OpenCloudOS-Kernel/include/linux/iocontext.h

153 lines
4.5 KiB
C
Raw Normal View History

#ifndef IOCONTEXT_H
#define IOCONTEXT_H
#include <linux/radix-tree.h>
#include <linux/rcupdate.h>
block, cfq: unlink cfq_io_context's immediately cic is association between io_context and request_queue. A cic is linked from both ioc and q and should be destroyed when either one goes away. As ioc and q both have their own locks, locking becomes a bit complex - both orders work for removal from one but not from the other. Currently, cfq tries to circumvent this locking order issue with RCU. ioc->lock nests inside queue_lock but the radix tree and cic's are also protected by RCU allowing either side to walk their lists without grabbing lock. This rather unconventional use of RCU quickly devolves into extremely fragile convolution. e.g. The following is from cfqd going away too soon after ioc and q exits raced. general protection fault: 0000 [#1] PREEMPT SMP CPU 2 Modules linked in: [ 88.503444] Pid: 599, comm: hexdump Not tainted 3.1.0-rc10-work+ #158 Bochs Bochs RIP: 0010:[<ffffffff81397628>] [<ffffffff81397628>] cfq_exit_single_io_context+0x58/0xf0 ... Call Trace: [<ffffffff81395a4a>] call_for_each_cic+0x5a/0x90 [<ffffffff81395ab5>] cfq_exit_io_context+0x15/0x20 [<ffffffff81389130>] exit_io_context+0x100/0x140 [<ffffffff81098a29>] do_exit+0x579/0x850 [<ffffffff81098d5b>] do_group_exit+0x5b/0xd0 [<ffffffff81098de7>] sys_exit_group+0x17/0x20 [<ffffffff81b02f2b>] system_call_fastpath+0x16/0x1b The only real hot path here is cic lookup during request initialization and avoiding extra locking requires very confined use of RCU. This patch makes cic removal from both ioc and request_queue perform double-locking and unlink immediately. * From q side, the change is almost trivial as ioc->lock nests inside queue_lock. It just needs to grab each ioc->lock as it walks cic_list and unlink it. * From ioc side, it's a bit more difficult because of inversed lock order. ioc needs its lock to walk its cic_list but can't grab the matching queue_lock and needs to perform unlock-relock dancing. Unlinking is now wholly done from put_io_context() and fast path is optimized by using the queue_lock the caller already holds, which is by far the most common case. If the ioc accessed multiple devices, it tries with trylock. In unlikely cases of fast path failure, it falls back to full double-locking dance from workqueue. Double-locking isn't the prettiest thing in the world but it's *far* simpler and more understandable than RCU trick without adding any meaningful overhead. This still leaves a lot of now unnecessary RCU logics. Future patches will trim them. -v2: Vivek pointed out that cic->q was being dereferenced after cic->release() was called. Updated to use local variable @this_q instead. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Vivek Goyal <vgoyal@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2011-12-14 07:33:39 +08:00
#include <linux/workqueue.h>
enum {
ICQ_IOPRIO_CHANGED = 1 << 0,
ICQ_CGROUP_CHANGED = 1 << 1,
block: exit_io_context() should call elevator_exit_icq_fn() While updating locking, b2efa05265 "block, cfq: unlink cfq_io_context's immediately" moved elevator_exit_icq_fn() invocation from exit_io_context() to the final ioc put. While this doesn't cause catastrophic failure, it effectively removes task exit notification to elevator and cause noticeable IO performance degradation with CFQ. On task exit, CFQ used to immediately expire the slice if it was being used by the exiting task as no more IO would be issued by the task; however, after b2efa05265, the notification is lost and disk could sit idle needlessly, leading to noticeable IO performance degradation for certain workloads. This patch renames ioc_exit_icq() to ioc_destroy_icq(), separates elevator_exit_icq_fn() invocation into ioc_exit_icq() and invokes it from exit_io_context(). ICQ_EXITED flag is added to avoid invoking the callback more than once for the same icq. Walking icq_list from ioc side and invoking elevator callback requires reverse double locking. This may be better implemented using RCU; unfortunately, using RCU isn't trivial. e.g. RCU protection would need to cover request_queue and queue_lock switch on cleanup makes grabbing queue_lock from RCU unsafe. Reverse double locking should do, at least for now. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-and-bisected-by: Shaohua Li <shli@kernel.org> LKML-Reference: <CANejiEVzs=pUhQSTvUppkDcc2TNZyfohBRLygW5zFmXyk5A-xQ@mail.gmail.com> Tested-by: Shaohua Li <shaohua.li@intel.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2012-02-15 16:45:53 +08:00
