Thomas Gleixner | 457c899 | 2019-05-19 13:08:55 +0100 | [diff] [blame] | 1 | // SPDX-License-Identifier: GPL-2.0-only |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 2 | /* |
| 3 | * RT-Mutexes: simple blocking mutual exclusion locks with PI support |
| 4 | * |
| 5 | * started by Ingo Molnar and Thomas Gleixner. |
| 6 | * |
| 7 | * Copyright (C) 2004-2006 Red Hat, Inc., Ingo Molnar <mingo@redhat.com> |
| 8 | * Copyright (C) 2005-2006 Timesys Corp., Thomas Gleixner <tglx@timesys.com> |
| 9 | * Copyright (C) 2005 Kihon Technologies Inc., Steven Rostedt |
| 10 | * Copyright (C) 2006 Esben Nielsen |
Steven Rostedt | 992caf7 | 2021-08-15 23:29:25 +0200 | [diff] [blame] | 11 | * Adaptive Spinlocks: |
| 12 | * Copyright (C) 2008 Novell, Inc., Gregory Haskins, Sven Dietrich, |
| 13 | * and Peter Morreale, |
| 14 | * Adaptive Spinlocks simplification: |
| 15 | * Copyright (C) 2008 Red Hat, Inc., Steven Rostedt <srostedt@redhat.com> |
Steven Rostedt | d07fe82c2 | 2006-07-30 03:04:03 -0700 | [diff] [blame] | 16 | * |
Mauro Carvalho Chehab | 387b146 | 2019-04-10 08:32:41 -0300 | [diff] [blame] | 17 | * See Documentation/locking/rt-mutex-design.rst for details. |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 18 | */ |
Thomas Gleixner | 531ae4b | 2021-08-15 23:27:57 +0200 | [diff] [blame] | 19 | #include <linux/sched.h> |
| 20 | #include <linux/sched/debug.h> |
| 21 | #include <linux/sched/deadline.h> |
Ingo Molnar | 174cd4b | 2017-02-02 19:15:33 +0100 | [diff] [blame] | 22 | #include <linux/sched/signal.h> |
Clark Williams | 8bd75c7 | 2013-02-07 09:47:07 -0600 | [diff] [blame] | 23 | #include <linux/sched/rt.h> |
Ingo Molnar | 84f001e | 2017-02-01 16:36:40 +0100 | [diff] [blame] | 24 | #include <linux/sched/wake_q.h> |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 25 | #include <linux/ww_mutex.h> |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 26 | |
| 27 | #include "rtmutex_common.h" |
| 28 | |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 29 | #ifndef WW_RT |
| 30 | # define build_ww_mutex() (false) |
| 31 | # define ww_container_of(rtm) NULL |
| 32 | |
| 33 | static inline int __ww_mutex_add_waiter(struct rt_mutex_waiter *waiter, |
| 34 | struct rt_mutex *lock, |
| 35 | struct ww_acquire_ctx *ww_ctx) |
| 36 | { |
| 37 | return 0; |
| 38 | } |
| 39 | |
| 40 | static inline void __ww_mutex_check_waiters(struct rt_mutex *lock, |
| 41 | struct ww_acquire_ctx *ww_ctx) |
| 42 | { |
| 43 | } |
| 44 | |
| 45 | static inline void ww_mutex_lock_acquired(struct ww_mutex *lock, |
| 46 | struct ww_acquire_ctx *ww_ctx) |
| 47 | { |
| 48 | } |
| 49 | |
| 50 | static inline int __ww_mutex_check_kill(struct rt_mutex *lock, |
| 51 | struct rt_mutex_waiter *waiter, |
| 52 | struct ww_acquire_ctx *ww_ctx) |
| 53 | { |
| 54 | return 0; |
| 55 | } |
| 56 | |
| 57 | #else |
| 58 | # define build_ww_mutex() (true) |
| 59 | # define ww_container_of(rtm) container_of(rtm, struct ww_mutex, base) |
| 60 | # include "ww_mutex.h" |
| 61 | #endif |
| 62 | |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 63 | /* |
| 64 | * lock->owner state tracking: |
| 65 | * |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 66 | * lock->owner holds the task_struct pointer of the owner. Bit 0 |
| 67 | * is used to keep track of the "lock has waiters" state. |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 68 | * |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 69 | * owner bit0 |
| 70 | * NULL 0 lock is free (fast acquire possible) |
| 71 | * NULL 1 lock is free and has waiters and the top waiter |
| 72 | * is going to take the lock* |
| 73 | * taskpointer 0 lock is held (fast release possible) |
| 74 | * taskpointer 1 lock is held and has waiters** |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 75 | * |
| 76 | * The fast atomic compare exchange based acquire and release is only |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 77 | * possible when bit 0 of lock->owner is 0. |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 78 | * |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 79 | * (*) It also can be a transitional state when grabbing the lock |
| 80 | * with ->wait_lock is held. To prevent any fast path cmpxchg to the lock, |
| 81 | * we need to set the bit0 before looking at the lock, and the owner may be |
| 82 | * NULL in this small time, hence this can be a transitional state. |
| 83 | * |
| 84 | * (**) There is a small time when bit 0 is set but there are no |
| 85 | * waiters. This can happen when grabbing the lock in the slow path. |
| 86 | * To prevent a cmpxchg of the owner releasing the lock, we need to |
| 87 | * set this bit before looking at the lock. |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 88 | */ |
| 89 | |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 90 | static __always_inline void |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 91 | rt_mutex_set_owner(struct rt_mutex_base *lock, struct task_struct *owner) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 92 | { |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 93 | unsigned long val = (unsigned long)owner; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 94 | |
| 95 | if (rt_mutex_has_waiters(lock)) |
| 96 | val |= RT_MUTEX_HAS_WAITERS; |
| 97 | |
Paul E. McKenney | 0050c7b | 2020-01-03 15:59:12 -0800 | [diff] [blame] | 98 | WRITE_ONCE(lock->owner, (struct task_struct *)val); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 99 | } |
| 100 | |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 101 | static __always_inline void clear_rt_mutex_waiters(struct rt_mutex_base *lock) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 102 | { |
| 103 | lock->owner = (struct task_struct *) |
| 104 | ((unsigned long)lock->owner & ~RT_MUTEX_HAS_WAITERS); |
| 105 | } |
| 106 | |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 107 | static __always_inline void fixup_rt_mutex_waiters(struct rt_mutex_base *lock) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 108 | { |
Thomas Gleixner | dbb2605 | 2016-11-30 21:04:41 +0000 | [diff] [blame] | 109 | unsigned long owner, *p = (unsigned long *) &lock->owner; |
| 110 | |
| 111 | if (rt_mutex_has_waiters(lock)) |
| 112 | return; |
| 113 | |
| 114 | /* |
| 115 | * The rbtree has no waiters enqueued, now make sure that the |
| 116 | * lock->owner still has the waiters bit set, otherwise the |
| 117 | * following can happen: |
| 118 | * |
| 119 | * CPU 0 CPU 1 CPU2 |
| 120 | * l->owner=T1 |
| 121 | * rt_mutex_lock(l) |
| 122 | * lock(l->lock) |
| 123 | * l->owner = T1 | HAS_WAITERS; |
| 124 | * enqueue(T2) |
| 125 | * boost() |
| 126 | * unlock(l->lock) |
| 127 | * block() |
| 128 | * |
| 129 | * rt_mutex_lock(l) |
| 130 | * lock(l->lock) |
| 131 | * l->owner = T1 | HAS_WAITERS; |
| 132 | * enqueue(T3) |
| 133 | * boost() |
| 134 | * unlock(l->lock) |
| 135 | * block() |
| 136 | * signal(->T2) signal(->T3) |
| 137 | * lock(l->lock) |
| 138 | * dequeue(T2) |
| 139 | * deboost() |
| 140 | * unlock(l->lock) |
| 141 | * lock(l->lock) |
| 142 | * dequeue(T3) |
| 143 | * ==> wait list is empty |
| 144 | * deboost() |
| 145 | * unlock(l->lock) |
| 146 | * lock(l->lock) |
| 147 | * fixup_rt_mutex_waiters() |
| 148 | * if (wait_list_empty(l) { |
| 149 | * l->owner = owner |
| 150 | * owner = l->owner & ~HAS_WAITERS; |
| 151 | * ==> l->owner = T1 |
| 152 | * } |
| 153 | * lock(l->lock) |
| 154 | * rt_mutex_unlock(l) fixup_rt_mutex_waiters() |
| 155 | * if (wait_list_empty(l) { |
| 156 | * owner = l->owner & ~HAS_WAITERS; |
| 157 | * cmpxchg(l->owner, T1, NULL) |
| 158 | * ===> Success (l->owner = NULL) |
| 159 | * |
| 160 | * l->owner = owner |
| 161 | * ==> l->owner = T1 |
| 162 | * } |
| 163 | * |
| 164 | * With the check for the waiter bit in place T3 on CPU2 will not |
| 165 | * overwrite. All tasks fiddling with the waiters bit are |
| 166 | * serialized by l->lock, so nothing else can modify the waiters |
| 167 | * bit. If the bit is set then nothing can change l->owner either |
| 168 | * so the simple RMW is safe. The cmpxchg() will simply fail if it |
| 169 | * happens in the middle of the RMW because the waiters bit is |
| 170 | * still set. |
| 171 | */ |
| 172 | owner = READ_ONCE(*p); |
| 173 | if (owner & RT_MUTEX_HAS_WAITERS) |
| 174 | WRITE_ONCE(*p, owner & ~RT_MUTEX_HAS_WAITERS); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 175 | } |
| 176 | |
| 177 | /* |
Sebastian Andrzej Siewior | cede884 | 2015-02-25 18:56:13 +0100 | [diff] [blame] | 178 | * We can speed up the acquire/release, if there's no debugging state to be |
| 179 | * set up. |
Thomas Gleixner | bd19723 | 2007-06-17 21:11:10 +0200 | [diff] [blame] | 180 | */ |
Sebastian Andrzej Siewior | cede884 | 2015-02-25 18:56:13 +0100 | [diff] [blame] | 181 | #ifndef CONFIG_DEBUG_RT_MUTEXES |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 182 | static __always_inline bool rt_mutex_cmpxchg_acquire(struct rt_mutex_base *lock, |
Sebastian Andrzej Siewior | 7851593 | 2021-08-15 23:27:54 +0200 | [diff] [blame] | 183 | struct task_struct *old, |
| 184 | struct task_struct *new) |
| 185 | { |
Thomas Gleixner | 709e0b6 | 2021-08-15 23:27:55 +0200 | [diff] [blame] | 186 | return try_cmpxchg_acquire(&lock->owner, &old, new); |
Sebastian Andrzej Siewior | 7851593 | 2021-08-15 23:27:54 +0200 | [diff] [blame] | 187 | } |
| 188 | |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 189 | static __always_inline bool rt_mutex_cmpxchg_release(struct rt_mutex_base *lock, |
Sebastian Andrzej Siewior | 7851593 | 2021-08-15 23:27:54 +0200 | [diff] [blame] | 190 | struct task_struct *old, |
| 191 | struct task_struct *new) |
| 192 | { |
Thomas Gleixner | 709e0b6 | 2021-08-15 23:27:55 +0200 | [diff] [blame] | 193 | return try_cmpxchg_release(&lock->owner, &old, new); |
Sebastian Andrzej Siewior | 7851593 | 2021-08-15 23:27:54 +0200 | [diff] [blame] | 194 | } |
Davidlohr Bueso | 700318d | 2015-09-30 13:03:13 -0700 | [diff] [blame] | 195 | |
| 196 | /* |
| 197 | * Callers must hold the ->wait_lock -- which is the whole purpose as we force |
| 198 | * all future threads that attempt to [Rmw] the lock to the slowpath. As such |
| 199 | * relaxed semantics suffice. |
| 200 | */ |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 201 | static __always_inline void mark_rt_mutex_waiters(struct rt_mutex_base *lock) |
Thomas Gleixner | bd19723 | 2007-06-17 21:11:10 +0200 | [diff] [blame] | 202 | { |
