return -ETIMEDOUT to be requeued
|
|
|
|
*/
|
|
|
|
int (*lookup_object)(struct fscache_object *object);
|
2009-04-03 22:42:36 +07:00
|
|
|
|
|
|
|
/* finished looking up */
|
|
|
|
void (*lookup_complete)(struct fscache_object *object);
|
|
|
|
|
|
|
|
/* increment the usage count on this object (may fail if unmounting) */
|
2018-04-04 19:41:27 +07:00
|
|
|
struct fscache_object *(*grab_object)(struct fscache_object *object,
|
|
|
|
enum fscache_obj_ref_trace why);
|
2009-04-03 22:42:36 +07:00
|
|
|
|
|
|
|
/* pin an object in the cache */
|
|
|
|
int (*pin_object)(struct fscache_object *object);
|
|
|
|
|
|
|
|
/* unpin an object in the cache */
|
|
|
|
void (*unpin_object)(struct fscache_object *object);
|
|
|
|
|
2013-08-22 04:29:38 +07:00
|
|
|
/* check the consistency between the backing cache and the FS-Cache
|
|
|
|
* cookie */
|
2016-05-20 17:32:31 +07:00
|
|
|
int (*check_consistency)(struct fscache_operation *op);
|
2013-08-22 04:29:38 +07:00
|
|
|
|
2011-03-31 08:57:33 +07:00
|
|
|
/* store the updated auxiliary data on an object */
|
2009-04-03 22:42:36 +07:00
|
|
|
void (*update_object)(struct fscache_object *object);
|
|
|
|
|
2012-12-21 04:52:36 +07:00
|
|
|
/* Invalidate an object */
|
|
|
|
void (*invalidate_object)(struct fscache_operation *op);
|
|
|
|
|
2009-04-03 22:42:36 +07:00
|
|
|
/* discard the resources pinned by an object and effect retirement if
|
|
|
|
* necessary */
|
|
|
|
void (*drop_object)(struct fscache_object *object);
|
|
|
|
|
|
|
|
/* dispose of a reference to an object */
|
2018-04-04 19:41:27 +07:00
|
|
|
void (*put_object)(struct fscache_object *object,
|
|
|
|
enum fscache_obj_ref_trace why);
|
2009-04-03 22:42:36 +07:00
|
|
|
|
|
|
|
/* sync a cache */
|
|
|
|
void (*sync_cache)(struct fscache_cache *cache);
|
|
|
|
|
|
|
|
/* notification that the attributes of a non-index object (such as
|
|
|
|
* i_size) have changed */
|
|
|
|
int (*attr_changed)(struct fscache_object *object);
|
|
|
|
|
|
|
|
/* reserve space for an object's data and associated metadata */
|
|
|
|
int (*reserve_space)(struct fscache_object *object, loff_t i_size);
|
|
|
|
|
|
|
|
/* request a backing block for a page be read or allocated in the
|
|
|
|
* cache */
|
|
|
|
fscache_page_retrieval_func_t read_or_alloc_page;
|
|
|
|
|
|
|
|
/* request backing blocks for a list of pages be read or allocated in
|
|
|
|
* the cache */
|
|
|
|
fscache_pages_retrieval_func_t read_or_alloc_pages;
|
|
|
|
|
|
|
|
/* request a backing block for a page be allocated in the cache so that
|
|
|
|
* it can be written directly */
|
|
|
|
fscache_page_retrieval_func_t allocate_page;
|
|
|
|
|
|
|
|
/* request backing blocks for pages be allocated in the cache so that
|
|
|
|
* they can be written directly */
|
|
|
|
fscache_pages_retrieval_func_t allocate_pages;
|
|
|
|
|
|
|
|
/* write a page to its backing block in the cache */
|
|
|
|
int (*write_page)(struct fscache_storage *op, struct page *page);
|
|
|
|
|
|
|
|
/* detach backing block from a page (optional)
|
|
|
|
* - must release the cookie lock before returning
|
|
|
|
* - may sleep
|
|
|
|
*/
|
|
|
|
void (*uncache_page)(struct fscache_object *object,
|
|
|
|
struct page *page);
|
|
|
|
|
|
|
|
/* dissociate a cache from all the pages it was backing */
|
|
|
|
void (*dissociate_pages)(struct fscache_cache *cache);
|
|
|
|
};
|
|
|
|
|
|
|
|
extern struct fscache_cookie fscache_fsdef_index;
|
|
|
|
|
2012-12-05 20:34:46 +07:00
|
|
|
/*
|
|
|
|
* Event list for fscache_object::{event_mask,events}
|
|
|
|
*/
|
|
|
|
enum {
|
FS-Cache: Fix object state machine to have separate work and wait states
Fix object state machine to have separate work and wait states as that makes
it easier to envision.
There are now three kinds of state:
(1) Work state. This is an execution state. No event processing is performed
by a work state. The function attached to a work state returns a pointer
indicating the next state to which the OSM should transition. Returning
NO_TRANSIT repeats the current state, but goes back to the scheduler
first.
(2) Wait state. This is an event processing state. No execution is
performed by a wait state. Wait states are just tables of "if event X
occurs, clear it and transition to state Y". The dispatcher returns to
the scheduler if none of the events in which the wait state has an
interest are currently pending.
(3) Out-of-band state. This is a special work state. Transitions to normal
states can be overridden when an unexpected event occurs (eg. I/O error).
Instead the dispatcher disables and clears the OOB event and transits to
the specified work state. This then acts as an ordinary work state,
though object->state points to the overridden destination. Returning
NO_TRANSIT resumes the overridden transition.
In addition, the states have names in their definitions, so there's no need for
tables of state names. Further, the EV_REQUEUE event is no longer necessary as
that is automatic for work states.
Since the states are now separate structs rather than values in an enum, it's
not possible to use comparisons other than (non-)equality between them, so use
some object->flags to indicate what phase an object is in.
The EV_RELEASE, EV_RETIRE and EV_WITHDRAW events have been squished into one
(EV_KILL). An object flag now carries the information about retirement.
Similarly, the RELEASING, RECYCLING and WITHDRAWING states have been merged
into an KILL_OBJECT state and additional states have been added for handling
waiting dependent objects (JUMPSTART_DEPS and KILL_DEPENDENTS).
A state has also been added for synchronising with parent object initialisation
(WAIT_FOR_PARENT) and another for initiating look up (PARENT_READY).
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
FSCACHE_OBJECT_EV_NEW_CHILD, /* T if object has a new child */
|
|
|
|
FSCACHE_OBJECT_EV_PARENT_READY, /* T if object's parent is ready */
|
2012-12-05 20:34:46 +07:00
|
|
|
FSCACHE_OBJECT_EV_UPDATE, /* T if object should be updated */
|
|
|
|
FSCACHE_OBJECT_EV_INVALIDATE, /* T if cache requested object invalidation */
|
|
|
|
FSCACHE_OBJECT_EV_CLEARED, /* T if accessors all gone */
|
|
|
|
FSCACHE_OBJECT_EV_ERROR, /* T if fatal error occurred during processing */
|
FS-Cache: Fix object state machine to have separate work and wait states
Fix object state machine to have separate work and wait states as that makes
it easier to envision.
There are now three kinds of state:
(1) Work state. This is an execution state. No event processing is performed
by a work state. The function attached to a work state returns a pointer
indicating the next state to which the OSM should transition. Returning
NO_TRANSIT repeats the current state, but goes back to the scheduler
first.
(2) Wait state. This is an event processing state. No execution is
performed by a wait state. Wait states are just tables of "if event X
occurs, clear it and transition to state Y". The dispatcher returns to
the scheduler if none of the events in which the wait state has an
interest are currently pending.
(3) Out-of-band state. This is a special work state. Transitions to normal
states can be overridden when an unexpected event occurs (eg. I/O error).
Instead the dispatcher disables and clears the OOB event and transits to
the specified work state. This then acts as an ordinary work state,
though object->state points to the overridden destination. Returning
NO_TRANSIT resumes the overridden transition.
In addition, the states have names in their definitions, so there's no need for
tables of state names. Further, the EV_REQUEUE event is no longer necessary as
that is automatic for work states.
Since the states are now separate structs rather than values in an enum, it's
not possible to use comparisons other than (non-)equality between them, so use
some object->flags to indicate what phase an object is in.
The EV_RELEASE, EV_RETIRE and EV_WITHDRAW events have been squished into one
(EV_KILL). An object flag now carries the information about retirement.
Similarly, the RELEASING, RECYCLING and WITHDRAWING states have been merged
into an KILL_OBJECT state and additional states have been added for handling
waiting dependent objects (JUMPSTART_DEPS and KILL_DEPENDENTS).
A state has also been added for synchronising with parent object initialisation
(WAIT_FOR_PARENT) and another for initiating look up (PARENT_READY).