ICQ_EXITED = 1 << 2,
ICQ_CHANGED_MASK = ICQ_IOPRIO_CHANGED | ICQ_CGROUP_CHANGED,
};
block, cfq: move icq creation and rq->elv.icq association to block core Now block layer knows everything necessary to create and associate icq's with requests. Move ioc_create_icq() to blk-ioc.c and update get_request() such that, if elevator_type->icq_size is set, requests are automatically associated with their matching icq's before elv_set_request(). io_context reference is also managed by block core on request alloc/free. * Only ioprio/cgroup changed handling remains from cfq_get_cic(). Collapsed into cfq_set_request(). * This removes queue kicking on icq allocation failure (for now). As icq allocation failure is rare and the only effect of queue kicking achieved was possibily accelerating queue processing, this change shouldn't be noticeable. There is a larger underlying problem. Unlike request allocation, icq allocation is not guaranteed to succeed eventually after retries. The number of icq is unbound and thus mempool can't be the solution either. This effectively adds allocation dependency on memory free path and thus possibility of deadlock. This usually wouldn't happen because icq allocation is not a hot path and, even when the condition triggers, it's highly unlikely that none of the writeback workers already has icq. However, this is still possible especially if elevator is being switched under high memory pressure, so we better get it fixed. Probably the only solution is just bypassing elevator and appending to dispatch queue on any elevator allocation failure. * Comment added to explain how icq's are managed and synchronized. This completes cleanup of io_context interface. Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2011-12-14 07:33:42 +08:00
/*
* An io_cq (icq) is association between an io_context (ioc) and a
* request_queue (q). This is used by elevators which need to track
* information per ioc - q pair.
*
* Elevator can request use of icq by setting elevator_type->icq_size and
* ->icq_align. Both size and align must be larger than that of struct
* io_cq and elevator can use the tail area for private information. The
* recommended way to do this is defining a struct which contains io_cq as
* the first member followed by private members and using its size and
* align. For example,
*
* struct snail_io_cq {
* struct io_cq icq;
* int poke_snail;
* int feed_snail;
* };
*
* struct elevator_type snail_elv_type {
* .ops = { ... },
* .icq_size = sizeof(struct snail_io_cq),
* .icq_align = __alignof__(struct snail_io_cq),
* ...
* };
*
* If icq_size is set, block core will manage icq's. All requests will
* have its ->elv.icq field set before elevator_ops->elevator_set_req_fn()
* is called and be holding a reference to the associated io_context.
*
* Whenever a new icq is created, elevator_ops->elevator_init_icq_fn() is
* called and, on destruction, ->elevator_exit_icq_fn(). Both functions
* are called with both the associated io_context and queue locks held.
*
* Elevator is allowed to lookup icq using ioc_lookup_icq() while holding
* queue lock but the returned icq is valid only until the queue lock is
* released. Elevators can not and should not try to create or destroy
* icq's.
*
* As icq's are linked from both ioc and q, the locking rules are a bit
* complex.
*
* - ioc lock nests inside q lock.
*
* - ioc->icq_list and icq->ioc_node are protected by ioc lock.
* q->icq_list and icq->q_node by q lock.