| 203 | unsigned long owner, *p = (unsigned long *) &lock->owner; |
| 204 | |
| 205 | do { |
| 206 | owner = *p; |
Davidlohr Bueso | 700318d | 2015-09-30 13:03:13 -0700 | [diff] [blame] | 207 | } while (cmpxchg_relaxed(p, owner, |
| 208 | owner | RT_MUTEX_HAS_WAITERS) != owner); |
Thomas Gleixner | bd19723 | 2007-06-17 21:11:10 +0200 | [diff] [blame] | 209 | } |
Thomas Gleixner | 27e3571 | 2014-06-11 18:44:04 +0000 | [diff] [blame] | 210 | |
| 211 | /* |
| 212 | * Safe fastpath aware unlock: |
| 213 | * 1) Clear the waiters bit |
| 214 | * 2) Drop lock->wait_lock |
| 215 | * 3) Try to unlock the lock with cmpxchg |
| 216 | */ |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 217 | static __always_inline bool unlock_rt_mutex_safe(struct rt_mutex_base *lock, |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 218 | unsigned long flags) |
Thomas Gleixner | 27e3571 | 2014-06-11 18:44:04 +0000 | [diff] [blame] | 219 | __releases(lock->wait_lock) |
| 220 | { |
| 221 | struct task_struct *owner = rt_mutex_owner(lock); |
| 222 | |
| 223 | clear_rt_mutex_waiters(lock); |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 224 | raw_spin_unlock_irqrestore(&lock->wait_lock, flags); |
Thomas Gleixner | 27e3571 | 2014-06-11 18:44:04 +0000 | [diff] [blame] | 225 | /* |
| 226 | * If a new waiter comes in between the unlock and the cmpxchg |
| 227 | * we have two situations: |
| 228 | * |
| 229 | * unlock(wait_lock); |
| 230 | * lock(wait_lock); |
| 231 | * cmpxchg(p, owner, 0) == owner |
| 232 | * mark_rt_mutex_waiters(lock); |
| 233 | * acquire(lock); |
| 234 | * or: |
| 235 | * |
| 236 | * unlock(wait_lock); |
| 237 | * lock(wait_lock); |
| 238 | * mark_rt_mutex_waiters(lock); |
| 239 | * |
| 240 | * cmpxchg(p, owner, 0) != owner |
| 241 | * enqueue_waiter(); |
| 242 | * unlock(wait_lock); |
| 243 | * lock(wait_lock); |
| 244 | * wake waiter(); |
| 245 | * unlock(wait_lock); |
| 246 | * lock(wait_lock); |
| 247 | * acquire(lock); |
| 248 | */ |
Davidlohr Bueso | 700318d | 2015-09-30 13:03:13 -0700 | [diff] [blame] | 249 | return rt_mutex_cmpxchg_release(lock, owner, NULL); |
Thomas Gleixner | 27e3571 | 2014-06-11 18:44:04 +0000 | [diff] [blame] | 250 | } |
| 251 | |
Thomas Gleixner | bd19723 | 2007-06-17 21:11:10 +0200 | [diff] [blame] | 252 | #else |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 253 | static __always_inline bool rt_mutex_cmpxchg_acquire(struct rt_mutex_base *lock, |
Sebastian Andrzej Siewior | 7851593 | 2021-08-15 23:27:54 +0200 | [diff] [blame] | 254 | struct task_struct *old, |
| 255 | struct task_struct *new) |
| 256 | { |
| 257 | return false; |
| 258 | |
| 259 | } |
| 260 | |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 261 | static __always_inline bool rt_mutex_cmpxchg_release(struct rt_mutex_base *lock, |
Sebastian Andrzej Siewior | 7851593 | 2021-08-15 23:27:54 +0200 | [diff] [blame] | 262 | struct task_struct *old, |
| 263 | struct task_struct *new) |
| 264 | { |
| 265 | return false; |
| 266 | } |
Davidlohr Bueso | 700318d | 2015-09-30 13:03:13 -0700 | [diff] [blame] | 267 | |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 268 | static __always_inline void mark_rt_mutex_waiters(struct rt_mutex_base *lock) |
Thomas Gleixner | bd19723 | 2007-06-17 21:11:10 +0200 | [diff] [blame] | 269 | { |
| 270 | lock->owner = (struct task_struct *) |
| 271 | ((unsigned long)lock->owner | RT_MUTEX_HAS_WAITERS); |
| 272 | } |
Thomas Gleixner | 27e3571 | 2014-06-11 18:44:04 +0000 | [diff] [blame] | 273 | |
| 274 | /* |
| 275 | * Simple slow path only version: lock->owner is protected by lock->wait_lock. |
| 276 | */ |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 277 | static __always_inline bool unlock_rt_mutex_safe(struct rt_mutex_base *lock, |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 278 | unsigned long flags) |
Thomas Gleixner | 27e3571 | 2014-06-11 18:44:04 +0000 | [diff] [blame] | 279 | __releases(lock->wait_lock) |
| 280 | { |
| 281 | lock->owner = NULL; |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 282 | raw_spin_unlock_irqrestore(&lock->wait_lock, flags); |
Thomas Gleixner | 27e3571 | 2014-06-11 18:44:04 +0000 | [diff] [blame] | 283 | return true; |
| 284 | } |
Thomas Gleixner | bd19723 | 2007-06-17 21:11:10 +0200 | [diff] [blame] | 285 | #endif |
| 286 | |
Peter Zijlstra | 715f7f9 | 2021-08-15 23:28:30 +0200 | [diff] [blame] | 287 | static __always_inline int __waiter_prio(struct task_struct *task) |
| 288 | { |
| 289 | int prio = task->prio; |
| 290 | |
| 291 | if (!rt_prio(prio)) |
| 292 | return DEFAULT_PRIO; |
| 293 | |
| 294 | return prio; |
| 295 | } |
| 296 | |
| 297 | static __always_inline void |
| 298 | waiter_update_prio(struct rt_mutex_waiter *waiter, struct task_struct *task) |
| 299 | { |
| 300 | waiter->prio = __waiter_prio(task); |
| 301 | waiter->deadline = task->dl.deadline; |
| 302 | } |
| 303 | |
Peter Zijlstra | 19830e5 | 2017-03-23 15:56:14 +0100 | [diff] [blame] | 304 | /* |
| 305 | * Only use with rt_mutex_waiter_{less,equal}() |
| 306 | */ |
| 307 | #define task_to_waiter(p) \ |
Peter Zijlstra | 715f7f9 | 2021-08-15 23:28:30 +0200 | [diff] [blame] | 308 | &(struct rt_mutex_waiter){ .prio = __waiter_prio(p), .deadline = (p)->dl.deadline } |
Peter Zijlstra | 19830e5 | 2017-03-23 15:56:14 +0100 | [diff] [blame] | 309 | |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 310 | static __always_inline int rt_mutex_waiter_less(struct rt_mutex_waiter *left, |
| 311 | struct rt_mutex_waiter *right) |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 312 | { |
Dario Faggioli | 2d3d891 | 2013-11-07 14:43:44 +0100 | [diff] [blame] | 313 | if (left->prio < right->prio) |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 314 | return 1; |
| 315 | |
| 316 | /* |
Dario Faggioli | 2d3d891 | 2013-11-07 14:43:44 +0100 | [diff] [blame] | 317 | * If both waiters have dl_prio(), we check the deadlines of the |
| 318 | * associated tasks. |
| 319 | * If left waiter has a dl_prio(), and we didn't return 1 above, |
| 320 | * then right waiter has a dl_prio() too. |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 321 | */ |
Dario Faggioli | 2d3d891 | 2013-11-07 14:43:44 +0100 | [diff] [blame] | 322 | if (dl_prio(left->prio)) |
Peter Zijlstra | e0aad5b | 2017-03-23 15:56:13 +0100 | [diff] [blame] | 323 | return dl_time_before(left->deadline, right->deadline); |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 324 | |
| 325 | return 0; |
| 326 | } |
| 327 | |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 328 | static __always_inline int rt_mutex_waiter_equal(struct rt_mutex_waiter *left, |
| 329 | struct rt_mutex_waiter *right) |
Peter Zijlstra | 19830e5 | 2017-03-23 15:56:14 +0100 | [diff] [blame] | 330 | { |
| 331 | if (left->prio != right->prio) |
| 332 | return 0; |
| 333 | |
| 334 | /* |
| 335 | * If both waiters have dl_prio(), we check the deadlines of the |
| 336 | * associated tasks. |
| 337 | * If left waiter has a dl_prio(), and we didn't return 0 above, |
| 338 | * then right waiter has a dl_prio() too. |
| 339 | */ |
| 340 | if (dl_prio(left->prio)) |
| 341 | return left->deadline == right->deadline; |
| 342 | |
| 343 | return 1; |
| 344 | } |
| 345 | |
Gregory Haskins | 48eb3f4 | 2021-08-15 23:29:23 +0200 | [diff] [blame] | 346 | static inline bool rt_mutex_steal(struct rt_mutex_waiter *waiter, |
| 347 | struct rt_mutex_waiter *top_waiter) |
| 348 | { |
| 349 | if (rt_mutex_waiter_less(waiter, top_waiter)) |
| 350 | return true; |
| 351 | |
| 352 | #ifdef RT_MUTEX_BUILD_SPINLOCKS |
| 353 | /* |
| 354 | * Note that RT tasks are excluded from same priority (lateral) |
| 355 | * steals to prevent the introduction of an unbounded latency. |
| 356 | */ |
| 357 | if (rt_prio(waiter->prio) || dl_prio(waiter->prio)) |
| 358 | return false; |
| 359 | |
| 360 | return rt_mutex_waiter_equal(waiter, top_waiter); |
| 361 | #else |
| 362 | return false; |
| 363 | #endif |
| 364 | } |
| 365 | |
Peter Zijlstra | 5a79872 | 2020-04-29 17:29:58 +0200 | [diff] [blame] | 366 | #define __node_2_waiter(node) \ |
| 367 | rb_entry((node), struct rt_mutex_waiter, tree_entry) |
| 368 | |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 369 | static __always_inline bool __waiter_less(struct rb_node *a, const struct rb_node *b) |
Peter Zijlstra | 5a79872 | 2020-04-29 17:29:58 +0200 | [diff] [blame] | 370 | { |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 371 | struct rt_mutex_waiter *aw = __node_2_waiter(a); |
| 372 | struct rt_mutex_waiter *bw = __node_2_waiter(b); |
| 373 | |
| 374 | if (rt_mutex_waiter_less(aw, bw)) |
| 375 | return 1; |
| 376 | |
| 377 | if (!build_ww_mutex()) |
| 378 | return 0; |
| 379 | |
| 380 | if (rt_mutex_waiter_less(bw, aw)) |
| 381 | return 0; |
| 382 | |
| 383 | /* NOTE: relies on waiter->ww_ctx being set before insertion */ |
| 384 | if (aw->ww_ctx) { |
| 385 | if (!bw->ww_ctx) |
| 386 | return 1; |
| 387 | |
| 388 | return (signed long)(aw->ww_ctx->stamp - |
| 389 | bw->ww_ctx->stamp) < 0; |
| 390 | } |
| 391 | |
| 392 | return 0; |
Peter Zijlstra | 5a79872 | 2020-04-29 17:29:58 +0200 | [diff] [blame] | 393 | } |
| 394 | |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 395 | static __always_inline void |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 396 | rt_mutex_enqueue(struct rt_mutex_base *lock, struct rt_mutex_waiter *waiter) |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 397 | { |
Peter Zijlstra | 5a79872 | 2020-04-29 17:29:58 +0200 | [diff] [blame] | 398 | rb_add_cached(&waiter->tree_entry, &lock->waiters, __waiter_less); |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 399 | } |
| 400 | |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 401 | static __always_inline void |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 402 | rt_mutex_dequeue(struct rt_mutex_base *lock, struct rt_mutex_waiter *waiter) |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 403 | { |
| 404 | if (RB_EMPTY_NODE(&waiter->tree_entry)) |
| 405 | return; |
| 406 | |
Davidlohr Bueso | a23ba90 | 2017-09-08 16:15:01 -0700 | [diff] [blame] | 407 | rb_erase_cached(&waiter->tree_entry, &lock->waiters); |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 408 | RB_CLEAR_NODE(&waiter->tree_entry); |
| 409 | } |
| 410 | |
Peter Zijlstra | 5a79872 | 2020-04-29 17:29:58 +0200 | [diff] [blame] | 411 | #define __node_2_pi_waiter(node) \ |
| 412 | rb_entry((node), struct rt_mutex_waiter, pi_tree_entry) |