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
FSCACHE_OBJECT_EV_KILL, /* T if netfs relinquished or cache withdrew object */
|
2012-12-05 20:34:46 +07:00
|
|
|
NR_FSCACHE_OBJECT_EVENTS
|
|
|
|
};
|
|
|
|
|
|
|
|
#define FSCACHE_OBJECT_EVENTS_MASK ((1UL << NR_FSCACHE_OBJECT_EVENTS) - 1)
|
|
|
|
|
FS-Cache: Fix object state machine to have separate work and wait states
Fix object state machine to have separate work and wait states as that makes
it easier to envision.
There are now three kinds of state:
(1) Work state. This is an execution state. No event processing is performed
by a work state. The function attached to a work state returns a pointer
indicating the next state to which the OSM should transition. Returning
NO_TRANSIT repeats the current state, but goes back to the scheduler
first.
(2) Wait state. This is an event processing state. No execution is
performed by a wait state. Wait states are just tables of "if event X
occurs, clear it and transition to state Y". The dispatcher returns to
the scheduler if none of the events in which the wait state has an
interest are currently pending.
(3) Out-of-band state. This is a special work state. Transitions to normal
states can be overridden when an unexpected event occurs (eg. I/O error).
Instead the dispatcher disables and clears the OOB event and transits to
the specified work state. This then acts as an ordinary work state,
though object->state points to the overridden destination. Returning
NO_TRANSIT resumes the overridden transition.
In addition, the states have names in their definitions, so there's no need for
tables of state names. Further, the EV_REQUEUE event is no longer necessary as
that is automatic for work states.
Since the states are now separate structs rather than values in an enum, it's
not possible to use comparisons other than (non-)equality between them, so use
some object->flags to indicate what phase an object is in.
The EV_RELEASE, EV_RETIRE and EV_WITHDRAW events have been squished into one
(EV_KILL). An object flag now carries the information about retirement.
Similarly, the RELEASING, RECYCLING and WITHDRAWING states have been merged
into an KILL_OBJECT state and additional states have been added for handling
waiting dependent objects (JUMPSTART_DEPS and KILL_DEPENDENTS).
A state has also been added for synchronising with parent object initialisation
(WAIT_FOR_PARENT) and another for initiating look up (PARENT_READY).
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
/*
|
|
|
|
* States for object state machine.
|
|
|
|
*/
|
|
|
|
struct fscache_transition {
|
|
|
|
unsigned long events;
|
|
|
|
const struct fscache_state *transit_to;
|
|
|
|
};
|
|
|
|
|
|
|
|
struct fscache_state {
|
|
|
|
char name[24];
|
|
|
|
char short_name[8];
|
|
|
|
const struct fscache_state *(*work)(struct fscache_object *object,
|
|
|
|
int event);
|
|
|
|
const struct fscache_transition transitions[];
|
|
|
|
};
|
|
|
|
|
2009-04-03 22:42:36 +07:00
|
|
|
/*
|
|
|
|
* on-disk cache file or index handle
|
|
|
|
*/
|
|
|
|
struct fscache_object {
|
FS-Cache: Fix object state machine to have separate work and wait states
Fix object state machine to have separate work and wait states as that makes
it easier to envision.
There are now three kinds of state:
(1) Work state. This is an execution state. No event processing is performed
by a work state. The function attached to a work state returns a pointer
indicating the next state to which the OSM should transition. Returning
NO_TRANSIT repeats the current state, but goes back to the scheduler
first.
(2) Wait state. This is an event processing state. No execution is
performed by a wait state. Wait states are just tables of "if event X
occurs, clear it and transition to state Y". The dispatcher returns to
the scheduler if none of the events in which the wait state has an
interest are currently pending.
(3) Out-of-band state. This is a special work state. Transitions to normal
states can be overridden when an unexpected event occurs (eg. I/O error).
Instead the dispatcher disables and clears the OOB event and transits to
the specified work state. This then acts as an ordinary work state,
though object->state points to the overridden destination. Returning
NO_TRANSIT resumes the overridden transition.
In addition, the states have names in their definitions, so there's no need for
tables of state names. Further, the EV_REQUEUE event is no longer necessary as
that is automatic for work states.
Since the states are now separate structs rather than values in an enum, it's
not possible to use comparisons other than (non-)equality between them, so use
some object->flags to indicate what phase an object is in.
The EV_RELEASE, EV_RETIRE and EV_WITHDRAW events have been squished into one
(EV_KILL). An object flag now carries the information about retirement.
Similarly, the RELEASING, RECYCLING and WITHDRAWING states have been merged
into an KILL_OBJECT state and additional states have been added for handling
waiting dependent objects (JUMPSTART_DEPS and KILL_DEPENDENTS).
A state has also been added for synchronising with parent object initialisation
(WAIT_FOR_PARENT) and another for initiating look up (PARENT_READY).
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
const struct fscache_state *state; /* Object state machine state */
|
|
|
|
const struct fscache_transition *oob_table; /* OOB state transition table */
|
2009-04-03 22:42:36 +07:00
|
|
|
int debug_id; /* debugging ID */
|
|
|
|
int n_children; /* number of child objects */
|
FS-Cache: Fix operation state management and accounting
Fix the state management of internal fscache operations and the accounting of
what operations are in what states.
This is done by:
(1) Give struct fscache_operation a enum variable that directly represents the
state it's currently in, rather than spreading this knowledge over a bunch
of flags, who's processing the operation at the moment and whether it is
queued or not.
This makes it easier to write assertions to check the state at various
points and to prevent invalid state transitions.
(2) Add an 'operation complete' state and supply a function to indicate the
completion of an operation (fscache_op_complete()) and make things call
it. The final call to fscache_put_operation() can then check that an op
in the appropriate state (complete or cancelled).
(3) Adjust the use of object->n_ops, ->n_in_progress, ->n_exclusive to better
govern the state of an object:
(a) The ->n_ops is now the number of extant operations on the object
and is now decremented by fscache_put_operation() only.
(b) The ->n_in_progress is simply the number of objects that have been
taken off of the object's pending queue for the purposes of being
run. This is decremented by fscache_op_complete() only.
(c) The ->n_exclusive is the number of exclusive ops that have been
submitted and queued or are in progress. It is decremented by
fscache_op_complete() and by fscache_cancel_op().
fscache_put_operation() and fscache_operation_gc() now no longer try to
clean up ->n_exclusive and ->n_in_progress. That was leading to double
decrements against fscache_cancel_op().
fscache_cancel_op() now no longer decrements ->n_ops. That was leading to
double decrements against fscache_put_operation().
fscache_submit_exclusive_op() now decides whether it has to queue an op
based on ->n_in_progress being > 0 rather than ->n_ops > 0 as the latter
will persist in being true even after all preceding operations have been
cancelled or completed. Furthermore, if an object is active and there are
runnable ops against it, there must be at least one op running.
(4) Add a remaining-pages counter (n_pages) to struct fscache_retrieval and
provide a function to record completion of the pages as they complete.
When n_pages reaches 0, the operation is deemed to be complete and
fscache_op_complete() is called.
Add calls to fscache_retrieval_complete() anywhere we've finished with a
page we've been given to read or allocate for. This includes places where
we just return pages to the netfs for reading from the server and where
accessing the cache fails and we discard the proposed netfs page.
The bugs in the unfixed state management manifest themselves as oopses like the
following where the operation completion gets out of sync with return of the
cookie by the netfs. This is possible because the cache unlocks and returns
all the netfs pages before recording its completion - which means that there's
nothing to stop the netfs discarding them and returning the cookie.
FS-Cache: Cookie 'NFS.fh' still has outstanding reads
------------[ cut here ]------------
kernel BUG at fs/fscache/cookie.c:519!