*
* - ioc->icq_tree and ioc->icq_hint are protected by ioc lock, while icq
* itself is protected by q lock. However, both the indexes and icq
* itself are also RCU managed and lookup can be performed holding only
* the q lock.
*
* - icq's are not reference counted. They are destroyed when either the
* ioc or q goes away. Each request with icq set holds an extra
* reference to ioc to ensure it stays until the request is completed.
*
* - Linking and unlinking icq's are performed while holding both ioc and q
* locks. Due to the lock ordering, q exit is simple but ioc exit
* requires reverse-order double lock dance.
*/
struct io_cq {
struct request_queue *q;
struct io_context *ioc;
/*
* q_node and ioc_node link io_cq through icq_list of q and ioc
* respectively. Both fields are unused once ioc_exit_icq() is
* called and shared with __rcu_icq_cache and __rcu_head which are
* used for RCU free of io_cq.
*/
union {
struct list_head q_node;
struct kmem_cache *__rcu_icq_cache;
};
union {
struct hlist_node ioc_node;
struct rcu_head __rcu_head;
};
unsigned int flags;
};
/*
* I/O subsystem state of the associated processes. It is refcounted
* and kmalloc'ed. These could be shared between processes.
*/
struct io_context {
atomic_long_t refcount;
atomic_t nr_tasks;
/* all the fields below are protected by this lock */
spinlock_t lock;
unsigned short ioprio;
/*
* For request batching
*/
int nr_batch_requests; /* Number of requests left in the batch */
unsigned long last_waited; /* Time last woken after wait for request */
struct radix_tree_root icq_tree;
struct io_cq __rcu *icq_hint;
struct hlist_head icq_list;
block, cfq: unlink cfq_io_context's immediately cic is association between io_context and request_queue. A cic is linked from both ioc and q and should be destroyed when either one goes away. As ioc and q both have their own locks, locking becomes a bit complex - both orders work for removal from one but not from the other. Currently, cfq tries to circumvent this locking order issue with RCU. ioc->lock nests inside queue_lock but the radix tree and cic's are also protected by RCU allowing either side to walk their lists without grabbing lock. This rather unconventional use of RCU quickly devolves into extremely fragile convolution. e.g. The following is from cfqd going away too soon after ioc and q exits raced. general protection fault: 0000 [#1] PREEMPT SMP CPU 2 Modules linked in: [ 88.503444] Pid: 599, comm: hexdump Not tainted 3.1.0-rc10-work+ #158 Bochs Bochs RIP: 0010:[<ffffffff81397628>] [<ffffffff81397628>] cfq_exit_single_io_context+0x58/0xf0 ... Call Trace: [<ffffffff81395a4a>] call_for_each_cic+0x5a/0x90 [<ffffffff81395ab5>] cfq_exit_io_context+0x15/0x20 [<ffffffff81389130>] exit_io_context+0x100/0x140 [<ffffffff81098a29>] do_exit+0x579/0x850 [<ffffffff81098d5b>] do_group_exit+0x5b/0xd0 [<ffffffff81098de7>] sys_exit_group+0x17/0x20 [<ffffffff81b02f2b>] system_call_fastpath+0x16/0x1b The only real hot path here is cic lookup during request initialization and avoiding extra locking requires very confined use of RCU. This patch makes cic removal from both ioc and request_queue perform double-locking and unlink immediately. * From q side, the change is almost trivial as ioc->lock nests inside queue_lock. It just needs to grab each ioc->lock as it walks cic_list and unlink it. * From ioc side, it's a bit more difficult because of inversed lock order. ioc needs its lock to walk its cic_list but can't grab the matching queue_lock and needs to perform unlock-relock dancing. Unlinking is now wholly done from put_io_context() and fast path is optimized by using the queue_lock the caller already holds, which is by far the most common case. If the ioc accessed multiple devices, it tries with trylock. In unlikely cases of fast path failure, it falls back to full double-locking dance from workqueue. Double-locking isn't the prettiest thing in the world but it's *far* simpler and more understandable than RCU trick without adding any meaningful overhead. This still leaves a lot of now unnecessary RCU logics. Future patches will trim them. -v2: Vivek pointed out that cic->q was being dereferenced after cic->release() was called. Updated to use local variable @this_q instead. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Vivek Goyal <vgoyal@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2011-12-14 07:33:39 +08:00
struct work_struct release_work;
};
static inline struct io_context *ioc_task_link(struct io_context *ioc)
{
/*
* if ref count is zero, don't allow sharing (ioc is going away, it's
* a race).