| 413 | |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 414 | static __always_inline bool |
| 415 | __pi_waiter_less(struct rb_node *a, const struct rb_node *b) |
Peter Zijlstra | 5a79872 | 2020-04-29 17:29:58 +0200 | [diff] [blame] | 416 | { |
| 417 | return rt_mutex_waiter_less(__node_2_pi_waiter(a), __node_2_pi_waiter(b)); |
| 418 | } |
| 419 | |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 420 | static __always_inline void |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 421 | rt_mutex_enqueue_pi(struct task_struct *task, struct rt_mutex_waiter *waiter) |
| 422 | { |
Peter Zijlstra | 5a79872 | 2020-04-29 17:29:58 +0200 | [diff] [blame] | 423 | rb_add_cached(&waiter->pi_tree_entry, &task->pi_waiters, __pi_waiter_less); |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 424 | } |
| 425 | |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 426 | static __always_inline void |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 427 | rt_mutex_dequeue_pi(struct task_struct *task, struct rt_mutex_waiter *waiter) |
| 428 | { |
| 429 | if (RB_EMPTY_NODE(&waiter->pi_tree_entry)) |
| 430 | return; |
| 431 | |
Davidlohr Bueso | a23ba90 | 2017-09-08 16:15:01 -0700 | [diff] [blame] | 432 | rb_erase_cached(&waiter->pi_tree_entry, &task->pi_waiters); |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 433 | RB_CLEAR_NODE(&waiter->pi_tree_entry); |
| 434 | } |
| 435 | |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 436 | static __always_inline void rt_mutex_adjust_prio(struct task_struct *p) |
Xunlei Pang | e96a7705 | 2017-03-23 15:56:08 +0100 | [diff] [blame] | 437 | { |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 438 | struct task_struct *pi_task = NULL; |
Xunlei Pang | e96a7705 | 2017-03-23 15:56:08 +0100 | [diff] [blame] | 439 | |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 440 | lockdep_assert_held(&p->pi_lock); |
Xunlei Pang | e96a7705 | 2017-03-23 15:56:08 +0100 | [diff] [blame] | 441 | |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 442 | if (task_has_pi_waiters(p)) |
| 443 | pi_task = task_top_pi_waiter(p)->task; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 444 | |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 445 | rt_mutex_setprio(p, pi_task); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 446 | } |
| 447 | |
Thomas Gleixner | b576e64 | 2021-08-15 23:28:08 +0200 | [diff] [blame] | 448 | /* RT mutex specific wake_q wrappers */ |
Thomas Gleixner | 9321f81 | 2021-09-28 17:00:06 +0200 | [diff] [blame] | 449 | static __always_inline void rt_mutex_wake_q_add_task(struct rt_wake_q_head *wqh, |
| 450 | struct task_struct *task, |
| 451 | unsigned int wake_state) |
| 452 | { |
| 453 | if (IS_ENABLED(CONFIG_PREEMPT_RT) && wake_state == TASK_RTLOCK_WAIT) { |
| 454 | if (IS_ENABLED(CONFIG_PROVE_LOCKING)) |
| 455 | WARN_ON_ONCE(wqh->rtlock_task); |
| 456 | get_task_struct(task); |
| 457 | wqh->rtlock_task = task; |
| 458 | } else { |
| 459 | wake_q_add(&wqh->head, task); |
| 460 | } |
| 461 | } |
| 462 | |
Thomas Gleixner | b576e64 | 2021-08-15 23:28:08 +0200 | [diff] [blame] | 463 | static __always_inline void rt_mutex_wake_q_add(struct rt_wake_q_head *wqh, |
| 464 | struct rt_mutex_waiter *w) |
| 465 | { |
Thomas Gleixner | 9321f81 | 2021-09-28 17:00:06 +0200 | [diff] [blame] | 466 | rt_mutex_wake_q_add_task(wqh, w->task, w->wake_state); |
Thomas Gleixner | b576e64 | 2021-08-15 23:28:08 +0200 | [diff] [blame] | 467 | } |
| 468 | |
| 469 | static __always_inline void rt_mutex_wake_up_q(struct rt_wake_q_head *wqh) |
| 470 | { |
Thomas Gleixner | 456cfbc | 2021-08-15 23:28:11 +0200 | [diff] [blame] | 471 | if (IS_ENABLED(CONFIG_PREEMPT_RT) && wqh->rtlock_task) { |
| 472 | wake_up_state(wqh->rtlock_task, TASK_RTLOCK_WAIT); |
| 473 | put_task_struct(wqh->rtlock_task); |
| 474 | wqh->rtlock_task = NULL; |
| 475 | } |
| 476 | |
| 477 | if (!wake_q_empty(&wqh->head)) |
| 478 | wake_up_q(&wqh->head); |
Thomas Gleixner | b576e64 | 2021-08-15 23:28:08 +0200 | [diff] [blame] | 479 | |
| 480 | /* Pairs with preempt_disable() in mark_wakeup_next_waiter() */ |
| 481 | preempt_enable(); |
| 482 | } |
| 483 | |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 484 | /* |
Thomas Gleixner | 8930ed8 | 2014-05-22 03:25:47 +0000 | [diff] [blame] | 485 | * Deadlock detection is conditional: |
| 486 | * |
| 487 | * If CONFIG_DEBUG_RT_MUTEXES=n, deadlock detection is only conducted |
| 488 | * if the detect argument is == RT_MUTEX_FULL_CHAINWALK. |
| 489 | * |
| 490 | * If CONFIG_DEBUG_RT_MUTEXES=y, deadlock detection is always |
| 491 | * conducted independent of the detect argument. |
| 492 | * |
| 493 | * If the waiter argument is NULL this indicates the deboost path and |
| 494 | * deadlock detection is disabled independent of the detect argument |
| 495 | * and the config settings. |
| 496 | */ |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 497 | static __always_inline bool |
| 498 | rt_mutex_cond_detect_deadlock(struct rt_mutex_waiter *waiter, |
| 499 | enum rtmutex_chainwalk chwalk) |
Thomas Gleixner | 8930ed8 | 2014-05-22 03:25:47 +0000 | [diff] [blame] | 500 | { |
Zhen Lei | 07d2597 | 2021-07-31 20:30:11 +0800 | [diff] [blame] | 501 | if (IS_ENABLED(CONFIG_DEBUG_RT_MUTEXES)) |
Thomas Gleixner | f7efc47 | 2021-03-26 16:29:36 +0100 | [diff] [blame] | 502 | return waiter != NULL; |
| 503 | return chwalk == RT_MUTEX_FULL_CHAINWALK; |
Thomas Gleixner | 8930ed8 | 2014-05-22 03:25:47 +0000 | [diff] [blame] | 504 | } |
| 505 | |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 506 | static __always_inline struct rt_mutex_base *task_blocked_on_lock(struct task_struct *p) |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 507 | { |
| 508 | return p->pi_blocked_on ? p->pi_blocked_on->lock : NULL; |
| 509 | } |
| 510 | |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 511 | /* |
| 512 | * Adjust the priority chain. Also used for deadlock detection. |
| 513 | * Decreases task's usage by one - may thus free the task. |
Juri Lelli | 0c10617 | 2013-05-15 11:04:10 +0200 | [diff] [blame] | 514 | * |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 515 | * @task: the task owning the mutex (owner) for which a chain walk is |
| 516 | * probably needed |
Tom(JeHyeon) Yeon | e6beaa36 | 2015-03-18 14:03:30 +0900 | [diff] [blame] | 517 | * @chwalk: do we have to carry out deadlock detection? |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 518 | * @orig_lock: the mutex (can be NULL if we are walking the chain to recheck |
| 519 | * things for a task that has just got its priority adjusted, and |
| 520 | * is waiting on a mutex) |
| 521 | * @next_lock: the mutex on which the owner of @orig_lock was blocked before |
| 522 | * we dropped its pi_lock. Is never dereferenced, only used for |
| 523 | * comparison to detect lock chain changes. |
Juri Lelli | 0c10617 | 2013-05-15 11:04:10 +0200 | [diff] [blame] | 524 | * @orig_waiter: rt_mutex_waiter struct for the task that has just donated |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 525 | * its priority to the mutex owner (can be NULL in the case |
| 526 | * depicted above or if the top waiter is gone away and we are |
| 527 | * actually deboosting the owner) |
| 528 | * @top_task: the current top waiter |
Juri Lelli | 0c10617 | 2013-05-15 11:04:10 +0200 | [diff] [blame] | 529 | * |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 530 | * Returns 0 or -EDEADLK. |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 531 | * |
| 532 | * Chain walk basics and protection scope |
| 533 | * |
| 534 | * [R] refcount on task |
| 535 | * [P] task->pi_lock held |
| 536 | * [L] rtmutex->wait_lock held |
| 537 | * |
| 538 | * Step Description Protected by |
| 539 | * function arguments: |
| 540 | * @task [R] |
| 541 | * @orig_lock if != NULL @top_task is blocked on it |
| 542 | * @next_lock Unprotected. Cannot be |
| 543 | * dereferenced. Only used for |
| 544 | * comparison. |
| 545 | * @orig_waiter if != NULL @top_task is blocked on it |
| 546 | * @top_task current, or in case of proxy |
| 547 | * locking protected by calling |
| 548 | * code |
| 549 | * again: |
| 550 | * loop_sanity_check(); |
| 551 | * retry: |
| 552 | * [1] lock(task->pi_lock); [R] acquire [P] |
| 553 | * [2] waiter = task->pi_blocked_on; [P] |
| 554 | * [3] check_exit_conditions_1(); [P] |
| 555 | * [4] lock = waiter->lock; [P] |
| 556 | * [5] if (!try_lock(lock->wait_lock)) { [P] try to acquire [L] |
| 557 | * unlock(task->pi_lock); release [P] |
| 558 | * goto retry; |
| 559 | * } |
| 560 | * [6] check_exit_conditions_2(); [P] + [L] |
| 561 | * [7] requeue_lock_waiter(lock, waiter); [P] + [L] |
| 562 | * [8] unlock(task->pi_lock); release [P] |
| 563 | * put_task_struct(task); release [R] |
| 564 | * [9] check_exit_conditions_3(); [L] |
| 565 | * [10] task = owner(lock); [L] |
| 566 | * get_task_struct(task); [L] acquire [R] |
| 567 | * lock(task->pi_lock); [L] acquire [P] |
| 568 | * [11] requeue_pi_waiter(tsk, waiters(lock));[P] + [L] |
| 569 | * [12] check_exit_conditions_4(); [P] + [L] |
| 570 | * [13] unlock(task->pi_lock); release [P] |
| 571 | * unlock(lock->wait_lock); release [L] |
| 572 | * goto again; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 573 | */ |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 574 | static int __sched rt_mutex_adjust_prio_chain(struct task_struct *task, |
| 575 | enum rtmutex_chainwalk chwalk, |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 576 | struct rt_mutex_base *orig_lock, |
| 577 | struct rt_mutex_base *next_lock, |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 578 | struct rt_mutex_waiter *orig_waiter, |
| 579 | struct task_struct *top_task) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 580 | { |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 581 | struct rt_mutex_waiter *waiter, *top_waiter = orig_waiter; |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 582 | struct rt_mutex_waiter *prerequeue_top_waiter; |
Thomas Gleixner | 8930ed8 | 2014-05-22 03:25:47 +0000 | [diff] [blame] | 583 | int ret = 0, depth = 0; |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 584 | struct rt_mutex_base *lock; |
Thomas Gleixner | 8930ed8 | 2014-05-22 03:25:47 +0000 | [diff] [blame] | 585 | bool detect_deadlock; |
Thomas Gleixner | 67792e2 | 2014-05-22 03:25:57 +0000 | [diff] [blame] | 586 | bool requeue = true; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 587 | |