invalid opcode: 0000 [#1] SMP
CPU 1
Modules linked in: cachefiles nfs fscache auth_rpcgss nfs_acl lockd sunrpc
Pid: 400, comm: kswapd0 Not tainted 3.1.0-rc7-fsdevel+ #1090 /DG965RY
RIP: 0010:[<ffffffffa007050a>] [<ffffffffa007050a>] __fscache_relinquish_cookie+0x170/0x343 [fscache]
RSP: 0018:ffff8800368cfb00 EFLAGS: 00010282
RAX: 000000000000003c RBX: ffff880023cc8790 RCX: 0000000000000000
RDX: 0000000000002f2e RSI: 0000000000000001 RDI: ffffffff813ab86c
RBP: ffff8800368cfb50 R08: 0000000000000002 R09: 0000000000000000
R10: ffff88003a1b7890 R11: ffff88001df6e488 R12: ffff880023d8ed98
R13: ffff880023cc8798 R14: 0000000000000004 R15: ffff88003b8bf370
FS: 0000000000000000(0000) GS:ffff88003bd00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00000000008ba008 CR3: 0000000023d93000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process kswapd0 (pid: 400, threadinfo ffff8800368ce000, task ffff88003b8bf040)
Stack:
ffff88003b8bf040 ffff88001df6e528 ffff88001df6e528 ffffffffa00b46b0
ffff88003b8bf040 ffff88001df6e488 ffff88001df6e620 ffffffffa00b46b0
ffff88001ebd04c8 0000000000000004 ffff8800368cfb70 ffffffffa00b2c91
Call Trace:
[<ffffffffa00b2c91>] nfs_fscache_release_inode_cookie+0x3b/0x47 [nfs]
[<ffffffffa008f25f>] nfs_clear_inode+0x3c/0x41 [nfs]
[<ffffffffa0090df1>] nfs4_evict_inode+0x2f/0x33 [nfs]
[<ffffffff810d8d47>] evict+0xa1/0x15c
[<ffffffff810d8e2e>] dispose_list+0x2c/0x38
[<ffffffff810d9ebd>] prune_icache_sb+0x28c/0x29b
[<ffffffff810c56b7>] prune_super+0xd5/0x140
[<ffffffff8109b615>] shrink_slab+0x102/0x1ab
[<ffffffff8109d690>] balance_pgdat+0x2f2/0x595
[<ffffffff8103e009>] ? process_timeout+0xb/0xb
[<ffffffff8109dba3>] kswapd+0x270/0x289
[<ffffffff8104c5ea>] ? __init_waitqueue_head+0x46/0x46
[<ffffffff8109d933>] ? balance_pgdat+0x595/0x595
[<ffffffff8104bf7a>] kthread+0x7f/0x87
[<ffffffff813ad6b4>] kernel_thread_helper+0x4/0x10
[<ffffffff81026b98>] ? finish_task_switch+0x45/0xc0
[<ffffffff813abcdd>] ? retint_restore_args+0xe/0xe
[<ffffffff8104befb>] ? __init_kthread_worker+0x53/0x53
[<ffffffff813ad6b0>] ? gs_change+0xb/0xb
Signed-off-by: David Howells <dhowells@redhat.com>
2012-12-21 04:52:35 +07:00
|
|
|
int n_ops; /* number of extant ops on object */
|
2009-04-03 22:42:36 +07:00
|
|
|
int n_obj_ops; /* number of object ops outstanding on object */
|
|
|
|
int n_in_progress; /* number of ops in progress */
|
FS-Cache: Fix operation state management and accounting
Fix the state management of internal fscache operations and the accounting of
what operations are in what states.
This is done by:
(1) Give struct fscache_operation a enum variable that directly represents the
state it's currently in, rather than spreading this knowledge over a bunch
of flags, who's processing the operation at the moment and whether it is
queued or not.
This makes it easier to write assertions to check the state at various
points and to prevent invalid state transitions.
(2) Add an 'operation complete' state and supply a function to indicate the
completion of an operation (fscache_op_complete()) and make things call
it. The final call to fscache_put_operation() can then check that an op
in the appropriate state (complete or cancelled).
(3) Adjust the use of object->n_ops, ->n_in_progress, ->n_exclusive to better
govern the state of an object:
(a) The ->n_ops is now the number of extant operations on the object
and is now decremented by fscache_put_operation() only.
(b) The ->n_in_progress is simply the number of objects that have been
taken off of the object's pending queue for the purposes of being
run. This is decremented by fscache_op_complete() only.
(c) The ->n_exclusive is the number of exclusive ops that have been
submitted and queued or are in progress. It is decremented by
fscache_op_complete() and by fscache_cancel_op().
fscache_put_operation() and fscache_operation_gc() now no longer try to
clean up ->n_exclusive and ->n_in_progress. That was leading to double
decrements against fscache_cancel_op().
fscache_cancel_op() now no longer decrements ->n_ops. That was leading to
double decrements against fscache_put_operation().
fscache_submit_exclusive_op() now decides whether it has to queue an op
based on ->n_in_progress being > 0 rather than ->n_ops > 0 as the latter
will persist in being true even after all preceding operations have been
cancelled or completed. Furthermore, if an object is active and there are
runnable ops against it, there must be at least one op running.
(4) Add a remaining-pages counter (n_pages) to struct fscache_retrieval and
provide a function to record completion of the pages as they complete.
When n_pages reaches 0, the operation is deemed to be complete and
fscache_op_complete() is called.
Add calls to fscache_retrieval_complete() anywhere we've finished with a
page we've been given to read or allocate for. This includes places where
we just return pages to the netfs for reading from the server and where
accessing the cache fails and we discard the proposed netfs page.
The bugs in the unfixed state management manifest themselves as oopses like the
following where the operation completion gets out of sync with return of the
cookie by the netfs. This is possible because the cache unlocks and returns
all the netfs pages before recording its completion - which means that there's
nothing to stop the netfs discarding them and returning the cookie.
FS-Cache: Cookie 'NFS.fh' still has outstanding reads
------------[ cut here ]------------
kernel BUG at fs/fscache/cookie.c:519!
invalid opcode: 0000 [#1] SMP
CPU 1
Modules linked in: cachefiles nfs fscache auth_rpcgss nfs_acl lockd sunrpc
Pid: 400, comm: kswapd0 Not tainted 3.1.0-rc7-fsdevel+ #1090 /DG965RY
RIP: 0010:[<ffffffffa007050a>] [<ffffffffa007050a>] __fscache_relinquish_cookie+0x170/0x343 [fscache]
RSP: 0018:ffff8800368cfb00 EFLAGS: 00010282
RAX: 000000000000003c RBX: ffff880023cc8790 RCX: 0000000000000000
RDX: 0000000000002f2e RSI: 0000000000000001 RDI: ffffffff813ab86c
RBP: ffff8800368cfb50 R08: 0000000000000002 R09: 0000000000000000
R10: ffff88003a1b7890 R11: ffff88001df6e488 R12: ffff880023d8ed98
R13: ffff880023cc8798 R14: 0000000000000004 R15: ffff88003b8bf370
FS: 0000000000000000(0000) GS:ffff88003bd00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00000000008ba008 CR3: 0000000023d93000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process kswapd0 (pid: 400, threadinfo ffff8800368ce000, task ffff88003b8bf040)
Stack:
ffff88003b8bf040 ffff88001df6e528 ffff88001df6e528 ffffffffa00b46b0
ffff88003b8bf040 ffff88001df6e488 ffff88001df6e620 ffffffffa00b46b0
ffff88001ebd04c8 0000000000000004 ffff8800368cfb70 ffffffffa00b2c91
Call Trace:
[<ffffffffa00b2c91>] nfs_fscache_release_inode_cookie+0x3b/0x47 [nfs]
[<ffffffffa008f25f>] nfs_clear_inode+0x3c/0x41 [nfs]
[<ffffffffa0090df1>] nfs4_evict_inode+0x2f/0x33 [nfs]
[<ffffffff810d8d47>] evict+0xa1/0x15c
[<ffffffff810d8e2e>] dispose_list+0x2c/0x38
[<ffffffff810d9ebd>] prune_icache_sb+0x28c/0x29b
[<ffffffff810c56b7>] prune_super+0xd5/0x140
[<ffffffff8109b615>] shrink_slab+0x102/0x1ab
[<ffffffff8109d690>] balance_pgdat+0x2f2/0x595
[<ffffffff8103e009>] ? process_timeout+0xb/0xb
[<ffffffff8109dba3>] kswapd+0x270/0x289
[<ffffffff8104c5ea>] ? __init_waitqueue_head+0x46/0x46
[<ffffffff8109d933>] ? balance_pgdat+0x595/0x595
[<ffffffff8104bf7a>] kthread+0x7f/0x87
[<ffffffff813ad6b4>] kernel_thread_helper+0x4/0x10
[<ffffffff81026b98>] ? finish_task_switch+0x45/0xc0
[<ffffffff813abcdd>] ? retint_restore_args+0xe/0xe
[<ffffffff8104befb>] ? __init_kthread_worker+0x53/0x53
[<ffffffff813ad6b0>] ? gs_change+0xb/0xb
Signed-off-by: David Howells <dhowells@redhat.com>
2012-12-21 04:52:35 +07:00
|
|
|
int n_exclusive; /* number of exclusive ops queued or in progress */
|
2009-11-20 01:11:04 +07:00
|
|
|
atomic_t n_reads; /* number of read ops in progress */
|
2009-04-03 22:42:36 +07:00
|
|
|
spinlock_t lock; /* state and operations lock */
|
|
|
|
|
|
|
|
unsigned long lookup_jif; /* time at which lookup started */
|
FS-Cache: Fix object state machine to have separate work and wait states
Fix object state machine to have separate work and wait states as that makes
it easier to envision.
There are now three kinds of state:
(1) Work state. This is an execution state. No event processing is performed
by a work state. The function attached to a work state returns a pointer
indicating the next state to which the OSM should transition. Returning
NO_TRANSIT repeats the current state, but goes back to the scheduler
first.
(2) Wait state. This is an event processing state. No execution is
performed by a wait state. Wait states are just tables of "if event X
occurs, clear it and transition to state Y". The dispatcher returns to
the scheduler if none of the events in which the wait state has an
interest are currently pending.