*/
if (ioc && atomic_long_inc_not_zero(&ioc->refcount)) {
atomic_inc(&ioc->nr_tasks);
return ioc;
}
return NULL;
}
struct task_struct;
#ifdef CONFIG_BLOCK
void put_io_context(struct io_context *ioc);
void exit_io_context(struct task_struct *task);
block: make ioc get/put interface more conventional and fix race on alloction Ignoring copy_io() during fork, io_context can be allocated from two places - current_io_context() and set_task_ioprio(). The former is always called from local task while the latter can be called from different task. The synchornization between them are peculiar and dubious. * current_io_context() doesn't grab task_lock() and assumes that if it saw %NULL ->io_context, it would stay that way until allocation and assignment is complete. It has smp_wmb() between alloc/init and assignment. * set_task_ioprio() grabs task_lock() for assignment and does smp_read_barrier_depends() between "ioc = task->io_context" and "if (ioc)". Unfortunately, this doesn't achieve anything - the latter is not a dependent load of the former. ie, if ioc itself were being dereferenced "ioc->xxx", it would mean something (not sure what tho) but as the code currently stands, the dependent read barrier is noop. As only one of the the two test-assignment sequences is task_lock() protected, the task_lock() can't do much about race between the two. Nothing prevents current_io_context() and set_task_ioprio() allocating its own ioc for the same task and overwriting the other's. Also, set_task_ioprio() can race with exiting task and create a new ioc after exit_io_context() is finished. ioc get/put doesn't have any reason to be complex. The only hot path is accessing the existing ioc of %current, which is simple to achieve given that ->io_context is never destroyed as long as the task is alive. All other paths can happily go through task_lock() like all other task sub structures without impacting anything. This patch updates ioc get/put so that it becomes more conventional. * alloc_io_context() is replaced with get_task_io_context(). This is the only interface which can acquire access to ioc of another task. On return, the caller has an explicit reference to the object which should be put using put_io_context() afterwards. * The functionality of current_io_context() remains the same but when creating a new ioc, it shares the code path with get_task_io_context() and always goes through task_lock(). * get_io_context() now means incrementing ref on an ioc which the caller already has access to (be that an explicit refcnt or implicit %current one). * PF_EXITING inhibits creation of new io_context and once exit_io_context() is finished, it's guaranteed that both ioc acquisition functions return %NULL. * All users are updated. Most are trivial but smp_read_barrier_depends() removal from cfq_get_io_context() needs a bit of explanation. I suppose the original intention was to ensure ioc->ioprio is visible when set_task_ioprio() allocates new io_context and installs it; however, this wouldn't have worked because set_task_ioprio() doesn't have wmb between init and install. There are other problems with this which will be fixed in another patch. * While at it, use NUMA_NO_NODE instead of -1 for wildcard node specification. -v2: Vivek spotted contamination from debug patch. Removed. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Vivek Goyal <vgoyal@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2011-12-14 07:33:38 +08:00
struct io_context *get_task_io_context(struct task_struct *task,
gfp_t gfp_flags, int node);
void ioc_ioprio_changed(struct io_context *ioc, int ioprio);
void ioc_cgroup_changed(struct io_context *ioc);
unsigned int icq_get_changed(struct io_cq *icq);
#else
struct io_context;
static inline void put_io_context(struct io_context *ioc) { }
static inline void exit_io_context(struct task_struct *task) { }
#endif
#endif