Thomas Gleixner | 8930ed8 | 2014-05-22 03:25:47 +0000 | [diff] [blame] | 588 | detect_deadlock = rt_mutex_cond_detect_deadlock(orig_waiter, chwalk); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 589 | |
| 590 | /* |
| 591 | * The (de)boosting is a step by step approach with a lot of |
| 592 | * pitfalls. We want this to be preemptible and we want hold a |
| 593 | * maximum of two locks per step. So we have to check |
| 594 | * carefully whether things change under us. |
| 595 | */ |
| 596 | again: |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 597 | /* |
| 598 | * We limit the lock chain length for each invocation. |
| 599 | */ |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 600 | if (++depth > max_lock_depth) { |
| 601 | static int prev_max; |
| 602 | |
| 603 | /* |
| 604 | * Print this only once. If the admin changes the limit, |
| 605 | * print a new message when reaching the limit again. |
| 606 | */ |
| 607 | if (prev_max != max_lock_depth) { |
| 608 | prev_max = max_lock_depth; |
| 609 | printk(KERN_WARNING "Maximum lock depth %d reached " |
| 610 | "task: %s (%d)\n", max_lock_depth, |
Pavel Emelyanov | ba25f9d | 2007-10-18 23:40:40 -0700 | [diff] [blame] | 611 | top_task->comm, task_pid_nr(top_task)); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 612 | } |
| 613 | put_task_struct(task); |
| 614 | |
Thomas Gleixner | 3d5c934 | 2014-06-05 12:34:23 +0200 | [diff] [blame] | 615 | return -EDEADLK; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 616 | } |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 617 | |
| 618 | /* |
| 619 | * We are fully preemptible here and only hold the refcount on |
| 620 | * @task. So everything can have changed under us since the |
| 621 | * caller or our own code below (goto retry/again) dropped all |
| 622 | * locks. |
| 623 | */ |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 624 | retry: |
| 625 | /* |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 626 | * [1] Task cannot go away as we did a get_task() before ! |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 627 | */ |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 628 | raw_spin_lock_irq(&task->pi_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 629 | |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 630 | /* |
| 631 | * [2] Get the waiter on which @task is blocked on. |
| 632 | */ |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 633 | waiter = task->pi_blocked_on; |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 634 | |
| 635 | /* |
| 636 | * [3] check_exit_conditions_1() protected by task->pi_lock. |
| 637 | */ |
| 638 | |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 639 | /* |
| 640 | * Check whether the end of the boosting chain has been |
| 641 | * reached or the state of the chain has changed while we |
| 642 | * dropped the locks. |
| 643 | */ |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 644 | if (!waiter) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 645 | goto out_unlock_pi; |
| 646 | |
Thomas Gleixner | 1a539a8 | 2007-06-08 13:46:58 -0700 | [diff] [blame] | 647 | /* |
| 648 | * Check the orig_waiter state. After we dropped the locks, |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 649 | * the previous owner of the lock might have released the lock. |
Thomas Gleixner | 1a539a8 | 2007-06-08 13:46:58 -0700 | [diff] [blame] | 650 | */ |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 651 | if (orig_waiter && !rt_mutex_owner(orig_lock)) |
Thomas Gleixner | 1a539a8 | 2007-06-08 13:46:58 -0700 | [diff] [blame] | 652 | goto out_unlock_pi; |
| 653 | |
| 654 | /* |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 655 | * We dropped all locks after taking a refcount on @task, so |
| 656 | * the task might have moved on in the lock chain or even left |
| 657 | * the chain completely and blocks now on an unrelated lock or |
| 658 | * on @orig_lock. |
| 659 | * |
| 660 | * We stored the lock on which @task was blocked in @next_lock, |
| 661 | * so we can detect the chain change. |
| 662 | */ |
| 663 | if (next_lock != waiter->lock) |
| 664 | goto out_unlock_pi; |
| 665 | |
| 666 | /* |
Peter Zijlstra | 6467822 | 2021-08-26 09:36:53 +0200 | [diff] [blame] | 667 | * There could be 'spurious' loops in the lock graph due to ww_mutex, |
| 668 | * consider: |
| 669 | * |
| 670 | * P1: A, ww_A, ww_B |
| 671 | * P2: ww_B, ww_A |
| 672 | * P3: A |
| 673 | * |
| 674 | * P3 should not return -EDEADLK because it gets trapped in the cycle |
| 675 | * created by P1 and P2 (which will resolve -- and runs into |
| 676 | * max_lock_depth above). Therefore disable detect_deadlock such that |
| 677 | * the below termination condition can trigger once all relevant tasks |
| 678 | * are boosted. |
| 679 | * |
| 680 | * Even when we start with ww_mutex we can disable deadlock detection, |
| 681 | * since we would supress a ww_mutex induced deadlock at [6] anyway. |
| 682 | * Supressing it here however is not sufficient since we might still |
| 683 | * hit [6] due to adjustment driven iteration. |
| 684 | * |
| 685 | * NOTE: if someone were to create a deadlock between 2 ww_classes we'd |
| 686 | * utterly fail to report it; lockdep should. |
| 687 | */ |
| 688 | if (IS_ENABLED(CONFIG_PREEMPT_RT) && waiter->ww_ctx && detect_deadlock) |
| 689 | detect_deadlock = false; |
| 690 | |
| 691 | /* |
Thomas Gleixner | 1a539a8 | 2007-06-08 13:46:58 -0700 | [diff] [blame] | 692 | * Drop out, when the task has no waiters. Note, |
| 693 | * top_waiter can be NULL, when we are in the deboosting |
| 694 | * mode! |
| 695 | */ |
Thomas Gleixner | 397335f | 2014-05-22 03:25:39 +0000 | [diff] [blame] | 696 | if (top_waiter) { |
| 697 | if (!task_has_pi_waiters(task)) |
| 698 | goto out_unlock_pi; |
| 699 | /* |
| 700 | * If deadlock detection is off, we stop here if we |
Thomas Gleixner | 67792e2 | 2014-05-22 03:25:57 +0000 | [diff] [blame] | 701 | * are not the top pi waiter of the task. If deadlock |
| 702 | * detection is enabled we continue, but stop the |
| 703 | * requeueing in the chain walk. |
Thomas Gleixner | 397335f | 2014-05-22 03:25:39 +0000 | [diff] [blame] | 704 | */ |
Thomas Gleixner | 67792e2 | 2014-05-22 03:25:57 +0000 | [diff] [blame] | 705 | if (top_waiter != task_top_pi_waiter(task)) { |
| 706 | if (!detect_deadlock) |
| 707 | goto out_unlock_pi; |
| 708 | else |
| 709 | requeue = false; |
| 710 | } |
Thomas Gleixner | 397335f | 2014-05-22 03:25:39 +0000 | [diff] [blame] | 711 | } |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 712 | |
| 713 | /* |
Thomas Gleixner | 67792e2 | 2014-05-22 03:25:57 +0000 | [diff] [blame] | 714 | * If the waiter priority is the same as the task priority |
| 715 | * then there is no further priority adjustment necessary. If |
| 716 | * deadlock detection is off, we stop the chain walk. If its |
| 717 | * enabled we continue, but stop the requeueing in the chain |
| 718 | * walk. |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 719 | */ |
Peter Zijlstra | 19830e5 | 2017-03-23 15:56:14 +0100 | [diff] [blame] | 720 | if (rt_mutex_waiter_equal(waiter, task_to_waiter(task))) { |
Thomas Gleixner | 67792e2 | 2014-05-22 03:25:57 +0000 | [diff] [blame] | 721 | if (!detect_deadlock) |
| 722 | goto out_unlock_pi; |
| 723 | else |
| 724 | requeue = false; |
| 725 | } |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 726 | |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 727 | /* |
| 728 | * [4] Get the next lock |
| 729 | */ |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 730 | lock = waiter->lock; |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 731 | /* |
| 732 | * [5] We need to trylock here as we are holding task->pi_lock, |
| 733 | * which is the reverse lock order versus the other rtmutex |
| 734 | * operations. |
| 735 | */ |
Thomas Gleixner | d209d74 | 2009-11-17 18:22:11 +0100 | [diff] [blame] | 736 | if (!raw_spin_trylock(&lock->wait_lock)) { |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 737 | raw_spin_unlock_irq(&task->pi_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 738 | cpu_relax(); |
| 739 | goto retry; |
| 740 | } |
| 741 | |
Thomas Gleixner | 397335f | 2014-05-22 03:25:39 +0000 | [diff] [blame] | 742 | /* |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 743 | * [6] check_exit_conditions_2() protected by task->pi_lock and |
| 744 | * lock->wait_lock. |
| 745 | * |
Thomas Gleixner | 397335f | 2014-05-22 03:25:39 +0000 | [diff] [blame] | 746 | * Deadlock detection. If the lock is the same as the original |
| 747 | * lock which caused us to walk the lock chain or if the |
| 748 | * current lock is owned by the task which initiated the chain |
| 749 | * walk, we detected a deadlock. |
| 750 | */ |
Thomas Gleixner | 95e02ca | 2006-06-27 02:55:02 -0700 | [diff] [blame] | 751 | if (lock == orig_lock || rt_mutex_owner(lock) == top_task) { |
Thomas Gleixner | 3d5c934 | 2014-06-05 12:34:23 +0200 | [diff] [blame] | 752 | ret = -EDEADLK; |
Peter Zijlstra | a055fcc | 2021-08-26 10:48:18 +0200 | [diff] [blame] | 753 | |
| 754 | /* |
| 755 | * When the deadlock is due to ww_mutex; also see above. Don't |
| 756 | * report the deadlock and instead let the ww_mutex wound/die |
| 757 | * logic pick which of the contending threads gets -EDEADLK. |
| 758 | * |
| 759 | * NOTE: assumes the cycle only contains a single ww_class; any |
| 760 | * other configuration and we fail to report; also, see |
| 761 | * lockdep. |
| 762 | */ |
Peter Zijlstra | e548057 | 2021-09-01 11:44:11 +0200 | [diff] [blame] | 763 | if (IS_ENABLED(CONFIG_PREEMPT_RT) && orig_waiter && orig_waiter->ww_ctx) |
Peter Zijlstra | a055fcc | 2021-08-26 10:48:18 +0200 | [diff] [blame] | 764 | ret = 0; |
| 765 | |
| 766 | raw_spin_unlock(&lock->wait_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 767 | goto out_unlock_pi; |
| 768 | } |
| 769 | |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 770 | /* |
Thomas Gleixner | 67792e2 | 2014-05-22 03:25:57 +0000 | [diff] [blame] | 771 | * If we just follow the lock chain for deadlock detection, no |
| 772 | * need to do all the requeue operations. To avoid a truckload |
| 773 | * of conditionals around the various places below, just do the |
| 774 | * minimum chain walk checks. |
| 775 | */ |
| 776 | if (!requeue) { |
| 777 | /* |
| 778 | * No requeue[7] here. Just release @task [8] |
| 779 | */ |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 780 | raw_spin_unlock(&task->pi_lock); |