(3) Out-of-band state. This is a special work state. Transitions to normal
states can be overridden when an unexpected event occurs (eg. I/O error).
Instead the dispatcher disables and clears the OOB event and transits to
the specified work state. This then acts as an ordinary work state,
though object->state points to the overridden destination. Returning
NO_TRANSIT resumes the overridden transition.
In addition, the states have names in their definitions, so there's no need for
tables of state names. Further, the EV_REQUEUE event is no longer necessary as
that is automatic for work states.
Since the states are now separate structs rather than values in an enum, it's
not possible to use comparisons other than (non-)equality between them, so use
some object->flags to indicate what phase an object is in.
The EV_RELEASE, EV_RETIRE and EV_WITHDRAW events have been squished into one
(EV_KILL). An object flag now carries the information about retirement.
Similarly, the RELEASING, RECYCLING and WITHDRAWING states have been merged
into an KILL_OBJECT state and additional states have been added for handling
waiting dependent objects (JUMPSTART_DEPS and KILL_DEPENDENTS).
A state has also been added for synchronising with parent object initialisation
(WAIT_FOR_PARENT) and another for initiating look up (PARENT_READY).
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
unsigned long oob_event_mask; /* OOB events this object is interested in */
|
2009-04-03 22:42:36 +07:00
|
|
|
unsigned long event_mask; /* events this object is interested in */
|
|
|
|
unsigned long events; /* events to be processed by this object
|
|
|
|
* (order is important - using fls) */
|
|
|
|
|
|
|
|
unsigned long flags;
|
|
|
|
#define FSCACHE_OBJECT_LOCK 0 /* T if object is busy being processed */
|
|
|
|
#define FSCACHE_OBJECT_PENDING_WRITE 1 /* T if object has pending write */
|
|
|
|
#define FSCACHE_OBJECT_WAITING 2 /* T if object is waiting on its parent */
|
FS-Cache: Simplify cookie retention for fscache_objects, fixing oops
Simplify the way fscache cache objects retain their cookie. The way I
implemented the cookie storage handling made synchronisation a pain (ie. the
object state machine can't rely on the cookie actually still being there).
Instead of the the object being detached from the cookie and the cookie being
freed in __fscache_relinquish_cookie(), we defer both operations:
(*) The detachment of the object from the list in the cookie now takes place
in fscache_drop_object() and is thus governed by the object state machine
(fscache_detach_from_cookie() has been removed).
(*) The release of the cookie is now in fscache_object_destroy() - which is
called by the cache backend just before it frees the object.
This means that the fscache_cookie struct is now available to the cache all the
way through from ->alloc_object() to ->drop_object() and ->put_object() -
meaning that it's no longer necessary to take object->lock to guarantee access.
However, __fscache_relinquish_cookie() doesn't wait for the object to go all
the way through to destruction before letting the netfs proceed. That would
massively slow down the netfs. Since __fscache_relinquish_cookie() leaves the
cookie around, in must therefore break all attachments to the netfs - which
includes ->def, ->netfs_data and any outstanding page read/writes.
To handle this, struct fscache_cookie now has an n_active counter:
(1) This starts off initialised to 1.
(2) Any time the cache needs to get at the netfs data, it calls
fscache_use_cookie() to increment it - if it is not zero. If it was zero,
then access is not permitted.
(3) When the cache has finished with the data, it calls fscache_unuse_cookie()
to decrement it. This does a wake-up on it if it reaches 0.
(4) __fscache_relinquish_cookie() decrements n_active and then waits for it to
reach 0. The initialisation to 1 in step (1) ensures that we only get
wake ups when we're trying to get rid of the cookie.
This leaves __fscache_relinquish_cookie() a lot simpler.
***
This fixes a problem in the current code whereby if fscache_invalidate() is
followed sufficiently quickly by fscache_relinquish_cookie() then it is
possible for __fscache_relinquish_cookie() to have detached the cookie from the
object and cleared the pointer before a thread is dispatched to process the
invalidation state in the object state machine.
Since the pending write clearance was deferred to the invalidation state to
make it asynchronous, we need to either wait in relinquishment for the stores
tree to be cleared in the invalidation state or we need to handle the clearance
in relinquishment.
Further, if the relinquishment code does clear the tree, then the invalidation
state need to make the clearance contingent on still having the cookie to hand
(since that's where the tree is rooted) and we have to prevent the cookie from
disappearing for the duration.
This can lead to an oops like the following:
BUG: unable to handle kernel NULL pointer dereference at 000000000000000c
...
RIP: 0010:[<ffffffff8151023e>] _spin_lock+0xe/0x30
...
CR2: 000000000000000c ...
...
Process kslowd002 (...)
....
Call Trace:
[<ffffffffa01c3278>] fscache_invalidate_writes+0x38/0xd0 [fscache]
[<ffffffff810096f0>] ? __switch_to+0xd0/0x320
[<ffffffff8105e759>] ? find_busiest_queue+0x69/0x150
[<ffffffff8110ddd4>] ? slow_work_enqueue+0x104/0x180
[<ffffffffa01c1303>] fscache_object_slow_work_execute+0x5e3/0x9d0 [fscache]
[<ffffffff81096b67>] ? bit_waitqueue+0x17/0xd0
[<ffffffff8110e233>] slow_work_execute+0x233/0x310
[<ffffffff8110e515>] slow_work_thread+0x205/0x360
[<ffffffff81096ca0>] ? autoremove_wake_function+0x0/0x40
[<ffffffff8110e310>] ? slow_work_thread+0x0/0x360
[<ffffffff81096936>] kthread+0x96/0xa0
[<ffffffff8100c0ca>] child_rip+0xa/0x20
[<ffffffff810968a0>] ? kthread+0x0/0xa0
[<ffffffff8100c0c0>] ? child_rip+0x0/0x20
The parameter to fscache_invalidate_writes() was object->cookie which is NULL.
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
#define FSCACHE_OBJECT_IS_LIVE 3 /* T if object is not withdrawn or relinquished */
|
|
|
|
#define FSCACHE_OBJECT_IS_LOOKED_UP 4 /* T if object has been looked up */
|
|
|
|
#define FSCACHE_OBJECT_IS_AVAILABLE 5 /* T if object has become active */
|
FS-Cache: Provide the ability to enable/disable cookies
Provide the ability to enable and disable fscache cookies. A disabled cookie
will reject or ignore further requests to:
Acquire a child cookie
Invalidate and update backing objects
Check the consistency of a backing object
Allocate storage for backing page
Read backing pages
Write to backing pages
but still allows:
Checks/waits on the completion of already in-progress objects
Uncaching of pages
Relinquishment of cookies
Two new operations are provided:
(1) Disable a cookie:
void fscache_disable_cookie(struct fscache_cookie *cookie,
bool invalidate);
If the cookie is not already disabled, this locks the cookie against other
dis/enablement ops, marks the cookie as being disabled, discards or
invalidates any backing objects and waits for cessation of activity on any
associated object.
This is a wrapper around a chunk split out of fscache_relinquish_cookie(),
but it reinitialises the cookie such that it can be reenabled.
All possible failures are handled internally. The caller should consider
calling fscache_uncache_all_inode_pages() afterwards to make sure all page
markings are cleared up.
(2) Enable a cookie:
void fscache_enable_cookie(struct fscache_cookie *cookie,
bool (*can_enable)(void *data),
void *data)
If the cookie is not already enabled, this locks the cookie against other
dis/enablement ops, invokes can_enable() and, if the cookie is not an
index cookie, will begin the procedure of acquiring backing objects.
The optional can_enable() function is passed the data argument and returns
a ruling as to whether or not enablement should actually be permitted to
begin.
All possible failures are handled internally. The cookie will only be
marked as enabled if provisional backing objects are allocated.
A later patch will introduce these to NFS. Cookie enablement during nfs_open()
is then contingent on i_writecount <= 0. can_enable() checks for a race
between open(O_RDONLY) and open(O_WRONLY/O_RDWR). This simplifies NFS's cookie
handling and allows us to get rid of open(O_RDONLY) accidentally introducing
caching to an inode that's open for writing already.
One operation has its API modified:
(3) Acquire a cookie.
struct fscache_cookie *fscache_acquire_cookie(
struct fscache_cookie *parent,
const struct fscache_cookie_def *def,
void *netfs_data,
bool enable);
This now has an additional argument that indicates whether the requested
cookie should be enabled by default. It doesn't need the can_enable()
function because the caller must prevent multiple calls for the same netfs
object and it doesn't need to take the enablement lock because no one else
can get at the cookie before this returns.