Thomas Gleixner | 67792e2 | 2014-05-22 03:25:57 +0000 | [diff] [blame] | 781 | put_task_struct(task); |
| 782 | |
| 783 | /* |
| 784 | * [9] check_exit_conditions_3 protected by lock->wait_lock. |
| 785 | * If there is no owner of the lock, end of chain. |
| 786 | */ |
| 787 | if (!rt_mutex_owner(lock)) { |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 788 | raw_spin_unlock_irq(&lock->wait_lock); |
Thomas Gleixner | 67792e2 | 2014-05-22 03:25:57 +0000 | [diff] [blame] | 789 | return 0; |
| 790 | } |
| 791 | |
| 792 | /* [10] Grab the next task, i.e. owner of @lock */ |
Matthew Wilcox (Oracle) | 7b3c92b | 2019-07-04 15:13:23 -0700 | [diff] [blame] | 793 | task = get_task_struct(rt_mutex_owner(lock)); |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 794 | raw_spin_lock(&task->pi_lock); |
Thomas Gleixner | 67792e2 | 2014-05-22 03:25:57 +0000 | [diff] [blame] | 795 | |
| 796 | /* |
| 797 | * No requeue [11] here. We just do deadlock detection. |
| 798 | * |
| 799 | * [12] Store whether owner is blocked |
| 800 | * itself. Decision is made after dropping the locks |
| 801 | */ |
| 802 | next_lock = task_blocked_on_lock(task); |
| 803 | /* |
| 804 | * Get the top waiter for the next iteration |
| 805 | */ |
| 806 | top_waiter = rt_mutex_top_waiter(lock); |
| 807 | |
| 808 | /* [13] Drop locks */ |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 809 | raw_spin_unlock(&task->pi_lock); |
| 810 | raw_spin_unlock_irq(&lock->wait_lock); |
Thomas Gleixner | 67792e2 | 2014-05-22 03:25:57 +0000 | [diff] [blame] | 811 | |
| 812 | /* If owner is not blocked, end of chain. */ |
| 813 | if (!next_lock) |
| 814 | goto out_put_task; |
| 815 | goto again; |
| 816 | } |
| 817 | |
| 818 | /* |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 819 | * Store the current top waiter before doing the requeue |
| 820 | * operation on @lock. We need it for the boost/deboost |
| 821 | * decision below. |
| 822 | */ |
| 823 | prerequeue_top_waiter = rt_mutex_top_waiter(lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 824 | |
Davidlohr Bueso | 9f40a51 | 2015-05-19 10:24:57 -0700 | [diff] [blame] | 825 | /* [7] Requeue the waiter in the lock waiter tree. */ |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 826 | rt_mutex_dequeue(lock, waiter); |
Peter Zijlstra | e0aad5b | 2017-03-23 15:56:13 +0100 | [diff] [blame] | 827 | |
| 828 | /* |
| 829 | * Update the waiter prio fields now that we're dequeued. |
| 830 | * |
| 831 | * These values can have changed through either: |
| 832 | * |
| 833 | * sys_sched_set_scheduler() / sys_sched_setattr() |
| 834 | * |
| 835 | * or |
| 836 | * |
| 837 | * DL CBS enforcement advancing the effective deadline. |
| 838 | * |
| 839 | * Even though pi_waiters also uses these fields, and that tree is only |
| 840 | * updated in [11], we can do this here, since we hold [L], which |
| 841 | * serializes all pi_waiters access and rb_erase() does not care about |
| 842 | * the values of the node being removed. |
| 843 | */ |
Peter Zijlstra | 715f7f9 | 2021-08-15 23:28:30 +0200 | [diff] [blame] | 844 | waiter_update_prio(waiter, task); |
Peter Zijlstra | e0aad5b | 2017-03-23 15:56:13 +0100 | [diff] [blame] | 845 | |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 846 | rt_mutex_enqueue(lock, waiter); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 847 | |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 848 | /* [8] Release the task */ |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 849 | raw_spin_unlock(&task->pi_lock); |
Thomas Gleixner | 2ffa5a5 | 2014-06-07 12:10:36 +0200 | [diff] [blame] | 850 | put_task_struct(task); |
| 851 | |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 852 | /* |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 853 | * [9] check_exit_conditions_3 protected by lock->wait_lock. |
| 854 | * |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 855 | * We must abort the chain walk if there is no lock owner even |
| 856 | * in the dead lock detection case, as we have nothing to |
| 857 | * follow here. This is the end of the chain we are walking. |
| 858 | */ |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 859 | if (!rt_mutex_owner(lock)) { |
| 860 | /* |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 861 | * If the requeue [7] above changed the top waiter, |
| 862 | * then we need to wake the new top waiter up to try |
| 863 | * to get the lock. |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 864 | */ |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 865 | if (prerequeue_top_waiter != rt_mutex_top_waiter(lock)) |
Thomas Gleixner | c014ef6 | 2021-08-15 23:28:06 +0200 | [diff] [blame] | 866 | wake_up_state(waiter->task, waiter->wake_state); |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 867 | raw_spin_unlock_irq(&lock->wait_lock); |
Thomas Gleixner | 2ffa5a5 | 2014-06-07 12:10:36 +0200 | [diff] [blame] | 868 | return 0; |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 869 | } |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 870 | |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 871 | /* [10] Grab the next task, i.e. the owner of @lock */ |
Matthew Wilcox (Oracle) | 7b3c92b | 2019-07-04 15:13:23 -0700 | [diff] [blame] | 872 | task = get_task_struct(rt_mutex_owner(lock)); |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 873 | raw_spin_lock(&task->pi_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 874 | |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 875 | /* [11] requeue the pi waiters if necessary */ |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 876 | if (waiter == rt_mutex_top_waiter(lock)) { |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 877 | /* |
| 878 | * The waiter became the new top (highest priority) |
| 879 | * waiter on the lock. Replace the previous top waiter |
Davidlohr Bueso | 9f40a51 | 2015-05-19 10:24:57 -0700 | [diff] [blame] | 880 | * in the owner tasks pi waiters tree with this waiter |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 881 | * and adjust the priority of the owner. |
| 882 | */ |
| 883 | rt_mutex_dequeue_pi(task, prerequeue_top_waiter); |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 884 | rt_mutex_enqueue_pi(task, waiter); |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 885 | rt_mutex_adjust_prio(task); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 886 | |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 887 | } else if (prerequeue_top_waiter == waiter) { |
| 888 | /* |
| 889 | * The waiter was the top waiter on the lock, but is |
Ingo Molnar | e2db759 | 2021-03-22 02:35:05 +0100 | [diff] [blame] | 890 | * no longer the top priority waiter. Replace waiter in |
Davidlohr Bueso | 9f40a51 | 2015-05-19 10:24:57 -0700 | [diff] [blame] | 891 | * the owner tasks pi waiters tree with the new top |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 892 | * (highest priority) waiter and adjust the priority |
| 893 | * of the owner. |
| 894 | * The new top waiter is stored in @waiter so that |
| 895 | * @waiter == @top_waiter evaluates to true below and |
| 896 | * we continue to deboost the rest of the chain. |
| 897 | */ |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 898 | rt_mutex_dequeue_pi(task, waiter); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 899 | waiter = rt_mutex_top_waiter(lock); |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 900 | rt_mutex_enqueue_pi(task, waiter); |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 901 | rt_mutex_adjust_prio(task); |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 902 | } else { |
| 903 | /* |
| 904 | * Nothing changed. No need to do any priority |
| 905 | * adjustment. |
| 906 | */ |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 907 | } |
| 908 | |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 909 | /* |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 910 | * [12] check_exit_conditions_4() protected by task->pi_lock |
| 911 | * and lock->wait_lock. The actual decisions are made after we |
| 912 | * dropped the locks. |
| 913 | * |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 914 | * Check whether the task which owns the current lock is pi |
| 915 | * blocked itself. If yes we store a pointer to the lock for |
| 916 | * the lock chain change detection above. After we dropped |
| 917 | * task->pi_lock next_lock cannot be dereferenced anymore. |
| 918 | */ |
| 919 | next_lock = task_blocked_on_lock(task); |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 920 | /* |
| 921 | * Store the top waiter of @lock for the end of chain walk |
| 922 | * decision below. |
| 923 | */ |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 924 | top_waiter = rt_mutex_top_waiter(lock); |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 925 | |
| 926 | /* [13] Drop the locks */ |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 927 | raw_spin_unlock(&task->pi_lock); |
| 928 | raw_spin_unlock_irq(&lock->wait_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 929 | |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 930 | /* |
Thomas Gleixner | 3eb65ae | 2014-06-09 19:40:34 +0200 | [diff] [blame] | 931 | * Make the actual exit decisions [12], based on the stored |
| 932 | * values. |
| 933 | * |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 934 | * We reached the end of the lock chain. Stop right here. No |
| 935 | * point to go back just to figure that out. |
| 936 | */ |
| 937 | if (!next_lock) |
| 938 | goto out_put_task; |
| 939 | |
Thomas Gleixner | a57594a | 2014-05-22 03:25:54 +0000 | [diff] [blame] | 940 | /* |
| 941 | * If the current waiter is not the top waiter on the lock, |
| 942 | * then we can stop the chain walk here if we are not in full |
| 943 | * deadlock detection mode. |
| 944 | */ |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 945 | if (!detect_deadlock && waiter != top_waiter) |
| 946 | goto out_put_task; |
| 947 | |
| 948 | goto again; |
| 949 | |
| 950 | out_unlock_pi: |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 951 | raw_spin_unlock_irq(&task->pi_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 952 | out_put_task: |
| 953 | put_task_struct(task); |
Ingo Molnar | 36c8b58 | 2006-07-03 00:25:41 -0700 | [diff] [blame] | 954 | |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 955 | return ret; |
| 956 | } |
| 957 | |
| 958 | /* |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 959 | * Try to take an rt-mutex |
| 960 | * |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 961 | * Must be called with lock->wait_lock held and interrupts disabled |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 962 | * |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 963 | * @lock: The lock to be acquired. |
| 964 | * @task: The task which wants to acquire the lock |
Davidlohr Bueso | 9f40a51 | 2015-05-19 10:24:57 -0700 | [diff] [blame] | 965 | * @waiter: The waiter that is queued to the lock's wait tree if the |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 966 | * callsite called task_blocked_on_lock(), otherwise NULL |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 967 | */ |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 968 | static int __sched |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 969 | try_to_take_rt_mutex(struct rt_mutex_base *lock, struct task_struct *task, |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 970 | struct rt_mutex_waiter *waiter) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 971 | { |
Peter Zijlstra | e0aad5b | 2017-03-23 15:56:13 +0100 | [diff] [blame] | 972 | lockdep_assert_held(&lock->wait_lock); |
| 973 | |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 974 | /* |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 975 | * Before testing whether we can acquire @lock, we set the |
| 976 | * RT_MUTEX_HAS_WAITERS bit in @lock->owner. This forces all |
| 977 | * other tasks which try to modify @lock into the slow path |
| 978 | * and they serialize on @lock->wait_lock. |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 979 | * |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 980 | * The RT_MUTEX_HAS_WAITERS bit can have a transitional state |
| 981 | * as explained at the top of this file if and only if: |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 982 | * |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 983 | * - There is a lock owner. The caller must fixup the |
| 984 | * transient state if it does a trylock or leaves the lock |
| 985 | * function due to a signal or timeout. |
| 986 | * |
| 987 | * - @task acquires the lock and there are no other |
| 988 | * waiters. This is undone in rt_mutex_set_owner(@task) at |
| 989 | * the end of this function. |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 990 | */ |
| 991 | mark_rt_mutex_waiters(lock); |
| 992 | |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 993 | /* |
| 994 | * If @lock has an owner, give up. |
| 995 | */ |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 996 | if (rt_mutex_owner(lock)) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 997 | return 0; |
| 998 | |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 999 | /* |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1000 | * If @waiter != NULL, @task has already enqueued the waiter |
Davidlohr Bueso | 9f40a51 | 2015-05-19 10:24:57 -0700 | [diff] [blame] | 1001 | * into @lock waiter tree. If @waiter == NULL then this is a |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1002 | * trylock attempt. |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 1003 | */ |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1004 | if (waiter) { |
Gregory Haskins | 48eb3f4 | 2021-08-15 23:29:23 +0200 | [diff] [blame] | 1005 | struct rt_mutex_waiter *top_waiter = rt_mutex_top_waiter(lock); |
| 1006 | |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1007 | /* |
Gregory Haskins | 48eb3f4 | 2021-08-15 23:29:23 +0200 | [diff] [blame] | 1008 | * If waiter is the highest priority waiter of @lock, |
| 1009 | * or allowed to steal it, take it over. |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1010 | */ |
Gregory Haskins | 48eb3f4 | 2021-08-15 23:29:23 +0200 | [diff] [blame] | 1011 | if (waiter == top_waiter || rt_mutex_steal(waiter, top_waiter)) { |
| 1012 | /* |
| 1013 | * We can acquire the lock. Remove the waiter from the |
| 1014 | * lock waiters tree. |
| 1015 | */ |
| 1016 | rt_mutex_dequeue(lock, waiter); |
| 1017 | } else { |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1018 | return 0; |
Gregory Haskins | 48eb3f4 | 2021-08-15 23:29:23 +0200 | [diff] [blame] | 1019 | } |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1020 | } else { |
| 1021 | /* |
| 1022 | * If the lock has waiters already we check whether @task is |
| 1023 | * eligible to take over the lock. |
| 1024 | * |
| 1025 | * If there are no other waiters, @task can acquire |
| 1026 | * the lock. @task->pi_blocked_on is NULL, so it does |
| 1027 | * not need to be dequeued. |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 1028 | */ |
| 1029 | if (rt_mutex_has_waiters(lock)) { |
Gregory Haskins | 48eb3f4 | 2021-08-15 23:29:23 +0200 | [diff] [blame] | 1030 | /* Check whether the trylock can steal it. */ |
| 1031 | if (!rt_mutex_steal(task_to_waiter(task), |
| 1032 | rt_mutex_top_waiter(lock))) |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1033 | return 0; |
| 1034 | |
| 1035 | /* |
| 1036 | * The current top waiter stays enqueued. We |
| 1037 | * don't have to change anything in the lock |
| 1038 | * waiters order. |
| 1039 | */ |
| 1040 | } else { |
| 1041 | /* |
| 1042 | * No waiters. Take the lock without the |
| 1043 | * pi_lock dance.@task->pi_blocked_on is NULL |
| 1044 | * and we have no waiters to enqueue in @task |
Davidlohr Bueso | 9f40a51 | 2015-05-19 10:24:57 -0700 | [diff] [blame] | 1045 | * pi waiters tree. |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1046 | */ |
| 1047 | goto takeit; |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 1048 | } |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 1049 | } |
| 1050 | |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1051 | /* |
| 1052 | * Clear @task->pi_blocked_on. Requires protection by |
| 1053 | * @task->pi_lock. Redundant operation for the @waiter == NULL |
| 1054 | * case, but conditionals are more expensive than a redundant |
| 1055 | * store. |
| 1056 | */ |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1057 | raw_spin_lock(&task->pi_lock); |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1058 | task->pi_blocked_on = NULL; |
| 1059 | /* |
| 1060 | * Finish the lock acquisition. @task is the new owner. If |
| 1061 | * other waiters exist we have to insert the highest priority |
Davidlohr Bueso | 9f40a51 | 2015-05-19 10:24:57 -0700 | [diff] [blame] | 1062 | * waiter into @task->pi_waiters tree. |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1063 | */ |
| 1064 | if (rt_mutex_has_waiters(lock)) |
| 1065 | rt_mutex_enqueue_pi(task, rt_mutex_top_waiter(lock)); |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1066 | raw_spin_unlock(&task->pi_lock); |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1067 | |
| 1068 | takeit: |
Thomas Gleixner | 358c331 | 2014-06-11 01:01:13 +0200 | [diff] [blame] | 1069 | /* |
| 1070 | * This either preserves the RT_MUTEX_HAS_WAITERS bit if there |
| 1071 | * are still waiters or clears it. |
| 1072 | */ |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 1073 | rt_mutex_set_owner(lock, task); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1074 | |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1075 | return 1; |
| 1076 | } |
| 1077 | |
| 1078 | /* |
| 1079 | * Task blocks on lock. |
| 1080 | * |
| 1081 | * Prepare waiter and propagate pi chain |
| 1082 | * |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1083 | * This must be called with lock->wait_lock held and interrupts disabled |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1084 | */ |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 1085 | static int __sched task_blocks_on_rt_mutex(struct rt_mutex_base *lock, |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 1086 | struct rt_mutex_waiter *waiter, |
| 1087 | struct task_struct *task, |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1088 | struct ww_acquire_ctx *ww_ctx, |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 1089 | enum rtmutex_chainwalk chwalk) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1090 | { |
Ingo Molnar | 36c8b58 | 2006-07-03 00:25:41 -0700 | [diff] [blame] | 1091 | struct task_struct *owner = rt_mutex_owner(lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1092 | struct rt_mutex_waiter *top_waiter = waiter; |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 1093 | struct rt_mutex_base *next_lock; |
Steven Rostedt | db63063 | 2006-09-29 01:59:44 -0700 | [diff] [blame] | 1094 | int chain_walk = 0, res; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1095 | |
Peter Zijlstra | e0aad5b | 2017-03-23 15:56:13 +0100 | [diff] [blame] | 1096 | lockdep_assert_held(&lock->wait_lock); |
| 1097 | |
Thomas Gleixner | 397335f | 2014-05-22 03:25:39 +0000 | [diff] [blame] | 1098 | /* |
| 1099 | * Early deadlock detection. We really don't want the task to |
| 1100 | * enqueue on itself just to untangle the mess later. It's not |
| 1101 | * only an optimization. We drop the locks, so another waiter |
| 1102 | * can come in before the chain walk detects the deadlock. So |
| 1103 | * the other will detect the deadlock and return -EDEADLOCK, |
| 1104 | * which is wrong, as the other waiter is not in a deadlock |
| 1105 | * situation. |
Peter Zijlstra | 02ea9fc | 2021-11-29 18:46:46 +0100 | [diff] [blame] | 1106 | * |
| 1107 | * Except for ww_mutex, in that case the chain walk must already deal |
| 1108 | * with spurious cycles, see the comments at [3] and [6]. |
Thomas Gleixner | 397335f | 2014-05-22 03:25:39 +0000 | [diff] [blame] | 1109 | */ |
Peter Zijlstra | 02ea9fc | 2021-11-29 18:46:46 +0100 | [diff] [blame] | 1110 | if (owner == task && !(build_ww_mutex() && ww_ctx)) |
Thomas Gleixner | 397335f | 2014-05-22 03:25:39 +0000 | [diff] [blame] | 1111 | return -EDEADLK; |
| 1112 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1113 | raw_spin_lock(&task->pi_lock); |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1114 | waiter->task = task; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1115 | waiter->lock = lock; |
Peter Zijlstra | 715f7f9 | 2021-08-15 23:28:30 +0200 | [diff] [blame] | 1116 | waiter_update_prio(waiter, task); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1117 | |
| 1118 | /* Get the top priority waiter on the lock */ |
| 1119 | if (rt_mutex_has_waiters(lock)) |