Signed-off-by: David Howells <dhowells@redhat.com
2013-09-21 06:09:31 +07:00
|
|
|
#define FSCACHE_OBJECT_RETIRED 6 /* T if object was retired on relinquishment */
|
2015-02-20 06:47:31 +07:00
|
|
|
#define FSCACHE_OBJECT_KILLED_BY_CACHE 7 /* T if object was killed by the cache */
|
fscache: Fix dead object requeue
Under some circumstances, an fscache object can become queued such that it
fscache_object_work_func() can be called once the object is in the
OBJECT_DEAD state. This results in the kernel oopsing when it tries to
invoke the handler for the state (which is hard coded to 0x2).
The way this comes about is something like the following:
(1) The object dispatcher is processing a work state for an object. This
is done in workqueue context.
(2) An out-of-band event comes in that isn't masked, causing the object to
be queued, say EV_KILL.
(3) The object dispatcher finishes processing the current work state on
that object and then sees there's another event to process, so,
without returning to the workqueue core, it processes that event too.
It then follows the chain of events that initiates until we reach
OBJECT_DEAD without going through a wait state (such as
WAIT_FOR_CLEARANCE).
At this point, object->events may be 0, object->event_mask will be 0
and oob_event_mask will be 0.
(4) The object dispatcher returns to the workqueue processor, and in due
course, this sees that the object's work item is still queued and
invokes it again.
(5) The current state is a work state (OBJECT_DEAD), so the dispatcher
jumps to it - resulting in an OOPS.
When I'm seeing this, the work state in (1) appears to have been either
LOOK_UP_OBJECT or CREATE_OBJECT (object->oob_table is
fscache_osm_lookup_oob).
The window for (2) is very small:
(A) object->event_mask is cleared whilst the event dispatch process is
underway - though there's no memory barrier to force this to the top
of the function.
The window, therefore is from the time the object was selected by the
workqueue processor and made requeueable to the time the mask was
cleared.
(B) fscache_raise_event() will only queue the object if it manages to set
the event bit and the corresponding event_mask bit was set.
The enqueuement is then deferred slightly whilst we get a ref on the
object and get the per-CPU variable for workqueue congestion. This
slight deferral slightly increases the probability by allowing extra
time for the workqueue to make the item requeueable.
Handle this by giving the dead state a processor function and checking the
for the dead state address rather than seeing if the processor function is
address 0x2. The dead state processor function can then set a flag to
indicate that it's occurred and give a warning if it occurs more than once
per object.
If this race occurs, an oops similar to the following is seen (note the RIP
value):
BUG: unable to handle kernel NULL pointer dereference at 0000000000000002
IP: [<0000000000000002>] 0x1
PGD 0
Oops: 0010 [#1] SMP
Modules linked in: ...
CPU: 17 PID: 16077 Comm: kworker/u48:9 Not tainted 3.10.0-327.18.2.el7.x86_64 #1
Hardware name: HP ProLiant DL380 Gen9/ProLiant DL380 Gen9, BIOS P89 12/27/2015
Workqueue: fscache_object fscache_object_work_func [fscache]
task: ffff880302b63980 ti: ffff880717544000 task.ti: ffff880717544000
RIP: 0010:[<0000000000000002>] [<0000000000000002>] 0x1
RSP: 0018:ffff880717547df8 EFLAGS: 00010202
RAX: ffffffffa0368640 RBX: ffff880edf7a4480 RCX: dead000000200200
RDX: 0000000000000002 RSI: 00000000ffffffff RDI: ffff880edf7a4480
RBP: ffff880717547e18 R08: 0000000000000000 R09: dfc40a25cb3a4510
R10: dfc40a25cb3a4510 R11: 0000000000000400 R12: 0000000000000000
R13: ffff880edf7a4510 R14: ffff8817f6153400 R15: 0000000000000600
FS: 0000000000000000(0000) GS:ffff88181f420000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000002 CR3: 000000000194a000 CR4: 00000000001407e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Stack:
ffffffffa0363695 ffff880edf7a4510 ffff88093f16f900 ffff8817faa4ec00
ffff880717547e60 ffffffff8109d5db 00000000faa4ec18 0000000000000000
ffff8817faa4ec18 ffff88093f16f930 ffff880302b63980 ffff88093f16f900
Call Trace:
[<ffffffffa0363695>] ? fscache_object_work_func+0xa5/0x200 [fscache]
[<ffffffff8109d5db>] process_one_work+0x17b/0x470
[<ffffffff8109e4ac>] worker_thread+0x21c/0x400
[<ffffffff8109e290>] ? rescuer_thread+0x400/0x400
[<ffffffff810a5acf>] kthread+0xcf/0xe0
[<ffffffff810a5a00>] ? kthread_create_on_node+0x140/0x140
[<ffffffff816460d8>] ret_from_fork+0x58/0x90
[<ffffffff810a5a00>] ? kthread_create_on_node+0x140/0x140
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Jeremy McNicoll <jeremymc@redhat.com>
Tested-by: Frank Sorenson <sorenson@redhat.com>
Tested-by: Benjamin Coddington <bcodding@redhat.com>
Reviewed-by: Benjamin Coddington <bcodding@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2017-01-31 16:45:28 +07:00
|
|
|
#define FSCACHE_OBJECT_RUN_AFTER_DEAD 8 /* T if object has been dispatched after death */
|
2009-04-03 22:42:36 +07:00
|
|
|
|
|
|
|
struct list_head cache_link; /* link in cache->object_list */
|
|
|
|
struct hlist_node cookie_link; /* link in cookie->backing_objects */
|
|
|
|
struct fscache_cache *cache; /* cache that supplied this object */
|
|
|
|
struct fscache_cookie *cookie; /* netfs's file/index object */
|
|
|
|
struct fscache_object *parent; /* parent object */
|
2010-07-21 03:09:01 +07:00
|
|
|
struct work_struct work; /* attention scheduling record */
|
2009-04-03 22:42:36 +07:00
|
|
|
struct list_head dependents; /* FIFO of dependent objects */
|
|
|
|
struct list_head dep_link; /* link in parent's dependents list */
|
|
|
|
struct list_head pending_ops; /* unstarted operations on this object */
|
2009-11-20 01:11:04 +07:00
|
|
|
#ifdef CONFIG_FSCACHE_OBJECT_LIST
|
|
|
|
struct rb_node objlist_link; /* link in global object list */
|
|
|
|
#endif
|
2009-04-03 22:42:36 +07:00
|
|
|
pgoff_t store_limit; /* current storage limit */
|
2009-11-20 01:11:52 +07:00
|
|
|
loff_t store_limit_l; /* current storage limit */
|
2009-04-03 22:42:36 +07:00
|
|
|
};
|
|
|
|
|
2013-05-11 01:50:25 +07:00
|
|
|
extern void fscache_object_init(struct fscache_object *, struct fscache_cookie *,
|
|
|
|
struct fscache_cache *);
|
FS-Cache: Simplify cookie retention for fscache_objects, fixing oops
Simplify the way fscache cache objects retain their cookie. The way I
implemented the cookie storage handling made synchronisation a pain (ie. the
object state machine can't rely on the cookie actually still being there).
Instead of the the object being detached from the cookie and the cookie being
freed in __fscache_relinquish_cookie(), we defer both operations:
(*) The detachment of the object from the list in the cookie now takes place
in fscache_drop_object() and is thus governed by the object state machine
(fscache_detach_from_cookie() has been removed).
(*) The release of the cookie is now in fscache_object_destroy() - which is
called by the cache backend just before it frees the object.
This means that the fscache_cookie struct is now available to the cache all the
way through from ->alloc_object() to ->drop_object() and ->put_object() -
meaning that it's no longer necessary to take object->lock to guarantee access.
However, __fscache_relinquish_cookie() doesn't wait for the object to go all
the way through to destruction before letting the netfs proceed. That would
massively slow down the netfs. Since __fscache_relinquish_cookie() leaves the
cookie around, in must therefore break all attachments to the netfs - which
includes ->def, ->netfs_data and any outstanding page read/writes.
To handle this, struct fscache_cookie now has an n_active counter:
(1) This starts off initialised to 1.
(2) Any time the cache needs to get at the netfs data, it calls
fscache_use_cookie() to increment it - if it is not zero. If it was zero,
then access is not permitted.
(3) When the cache has finished with the data, it calls fscache_unuse_cookie()
to decrement it. This does a wake-up on it if it reaches 0.
(4) __fscache_relinquish_cookie() decrements n_active and then waits for it to
reach 0. The initialisation to 1 in step (1) ensures that we only get
wake ups when we're trying to get rid of the cookie.
This leaves __fscache_relinquish_cookie() a lot simpler.
***
This fixes a problem in the current code whereby if fscache_invalidate() is
followed sufficiently quickly by fscache_relinquish_cookie() then it is
possible for __fscache_relinquish_cookie() to have detached the cookie from the
object and cleared the pointer before a thread is dispatched to process the
invalidation state in the object state machine.
Since the pending write clearance was deferred to the invalidation state to
make it asynchronous, we need to either wait in relinquishment for the stores
tree to be cleared in the invalidation state or we need to handle the clearance
in relinquishment.