| 1120 | top_waiter = rt_mutex_top_waiter(lock); |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 1121 | rt_mutex_enqueue(lock, waiter); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1122 | |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1123 | task->pi_blocked_on = waiter; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1124 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1125 | raw_spin_unlock(&task->pi_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1126 | |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1127 | if (build_ww_mutex() && ww_ctx) { |
| 1128 | struct rt_mutex *rtm; |
| 1129 | |
| 1130 | /* Check whether the waiter should back out immediately */ |
| 1131 | rtm = container_of(lock, struct rt_mutex, rtmutex); |
| 1132 | res = __ww_mutex_add_waiter(waiter, rtm, ww_ctx); |
Thomas Gleixner | 37e8abf | 2021-08-25 12:33:14 +0200 | [diff] [blame] | 1133 | if (res) { |
| 1134 | raw_spin_lock(&task->pi_lock); |
| 1135 | rt_mutex_dequeue(lock, waiter); |
| 1136 | task->pi_blocked_on = NULL; |
| 1137 | raw_spin_unlock(&task->pi_lock); |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1138 | return res; |
Thomas Gleixner | 37e8abf | 2021-08-25 12:33:14 +0200 | [diff] [blame] | 1139 | } |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1140 | } |
| 1141 | |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 1142 | if (!owner) |
| 1143 | return 0; |
| 1144 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1145 | raw_spin_lock(&owner->pi_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1146 | if (waiter == rt_mutex_top_waiter(lock)) { |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 1147 | rt_mutex_dequeue_pi(owner, top_waiter); |
| 1148 | rt_mutex_enqueue_pi(owner, waiter); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1149 | |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 1150 | rt_mutex_adjust_prio(owner); |
Steven Rostedt | db63063 | 2006-09-29 01:59:44 -0700 | [diff] [blame] | 1151 | if (owner->pi_blocked_on) |
| 1152 | chain_walk = 1; |
Thomas Gleixner | 8930ed8 | 2014-05-22 03:25:47 +0000 | [diff] [blame] | 1153 | } else if (rt_mutex_cond_detect_deadlock(waiter, chwalk)) { |
Steven Rostedt | db63063 | 2006-09-29 01:59:44 -0700 | [diff] [blame] | 1154 | chain_walk = 1; |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 1155 | } |
Steven Rostedt | db63063 | 2006-09-29 01:59:44 -0700 | [diff] [blame] | 1156 | |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 1157 | /* Store the lock on which owner is blocked or NULL */ |
| 1158 | next_lock = task_blocked_on_lock(owner); |
| 1159 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1160 | raw_spin_unlock(&owner->pi_lock); |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 1161 | /* |
| 1162 | * Even if full deadlock detection is on, if the owner is not |
| 1163 | * blocked itself, we can avoid finding this out in the chain |
| 1164 | * walk. |
| 1165 | */ |
| 1166 | if (!chain_walk || !next_lock) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1167 | return 0; |
| 1168 | |
Steven Rostedt | db63063 | 2006-09-29 01:59:44 -0700 | [diff] [blame] | 1169 | /* |
| 1170 | * The owner can't disappear while holding a lock, |
| 1171 | * so the owner struct is protected by wait_lock. |
| 1172 | * Gets dropped in rt_mutex_adjust_prio_chain()! |
| 1173 | */ |
| 1174 | get_task_struct(owner); |
| 1175 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1176 | raw_spin_unlock_irq(&lock->wait_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1177 | |
Thomas Gleixner | 8930ed8 | 2014-05-22 03:25:47 +0000 | [diff] [blame] | 1178 | res = rt_mutex_adjust_prio_chain(owner, chwalk, lock, |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 1179 | next_lock, waiter, task); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1180 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1181 | raw_spin_lock_irq(&lock->wait_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1182 | |
| 1183 | return res; |
| 1184 | } |
| 1185 | |
| 1186 | /* |
Davidlohr Bueso | 9f40a51 | 2015-05-19 10:24:57 -0700 | [diff] [blame] | 1187 | * Remove the top waiter from the current tasks pi waiter tree and |
Davidlohr Bueso | 45ab4ef | 2015-05-19 10:24:55 -0700 | [diff] [blame] | 1188 | * queue it up. |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1189 | * |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1190 | * Called with lock->wait_lock held and interrupts disabled. |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1191 | */ |
Thomas Gleixner | 7980aa3 | 2021-08-15 23:28:09 +0200 | [diff] [blame] | 1192 | static void __sched mark_wakeup_next_waiter(struct rt_wake_q_head *wqh, |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 1193 | struct rt_mutex_base *lock) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1194 | { |
| 1195 | struct rt_mutex_waiter *waiter; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1196 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1197 | raw_spin_lock(¤t->pi_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1198 | |
| 1199 | waiter = rt_mutex_top_waiter(lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1200 | |
| 1201 | /* |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 1202 | * Remove it from current->pi_waiters and deboost. |
| 1203 | * |
| 1204 | * We must in fact deboost here in order to ensure we call |
| 1205 | * rt_mutex_setprio() to update p->pi_top_task before the |
| 1206 | * task unblocks. |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1207 | */ |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 1208 | rt_mutex_dequeue_pi(current, waiter); |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 1209 | rt_mutex_adjust_prio(current); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1210 | |
Thomas Gleixner | 27e3571 | 2014-06-11 18:44:04 +0000 | [diff] [blame] | 1211 | /* |
| 1212 | * As we are waking up the top waiter, and the waiter stays |
| 1213 | * queued on the lock until it gets the lock, this lock |
| 1214 | * obviously has waiters. Just set the bit here and this has |
| 1215 | * the added benefit of forcing all new tasks into the |
| 1216 | * slow path making sure no task of lower priority than |
| 1217 | * the top waiter can steal this lock. |
| 1218 | */ |
| 1219 | lock->owner = (void *) RT_MUTEX_HAS_WAITERS; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1220 | |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 1221 | /* |
| 1222 | * We deboosted before waking the top waiter task such that we don't |
| 1223 | * run two tasks with the 'same' priority (and ensure the |
| 1224 | * p->pi_top_task pointer points to a blocked task). This however can |
| 1225 | * lead to priority inversion if we would get preempted after the |
| 1226 | * deboost but before waking our donor task, hence the preempt_disable() |
| 1227 | * before unlock. |
| 1228 | * |
Thomas Gleixner | 7980aa3 | 2021-08-15 23:28:09 +0200 | [diff] [blame] | 1229 | * Pairs with preempt_enable() in rt_mutex_wake_up_q(); |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 1230 | */ |
| 1231 | preempt_disable(); |
Thomas Gleixner | 7980aa3 | 2021-08-15 23:28:09 +0200 | [diff] [blame] | 1232 | rt_mutex_wake_q_add(wqh, waiter); |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 1233 | raw_spin_unlock(¤t->pi_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1234 | } |
| 1235 | |
Thomas Gleixner | e17ba59b | 2021-08-15 23:28:12 +0200 | [diff] [blame] | 1236 | static int __sched __rt_mutex_slowtrylock(struct rt_mutex_base *lock) |
| 1237 | { |
| 1238 | int ret = try_to_take_rt_mutex(lock, current, NULL); |
| 1239 | |
| 1240 | /* |
| 1241 | * try_to_take_rt_mutex() sets the lock waiters bit |
| 1242 | * unconditionally. Clean this up. |
| 1243 | */ |
| 1244 | fixup_rt_mutex_waiters(lock); |
| 1245 | |
| 1246 | return ret; |
| 1247 | } |
| 1248 | |
| 1249 | /* |
| 1250 | * Slow path try-lock function: |
| 1251 | */ |
| 1252 | static int __sched rt_mutex_slowtrylock(struct rt_mutex_base *lock) |
| 1253 | { |
| 1254 | unsigned long flags; |
| 1255 | int ret; |
| 1256 | |
| 1257 | /* |
| 1258 | * If the lock already has an owner we fail to get the lock. |
| 1259 | * This can be done without taking the @lock->wait_lock as |
| 1260 | * it is only being read, and this is a trylock anyway. |
| 1261 | */ |
| 1262 | if (rt_mutex_owner(lock)) |
| 1263 | return 0; |
| 1264 | |
| 1265 | /* |
| 1266 | * The mutex has currently no owner. Lock the wait lock and try to |
| 1267 | * acquire the lock. We use irqsave here to support early boot calls. |
| 1268 | */ |
| 1269 | raw_spin_lock_irqsave(&lock->wait_lock, flags); |
| 1270 | |
| 1271 | ret = __rt_mutex_slowtrylock(lock); |
| 1272 | |
| 1273 | raw_spin_unlock_irqrestore(&lock->wait_lock, flags); |
| 1274 | |
| 1275 | return ret; |
| 1276 | } |
| 1277 | |
| 1278 | static __always_inline int __rt_mutex_trylock(struct rt_mutex_base *lock) |
| 1279 | { |
| 1280 | if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) |
| 1281 | return 1; |
| 1282 | |
| 1283 | return rt_mutex_slowtrylock(lock); |
| 1284 | } |
| 1285 | |
| 1286 | /* |
| 1287 | * Slow path to release a rt-mutex. |
| 1288 | */ |
| 1289 | static void __sched rt_mutex_slowunlock(struct rt_mutex_base *lock) |
| 1290 | { |
| 1291 | DEFINE_RT_WAKE_Q(wqh); |
| 1292 | unsigned long flags; |
| 1293 | |
| 1294 | /* irqsave required to support early boot calls */ |
| 1295 | raw_spin_lock_irqsave(&lock->wait_lock, flags); |
| 1296 | |
| 1297 | debug_rt_mutex_unlock(lock); |
| 1298 | |
| 1299 | /* |
| 1300 | * We must be careful here if the fast path is enabled. If we |
| 1301 | * have no waiters queued we cannot set owner to NULL here |
| 1302 | * because of: |
| 1303 | * |
| 1304 | * foo->lock->owner = NULL; |
| 1305 | * rtmutex_lock(foo->lock); <- fast path |
| 1306 | * free = atomic_dec_and_test(foo->refcnt); |
| 1307 | * rtmutex_unlock(foo->lock); <- fast path |
| 1308 | * if (free) |
| 1309 | * kfree(foo); |
| 1310 | * raw_spin_unlock(foo->lock->wait_lock); |
| 1311 | * |
| 1312 | * So for the fastpath enabled kernel: |
| 1313 | * |
| 1314 | * Nothing can set the waiters bit as long as we hold |
| 1315 | * lock->wait_lock. So we do the following sequence: |
| 1316 | * |
| 1317 | * owner = rt_mutex_owner(lock); |
| 1318 | * clear_rt_mutex_waiters(lock); |
| 1319 | * raw_spin_unlock(&lock->wait_lock); |
| 1320 | * if (cmpxchg(&lock->owner, owner, 0) == owner) |
| 1321 | * return; |
| 1322 | * goto retry; |