Further, if the relinquishment code does clear the tree, then the invalidation
state need to make the clearance contingent on still having the cookie to hand
(since that's where the tree is rooted) and we have to prevent the cookie from
disappearing for the duration.
This can lead to an oops like the following:
BUG: unable to handle kernel NULL pointer dereference at 000000000000000c
...
RIP: 0010:[<ffffffff8151023e>] _spin_lock+0xe/0x30
...
CR2: 000000000000000c ...
...
Process kslowd002 (...)
....
Call Trace:
[<ffffffffa01c3278>] fscache_invalidate_writes+0x38/0xd0 [fscache]
[<ffffffff810096f0>] ? __switch_to+0xd0/0x320
[<ffffffff8105e759>] ? find_busiest_queue+0x69/0x150
[<ffffffff8110ddd4>] ? slow_work_enqueue+0x104/0x180
[<ffffffffa01c1303>] fscache_object_slow_work_execute+0x5e3/0x9d0 [fscache]
[<ffffffff81096b67>] ? bit_waitqueue+0x17/0xd0
[<ffffffff8110e233>] slow_work_execute+0x233/0x310
[<ffffffff8110e515>] slow_work_thread+0x205/0x360
[<ffffffff81096ca0>] ? autoremove_wake_function+0x0/0x40
[<ffffffff8110e310>] ? slow_work_thread+0x0/0x360
[<ffffffff81096936>] kthread+0x96/0xa0
[<ffffffff8100c0ca>] child_rip+0xa/0x20
[<ffffffff810968a0>] ? kthread+0x0/0xa0
[<ffffffff8100c0c0>] ? child_rip+0x0/0x20
The parameter to fscache_invalidate_writes() was object->cookie which is NULL.
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
extern void fscache_object_destroy(struct fscache_object *);
|
2009-04-03 22:42:36 +07:00
|
|
|
|
|
|
|
extern void fscache_object_lookup_negative(struct fscache_object *object);
|
|
|
|
extern void fscache_obtained_object(struct fscache_object *object);
|
|
|
|
|
2013-05-11 01:50:26 +07:00
|
|
|
static inline bool fscache_object_is_live(struct fscache_object *object)
|
|
|
|
{
|
FS-Cache: Fix object state machine to have separate work and wait states
Fix object state machine to have separate work and wait states as that makes
it easier to envision.
There are now three kinds of state:
(1) Work state. This is an execution state. No event processing is performed
by a work state. The function attached to a work state returns a pointer
indicating the next state to which the OSM should transition. Returning
NO_TRANSIT repeats the current state, but goes back to the scheduler
first.
(2) Wait state. This is an event processing state. No execution is
performed by a wait state. Wait states are just tables of "if event X
occurs, clear it and transition to state Y". The dispatcher returns to
the scheduler if none of the events in which the wait state has an
interest are currently pending.
(3) Out-of-band state. This is a special work state. Transitions to normal
states can be overridden when an unexpected event occurs (eg. I/O error).
Instead the dispatcher disables and clears the OOB event and transits to
the specified work state. This then acts as an ordinary work state,
though object->state points to the overridden destination. Returning
NO_TRANSIT resumes the overridden transition.
In addition, the states have names in their definitions, so there's no need for
tables of state names. Further, the EV_REQUEUE event is no longer necessary as
that is automatic for work states.
Since the states are now separate structs rather than values in an enum, it's
not possible to use comparisons other than (non-)equality between them, so use
some object->flags to indicate what phase an object is in.
The EV_RELEASE, EV_RETIRE and EV_WITHDRAW events have been squished into one
(EV_KILL). An object flag now carries the information about retirement.
Similarly, the RELEASING, RECYCLING and WITHDRAWING states have been merged
into an KILL_OBJECT state and additional states have been added for handling
waiting dependent objects (JUMPSTART_DEPS and KILL_DEPENDENTS).
A state has also been added for synchronising with parent object initialisation
(WAIT_FOR_PARENT) and another for initiating look up (PARENT_READY).
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
return test_bit(FSCACHE_OBJECT_IS_LIVE, &object->flags);
|
2013-05-11 01:50:26 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline bool fscache_object_is_dying(struct fscache_object *object)
|
|
|
|
{
|
|
|
|
return !fscache_object_is_live(object);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline bool fscache_object_is_available(struct fscache_object *object)
|
|
|
|
{
|
FS-Cache: Fix object state machine to have separate work and wait states
Fix object state machine to have separate work and wait states as that makes
it easier to envision.
There are now three kinds of state:
(1) Work state. This is an execution state. No event processing is performed
by a work state. The function attached to a work state returns a pointer
indicating the next state to which the OSM should transition. Returning
NO_TRANSIT repeats the current state, but goes back to the scheduler
first.
(2) Wait state. This is an event processing state. No execution is
performed by a wait state. Wait states are just tables of "if event X
occurs, clear it and transition to state Y". The dispatcher returns to
the scheduler if none of the events in which the wait state has an
interest are currently pending.
(3) Out-of-band state. This is a special work state. Transitions to normal
states can be overridden when an unexpected event occurs (eg. I/O error).
Instead the dispatcher disables and clears the OOB event and transits to
the specified work state. This then acts as an ordinary work state,
though object->state points to the overridden destination. Returning
NO_TRANSIT resumes the overridden transition.
In addition, the states have names in their definitions, so there's no need for
tables of state names. Further, the EV_REQUEUE event is no longer necessary as
that is automatic for work states.
Since the states are now separate structs rather than values in an enum, it's
not possible to use comparisons other than (non-)equality between them, so use
some object->flags to indicate what phase an object is in.
The EV_RELEASE, EV_RETIRE and EV_WITHDRAW events have been squished into one
(EV_KILL). An object flag now carries the information about retirement.
Similarly, the RELEASING, RECYCLING and WITHDRAWING states have been merged
into an KILL_OBJECT state and additional states have been added for handling
waiting dependent objects (JUMPSTART_DEPS and KILL_DEPENDENTS).
A state has also been added for synchronising with parent object initialisation
(WAIT_FOR_PARENT) and another for initiating look up (PARENT_READY).
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
return test_bit(FSCACHE_OBJECT_IS_AVAILABLE, &object->flags);
|
2013-05-11 01:50:26 +07:00
|
|
|
}
|
|
|
|
|
2015-02-24 17:05:27 +07:00
|
|
|
static inline bool fscache_cache_is_broken(struct fscache_object *object)
|
|
|
|
{
|
|
|
|
return test_bit(FSCACHE_IOERROR, &object->cache->flags);
|
|
|
|
}
|
|
|
|
|
2013-05-11 01:50:26 +07:00
|
|
|
static inline bool fscache_object_is_active(struct fscache_object *object)
|
|
|
|
{
|
|
|
|
return fscache_object_is_available(object) &&
|
|
|
|
fscache_object_is_live(object) &&
|
2015-02-24 17:05:27 +07:00
|
|
|
!fscache_cache_is_broken(object);
|
2013-05-11 01:50:26 +07:00
|
|
|
}
|
|
|
|
|
2009-04-03 22:42:36 +07:00
|
|
|
/**
|
|
|
|
* fscache_object_destroyed - Note destruction of an object in a cache
|
|
|
|
* @cache: The cache from which the object came
|
|
|
|
*
|
|
|
|
* Note the destruction and deallocation of an object record in a cache.
|
|
|
|
*/
|
|
|
|
static inline void fscache_object_destroyed(struct fscache_cache *cache)
|
|
|
|
{
|
|
|
|
if (atomic_dec_and_test(&cache->object_count))
|
|
|
|
wake_up_all(&fscache_cache_cleared_wq);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* fscache_object_lookup_error - Note an object encountered an error
|
|
|
|
* @object: The object on which the error was encountered
|
|
|
|
*
|
|
|
|
* Note that an object encountered a fatal error (usually an I/O error) and
|
|
|
|
* that it should be withdrawn as soon as possible.
|
|
|
|
*/
|
|
|
|
static inline void fscache_object_lookup_error(struct fscache_object *object)
|
|
|
|
{
|
|
|
|
set_bit(FSCACHE_OBJECT_EV_ERROR, &object->events);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* fscache_set_store_limit - Set the maximum size to be stored in an object
|
|
|
|
* @object: The object to set the maximum on
|
|
|
|
* @i_size: The limit to set in bytes
|
|
|
|
*
|
|
|
|
* Set the maximum size an object is permitted to reach, implying the highest
|
|
|
|
* byte that may be written. Intended to be called by the attr_changed() op.
|
|
|
|
*
|
|
|
|
* See Documentation/filesystems/caching/backend-api.txt for a complete
|
|
|
|
* description.
|
|
|
|
*/
|
|
|
|
static inline
|
|
|
|
void fscache_set_store_limit(struct fscache_object *object, loff_t i_size)
|
|
|
|
{
|
2009-11-20 01:11:52 +07:00
|
|
|
object->store_limit_l = i_size;
|
2009-04-03 22:42:36 +07:00
|
|
|
object->store_limit = i_size >> PAGE_SHIFT;
|
|
|
|
if (i_size & ~PAGE_MASK)
|
|
|
|
object->store_limit++;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* fscache_end_io - End a retrieval operation on a page
|
|
|
|
* @op: The FS-Cache operation covering the retrieval
|
|
|
|
* @page: The page that was to be fetched
|
|
|
|
* @error: The error code (0 if successful)
|
|
|
|
*
|
|
|
|
* Note the end of an operation to retrieve a page, as covered by a particular
|
|
|
|
* operation record.
|
|
|
|
*/
|
|
|
|
static inline void fscache_end_io(struct fscache_retrieval *op,
|
|
|
|
struct page *page, int error)
|
|
|
|
{
|
|
|
|
op->end_io_func(page, op->context, error);
|
|
|
|
}
|
|
|
|
|
2013-09-21 06:09:31 +07:00
|
|
|
static inline void __fscache_use_cookie(struct fscache_cookie *cookie)
|
|
|
|
{
|
|
|
|
atomic_inc(&cookie->n_active);
|
|
|
|
}
|
|
|
|
|
FS-Cache: Simplify cookie retention for fscache_objects, fixing oops
Simplify the way fscache cache objects retain their cookie. The way I
implemented the cookie storage handling made synchronisation a pain (ie. the
object state machine can't rely on the cookie actually still being there).
Instead of the the object being detached from the cookie and the cookie being
freed in __fscache_relinquish_cookie(), we defer both operations:
(*) The detachment of the object from the list in the cookie now takes place
in fscache_drop_object() and is thus governed by the object state machine
(fscache_detach_from_cookie() has been removed).
(*) The release of the cookie is now in fscache_object_destroy() - which is
called by the cache backend just before it frees the object.
This means that the fscache_cookie struct is now available to the cache all the
way through from ->alloc_object() to ->drop_object() and ->put_object() -
meaning that it's no longer necessary to take object->lock to guarantee access.
However, __fscache_relinquish_cookie() doesn't wait for the object to go all
the way through to destruction before letting the netfs proceed. That would
massively slow down the netfs. Since __fscache_relinquish_cookie() leaves the
cookie around, in must therefore break all attachments to the netfs - which
includes ->def, ->netfs_data and any outstanding page read/writes.
To handle this, struct fscache_cookie now has an n_active counter:
(1) This starts off initialised to 1.
(2) Any time the cache needs to get at the netfs data, it calls
fscache_use_cookie() to increment it - if it is not zero. If it was zero,
then access is not permitted.
(3) When the cache has finished with the data, it calls fscache_unuse_cookie()
to decrement it. This does a wake-up on it if it reaches 0.
(4) __fscache_relinquish_cookie() decrements n_active and then waits for it to
reach 0. The initialisation to 1 in step (1) ensures that we only get
wake ups when we're trying to get rid of the cookie.
This leaves __fscache_relinquish_cookie() a lot simpler.
***
This fixes a problem in the current code whereby if fscache_invalidate() is
followed sufficiently quickly by fscache_relinquish_cookie() then it is
possible for __fscache_relinquish_cookie() to have detached the cookie from the
object and cleared the pointer before a thread is dispatched to process the
invalidation state in the object state machine.
Since the pending write clearance was deferred to the invalidation state to
make it asynchronous, we need to either wait in relinquishment for the stores
tree to be cleared in the invalidation state or we need to handle the clearance
in relinquishment.
Further, if the relinquishment code does clear the tree, then the invalidation
state need to make the clearance contingent on still having the cookie to hand
(since that's where the tree is rooted) and we have to prevent the cookie from
disappearing for the duration.
This can lead to an oops like the following:
BUG: unable to handle kernel NULL pointer dereference at 000000000000000c
...
RIP: 0010:[<ffffffff8151023e>] _spin_lock+0xe/0x30
...
CR2: 000000000000000c ...
...
Process kslowd002 (...)
....
Call Trace:
[<ffffffffa01c3278>] fscache_invalidate_writes+0x38/0xd0 [fscache]
[<ffffffff810096f0>] ? __switch_to+0xd0/0x320
[<ffffffff8105e759>] ? find_busiest_queue+0x69/0x150
[<ffffffff8110ddd4>] ? slow_work_enqueue+0x104/0x180
[<ffffffffa01c1303>] fscache_object_slow_work_execute+0x5e3/0x9d0 [fscache]
[<ffffffff81096b67>] ? bit_waitqueue+0x17/0xd0
[<ffffffff8110e233>] slow_work_execute+0x233/0x310
[<ffffffff8110e515>] slow_work_thread+0x205/0x360
[<ffffffff81096ca0>] ? autoremove_wake_function+0x0/0x40
[<ffffffff8110e310>] ? slow_work_thread+0x0/0x360
[<ffffffff81096936>] kthread+0x96/0xa0
[<ffffffff8100c0ca>] child_rip+0xa/0x20
[<ffffffff810968a0>] ? kthread+0x0/0xa0
[<ffffffff8100c0c0>] ? child_rip+0x0/0x20
The parameter to fscache_invalidate_writes() was object->cookie which is NULL.
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
/**
|
|
|
|
* fscache_use_cookie - Request usage of cookie attached to an object
|
|
|
|
* @object: Object description
|
|
|
|
*
|
|
|
|
* Request usage of the cookie attached to an object. NULL is returned if the
|
|
|
|
* relinquishment had reduced the cookie usage count to 0.
|
|
|
|
*/
|
|
|
|
static inline bool fscache_use_cookie(struct fscache_object *object)
|
|
|
|
{
|
|
|
|
struct fscache_cookie *cookie = object->cookie;
|
|
|
|
return atomic_inc_not_zero(&cookie->n_active) != 0;
|
|
|
|
}
|
|
|
|
|
2013-09-21 06:09:31 +07:00
|
|
|
static inline bool __fscache_unuse_cookie(struct fscache_cookie *cookie)
|
|
|
|
{
|
|
|
|
return atomic_dec_and_test(&cookie->n_active);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void __fscache_wake_unused_cookie(struct fscache_cookie *cookie)
|
|
|
|
{
|
2018-03-15 17:43:43 +07:00
|
|
|
wake_up_var(&cookie->n_active);
|
2013-09-21 06:09:31 +07:00
|
|
|
}
|
|
|
|
|
FS-Cache: Simplify cookie retention for fscache_objects, fixing oops
Simplify the way fscache cache objects retain their cookie. The way I
implemented the cookie storage handling made synchronisation a pain (ie. the
object state machine can't rely on the cookie actually still being there).
Instead of the the object being detached from the cookie and the cookie being
freed in __fscache_relinquish_cookie(), we defer both operations:
(*) The detachment of the object from the list in the cookie now takes place
in fscache_drop_object() and is thus governed by the object state machine
(fscache_detach_from_cookie() has been removed).
(*) The release of the cookie is now in fscache_object_destroy() - which is
called by the cache backend just before it frees the object.
This means that the fscache_cookie struct is now available to the cache all the
way through from ->alloc_object() to ->drop_object() and ->put_object() -
meaning that it's no longer necessary to take object->lock to guarantee access.
However, __fscache_relinquish_cookie() doesn't wait for the object to go all
the way through to destruction before letting the netfs proceed. That would
massively slow down the netfs. Since __fscache_relinquish_cookie() leaves the
cookie around, in must therefore break all attachments to the netfs - which
includes ->def, ->netfs_data and any outstanding page read/writes.
To handle this, struct fscache_cookie now has an n_active counter:
(1) This starts off initialised to 1.
(2) Any time the cache needs to get at the netfs data, it calls
fscache_use_cookie() to increment it - if it is not zero. If it was zero,
then access is not permitted.
(3) When the cache has finished with the data, it calls fscache_unuse_cookie()
to decrement it. This does a wake-up on it if it reaches 0.
(4) __fscache_relinquish_cookie() decrements n_active and then waits for it to
reach 0. The initialisation to 1 in step (1) ensures that we only get
wake ups when we're trying to get rid of the cookie.
This leaves __fscache_relinquish_cookie() a lot simpler.
***
This fixes a problem in the current code whereby if fscache_invalidate() is
followed sufficiently quickly by fscache_relinquish_cookie() then it is
possible for __fscache_relinquish_cookie() to have detached the cookie from the
object and cleared the pointer before a thread is dispatched to process the
invalidation state in the object state machine.
Since the pending write clearance was deferred to the invalidation state to
make it asynchronous, we need to either wait in relinquishment for the stores
tree to be cleared in the invalidation state or we need to handle the clearance
in relinquishment.
Further, if the relinquishment code does clear the tree, then the invalidation
state need to make the clearance contingent on still having the cookie to hand
(since that's where the tree is rooted) and we have to prevent the cookie from
disappearing for the duration.
This can lead to an oops like the following:
BUG: unable to handle kernel NULL pointer dereference at 000000000000000c
...
RIP: 0010:[<ffffffff8151023e>] _spin_lock+0xe/0x30
...
CR2: 000000000000000c ...
...
Process kslowd002 (...)
....
Call Trace:
[<ffffffffa01c3278>] fscache_invalidate_writes+0x38/0xd0 [fscache]
[<ffffffff810096f0>] ? __switch_to+0xd0/0x320
[<ffffffff8105e759>] ? find_busiest_queue+0x69/0x150
[<ffffffff8110ddd4>] ? slow_work_enqueue+0x104/0x180
[<ffffffffa01c1303>] fscache_object_slow_work_execute+0x5e3/0x9d0 [fscache]
[<ffffffff81096b67>] ? bit_waitqueue+0x17/0xd0
[<ffffffff8110e233>] slow_work_execute+0x233/0x310
[<ffffffff8110e515>] slow_work_thread+0x205/0x360
[<ffffffff81096ca0>] ? autoremove_wake_function+0x0/0x40
[<ffffffff8110e310>] ? slow_work_thread+0x0/0x360
[<ffffffff81096936>] kthread+0x96/0xa0
[<ffffffff8100c0ca>] child_rip+0xa/0x20
[<ffffffff810968a0>] ? kthread+0x0/0xa0
[<ffffffff8100c0c0>] ? child_rip+0x0/0x20
The parameter to fscache_invalidate_writes() was object->cookie which is NULL.
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
/**
|
|
|
|
* fscache_unuse_cookie - Cease usage of cookie attached to an object
|
|
|
|
* @object: Object description
|
|
|
|
*
|
|
|
|
* Cease usage of the cookie attached to an object. When the users count
|
|
|
|
* reaches zero then the cookie relinquishment will be permitted to proceed.
|
|
|
|
*/
|
|
|
|
static inline void fscache_unuse_cookie(struct fscache_object *object)
|
|
|
|
{
|
|
|
|
struct fscache_cookie *cookie = object->cookie;
|
2013-09-21 06:09:31 +07:00
|
|
|
if (__fscache_unuse_cookie(cookie))
|
|
|
|
__fscache_wake_unused_cookie(cookie);
|
FS-Cache: Simplify cookie retention for fscache_objects, fixing oops
Simplify the way fscache cache objects retain their cookie. The way I
implemented the cookie storage handling made synchronisation a pain (ie. the
object state machine can't rely on the cookie actually still being there).
Instead of the the object being detached from the cookie and the cookie being
freed in __fscache_relinquish_cookie(), we defer both operations:
(*) The detachment of the object from the list in the cookie now takes place
in fscache_drop_object() and is thus governed by the object state machine
(fscache_detach_from_cookie() has been removed).
(*) The release of the cookie is now in fscache_object_destroy() - which is
called by the cache backend just before it frees the object.
This means that the fscache_cookie struct is now available to the cache all the
way through from ->alloc_object() to ->drop_object() and ->put_object() -
meaning that it's no longer necessary to take object->lock to guarantee access.
However, __fscache_relinquish_cookie() doesn't wait for the object to go all
the way through to destruction before letting the netfs proceed. That would
massively slow down the netfs. Since __fscache_relinquish_cookie() leaves the
cookie around, in must therefore break all attachments to the netfs - which
includes ->def, ->netfs_data and any outstanding page read/writes.
To handle this, struct fscache_cookie now has an n_active counter:
(1) This starts off initialised to 1.
(2) Any time the cache needs to get at the netfs data, it calls
fscache_use_cookie() to increment it - if it is not zero. If it was zero,
then access is not permitted.
(3) When the cache has finished with the data, it calls fscache_unuse_cookie()
to decrement it. This does a wake-up on it if it reaches 0.
(4) __fscache_relinquish_cookie() decrements n_active and then waits for it to
reach 0. The initialisation to 1 in step (1) ensures that we only get
wake ups when we're trying to get rid of the cookie.
This leaves __fscache_relinquish_cookie() a lot simpler.
***
This fixes a problem in the current code whereby if fscache_invalidate() is
followed sufficiently quickly by fscache_relinquish_cookie() then it is
possible for __fscache_relinquish_cookie() to have detached the cookie from the
object and cleared the pointer before a thread is dispatched to process the
invalidation state in the object state machine.
Since the pending write clearance was deferred to the invalidation state to
make it asynchronous, we need to either wait in relinquishment for the stores
tree to be cleared in the invalidation state or we need to handle the clearance
in relinquishment.
Further, if the relinquishment code does clear the tree, then the invalidation
state need to make the clearance contingent on still having the cookie to hand
(since that's where the tree is rooted) and we have to prevent the cookie from
disappearing for the duration.
This can lead to an oops like the following:
BUG: unable to handle kernel NULL pointer dereference at 000000000000000c
...
RIP: 0010:[<ffffffff8151023e>] _spin_lock+0xe/0x30
...
CR2: 000000000000000c ...
...
Process kslowd002 (...)
....
Call Trace:
[<ffffffffa01c3278>] fscache_invalidate_writes+0x38/0xd0 [fscache]
[<ffffffff810096f0>] ? __switch_to+0xd0/0x320
[<ffffffff8105e759>] ? find_busiest_queue+0x69/0x150
[<ffffffff8110ddd4>] ? slow_work_enqueue+0x104/0x180
[<ffffffffa01c1303>] fscache_object_slow_work_execute+0x5e3/0x9d0 [fscache]
[<ffffffff81096b67>] ? bit_waitqueue+0x17/0xd0
[<ffffffff8110e233>] slow_work_execute+0x233/0x310
[<ffffffff8110e515>] slow_work_thread+0x205/0x360
[<ffffffff81096ca0>] ? autoremove_wake_function+0x0/0x40
[<ffffffff8110e310>] ? slow_work_thread+0x0/0x360
[<ffffffff81096936>] kthread+0x96/0xa0
[<ffffffff8100c0ca>] child_rip+0xa/0x20
[<ffffffff810968a0>] ? kthread+0x0/0xa0
[<ffffffff8100c0c0>] ? child_rip+0x0/0x20
The parameter to fscache_invalidate_writes() was object->cookie which is NULL.
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-By: Milosz Tanski <milosz@adfin.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
2013-05-11 01:50:26 +07:00
|
|
|
}
|
|
|
|
|
2009-04-03 22:42:36 +07:00
|
|
|
/*
|
|
|
|
* out-of-line cache backend functions
|
|
|
|
*/
|
2011-11-01 07:11:33 +07:00
|
|
|
extern __printf(3, 4)
|
|
|
|
void fscache_init_cache(struct fscache_cache *cache,
|
|
|
|
const struct fscache_cache_ops *ops,
|
|
|
|
const char *idfmt, ...);
|
2009-04-03 22:42:36 +07:00
|
|
|
|
|
|
|
extern int fscache_add_cache(struct fscache_cache *cache,
|
|
|
|
struct fscache_object *fsdef,
|
|
|
|
const char *tagname);
|
|
|
|
extern void fscache_withdraw_cache(struct fscache_cache *cache);
|
|
|
|
|
|
|
|
extern void fscache_io_error(struct fscache_cache *cache);
|
|
|
|
|
2012-12-21 04:52:32 +07:00
|
|
|
extern void fscache_mark_page_cached(struct fscache_retrieval *op,
|
|
|
|
struct page *page);
|
|
|
|
|
2009-04-03 22:42:36 +07:00
|
|
|
extern void fscache_mark_pages_cached(struct fscache_retrieval *op,
|
|
|
|
struct pagevec *pagevec);
|
|
|
|
|
2010-07-21 03:09:01 +07:00
|
|
|
extern bool fscache_object_sleep_till_congested(signed long *timeoutp);
|
|
|
|
|
2009-04-03 22:42:36 +07:00
|
|
|
extern enum fscache_checkaux fscache_check_aux(struct fscache_object *object,
|
|
|
|
const void *data,
|
2018-04-04 19:41:28 +07:00
|
|
|
uint16_t datalen,
|
|
|
|
loff_t object_size);
|
2009-04-03 22:42:36 +07:00
|
|
|
|
2015-02-20 06:47:31 +07:00
|
|
|
extern void fscache_object_retrying_stale(struct fscache_object *object);
|
|
|
|
|
|
|
|
enum fscache_why_object_killed {
|
|
|
|
FSCACHE_OBJECT_IS_STALE,
|
|
|
|
FSCACHE_OBJECT_NO_SPACE,
|
|
|
|
FSCACHE_OBJECT_WAS_RETIRED,
|
|
|
|
FSCACHE_OBJECT_WAS_CULLED,
|
|
|
|
};
|
|
|
|
extern void fscache_object_mark_killed(struct fscache_object *object,
|
|
|
|
enum fscache_why_object_killed why);
|
|
|
|
|
2009-04-03 22:42:36 +07:00
|
|
|
#endif /* _LINUX_FSCACHE_CACHE_H */
|