| 1323 | * |
| 1324 | * The fastpath disabled variant is simple as all access to |
| 1325 | * lock->owner is serialized by lock->wait_lock: |
| 1326 | * |
| 1327 | * lock->owner = NULL; |
| 1328 | * raw_spin_unlock(&lock->wait_lock); |
| 1329 | */ |
| 1330 | while (!rt_mutex_has_waiters(lock)) { |
| 1331 | /* Drops lock->wait_lock ! */ |
| 1332 | if (unlock_rt_mutex_safe(lock, flags) == true) |
| 1333 | return; |
| 1334 | /* Relock the rtmutex and try again */ |
| 1335 | raw_spin_lock_irqsave(&lock->wait_lock, flags); |
| 1336 | } |
| 1337 | |
| 1338 | /* |
| 1339 | * The wakeup next waiter path does not suffer from the above |
| 1340 | * race. See the comments there. |
| 1341 | * |
| 1342 | * Queue the next waiter for wakeup once we release the wait_lock. |
| 1343 | */ |
| 1344 | mark_wakeup_next_waiter(&wqh, lock); |
| 1345 | raw_spin_unlock_irqrestore(&lock->wait_lock, flags); |
| 1346 | |
| 1347 | rt_mutex_wake_up_q(&wqh); |
| 1348 | } |
| 1349 | |
| 1350 | static __always_inline void __rt_mutex_unlock(struct rt_mutex_base *lock) |
| 1351 | { |
| 1352 | if (likely(rt_mutex_cmpxchg_release(lock, current, NULL))) |
| 1353 | return; |
| 1354 | |
| 1355 | rt_mutex_slowunlock(lock); |
| 1356 | } |
| 1357 | |
Steven Rostedt | 992caf7 | 2021-08-15 23:29:25 +0200 | [diff] [blame] | 1358 | #ifdef CONFIG_SMP |
| 1359 | static bool rtmutex_spin_on_owner(struct rt_mutex_base *lock, |
| 1360 | struct rt_mutex_waiter *waiter, |
| 1361 | struct task_struct *owner) |
| 1362 | { |
| 1363 | bool res = true; |
| 1364 | |
| 1365 | rcu_read_lock(); |
| 1366 | for (;;) { |
| 1367 | /* If owner changed, trylock again. */ |
| 1368 | if (owner != rt_mutex_owner(lock)) |
| 1369 | break; |
| 1370 | /* |
| 1371 | * Ensure that @owner is dereferenced after checking that |
| 1372 | * the lock owner still matches @owner. If that fails, |
| 1373 | * @owner might point to freed memory. If it still matches, |
| 1374 | * the rcu_read_lock() ensures the memory stays valid. |
| 1375 | */ |
| 1376 | barrier(); |
| 1377 | /* |
| 1378 | * Stop spinning when: |
| 1379 | * - the lock owner has been scheduled out |
| 1380 | * - current is not longer the top waiter |
| 1381 | * - current is requested to reschedule (redundant |
| 1382 | * for CONFIG_PREEMPT_RCU=y) |
| 1383 | * - the VCPU on which owner runs is preempted |
| 1384 | */ |
Kefeng Wang | c0bed69 | 2021-12-03 15:59:34 +0800 | [diff] [blame] | 1385 | if (!owner_on_cpu(owner) || need_resched() || |
Thomas Gleixner | f16cc98 | 2021-12-18 10:57:03 +0100 | [diff] [blame] | 1386 | !rt_mutex_waiter_is_top_waiter(lock, waiter)) { |
Steven Rostedt | 992caf7 | 2021-08-15 23:29:25 +0200 | [diff] [blame] | 1387 | res = false; |
| 1388 | break; |
| 1389 | } |
| 1390 | cpu_relax(); |
| 1391 | } |
| 1392 | rcu_read_unlock(); |
| 1393 | return res; |
| 1394 | } |
| 1395 | #else |
| 1396 | static bool rtmutex_spin_on_owner(struct rt_mutex_base *lock, |
| 1397 | struct rt_mutex_waiter *waiter, |
| 1398 | struct task_struct *owner) |
| 1399 | { |
| 1400 | return false; |
| 1401 | } |
| 1402 | #endif |
| 1403 | |
Thomas Gleixner | e17ba59b | 2021-08-15 23:28:12 +0200 | [diff] [blame] | 1404 | #ifdef RT_MUTEX_BUILD_MUTEX |
| 1405 | /* |
| 1406 | * Functions required for: |
| 1407 | * - rtmutex, futex on all kernels |
| 1408 | * - mutex and rwsem substitutions on RT kernels |
| 1409 | */ |
| 1410 | |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1411 | /* |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 1412 | * Remove a waiter from a lock and give up |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1413 | * |
Thomas Gleixner | e17ba59b | 2021-08-15 23:28:12 +0200 | [diff] [blame] | 1414 | * Must be called with lock->wait_lock held and interrupts disabled. It must |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 1415 | * have just failed to try_to_take_rt_mutex(). |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1416 | */ |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 1417 | static void __sched remove_waiter(struct rt_mutex_base *lock, |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 1418 | struct rt_mutex_waiter *waiter) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1419 | { |
Thomas Gleixner | 1ca7b86 | 2014-06-07 09:36:13 +0200 | [diff] [blame] | 1420 | bool is_top_waiter = (waiter == rt_mutex_top_waiter(lock)); |
Ingo Molnar | 36c8b58 | 2006-07-03 00:25:41 -0700 | [diff] [blame] | 1421 | struct task_struct *owner = rt_mutex_owner(lock); |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 1422 | struct rt_mutex_base *next_lock; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1423 | |
Peter Zijlstra | e0aad5b | 2017-03-23 15:56:13 +0100 | [diff] [blame] | 1424 | lockdep_assert_held(&lock->wait_lock); |
| 1425 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1426 | raw_spin_lock(¤t->pi_lock); |
Peter Zijlstra | fb00aca | 2013-11-07 14:43:43 +0100 | [diff] [blame] | 1427 | rt_mutex_dequeue(lock, waiter); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1428 | current->pi_blocked_on = NULL; |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1429 | raw_spin_unlock(¤t->pi_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1430 | |
Thomas Gleixner | 1ca7b86 | 2014-06-07 09:36:13 +0200 | [diff] [blame] | 1431 | /* |
| 1432 | * Only update priority if the waiter was the highest priority |
| 1433 | * waiter of the lock and there is an owner to update. |
| 1434 | */ |
| 1435 | if (!owner || !is_top_waiter) |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 1436 | return; |
| 1437 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1438 | raw_spin_lock(&owner->pi_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1439 | |
Thomas Gleixner | 1ca7b86 | 2014-06-07 09:36:13 +0200 | [diff] [blame] | 1440 | rt_mutex_dequeue_pi(owner, waiter); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1441 | |
Thomas Gleixner | 1ca7b86 | 2014-06-07 09:36:13 +0200 | [diff] [blame] | 1442 | if (rt_mutex_has_waiters(lock)) |
| 1443 | rt_mutex_enqueue_pi(owner, rt_mutex_top_waiter(lock)); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1444 | |
Peter Zijlstra | acd5862 | 2017-03-23 15:56:11 +0100 | [diff] [blame] | 1445 | rt_mutex_adjust_prio(owner); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1446 | |
Thomas Gleixner | 1ca7b86 | 2014-06-07 09:36:13 +0200 | [diff] [blame] | 1447 | /* Store the lock on which owner is blocked or NULL */ |
| 1448 | next_lock = task_blocked_on_lock(owner); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1449 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1450 | raw_spin_unlock(&owner->pi_lock); |
Steven Rostedt | db63063 | 2006-09-29 01:59:44 -0700 | [diff] [blame] | 1451 | |
Thomas Gleixner | 1ca7b86 | 2014-06-07 09:36:13 +0200 | [diff] [blame] | 1452 | /* |
| 1453 | * Don't walk the chain, if the owner task is not blocked |
| 1454 | * itself. |
| 1455 | */ |
Thomas Gleixner | 8208498 | 2014-06-05 11:16:12 +0200 | [diff] [blame] | 1456 | if (!next_lock) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1457 | return; |
| 1458 | |
Steven Rostedt | db63063 | 2006-09-29 01:59:44 -0700 | [diff] [blame] | 1459 | /* gets dropped in rt_mutex_adjust_prio_chain()! */ |
| 1460 | get_task_struct(owner); |
| 1461 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1462 | raw_spin_unlock_irq(&lock->wait_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1463 | |
Thomas Gleixner | 8930ed8 | 2014-05-22 03:25:47 +0000 | [diff] [blame] | 1464 | rt_mutex_adjust_prio_chain(owner, RT_MUTEX_MIN_CHAINWALK, lock, |
| 1465 | next_lock, NULL, current); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1466 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1467 | raw_spin_lock_irq(&lock->wait_lock); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1468 | } |
| 1469 | |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1470 | /** |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1471 | * rt_mutex_slowlock_block() - Perform the wait-wake-try-to-take loop |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1472 | * @lock: the rt_mutex to take |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1473 | * @ww_ctx: WW mutex context pointer |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1474 | * @state: the state the task should block in (TASK_INTERRUPTIBLE |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1475 | * or TASK_UNINTERRUPTIBLE) |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1476 | * @timeout: the pre-initialized and started timer, or NULL for none |
| 1477 | * @waiter: the pre-initialized rt_mutex_waiter |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1478 | * |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1479 | * Must be called with lock->wait_lock held and interrupts disabled |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1480 | */ |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1481 | static int __sched rt_mutex_slowlock_block(struct rt_mutex_base *lock, |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1482 | struct ww_acquire_ctx *ww_ctx, |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1483 | unsigned int state, |
| 1484 | struct hrtimer_sleeper *timeout, |
| 1485 | struct rt_mutex_waiter *waiter) |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1486 | { |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1487 | struct rt_mutex *rtm = container_of(lock, struct rt_mutex, rtmutex); |
Steven Rostedt | 992caf7 | 2021-08-15 23:29:25 +0200 | [diff] [blame] | 1488 | struct task_struct *owner; |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1489 | int ret = 0; |
| 1490 | |
| 1491 | for (;;) { |
| 1492 | /* Try to acquire the lock: */ |
Lai Jiangshan | 8161239 | 2011-01-14 17:09:41 +0800 | [diff] [blame] | 1493 | if (try_to_take_rt_mutex(lock, current, waiter)) |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1494 | break; |
| 1495 | |
Thomas Gleixner | a51a327 | 2021-03-26 16:29:44 +0100 | [diff] [blame] | 1496 | if (timeout && !timeout->task) { |
| 1497 | ret = -ETIMEDOUT; |
| 1498 | break; |
| 1499 | } |
| 1500 | if (signal_pending_state(state, current)) { |
| 1501 | ret = -EINTR; |
| 1502 | break; |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1503 | } |
| 1504 | |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1505 | if (build_ww_mutex() && ww_ctx) { |
| 1506 | ret = __ww_mutex_check_kill(rtm, waiter, ww_ctx); |
| 1507 | if (ret) |
| 1508 | break; |
| 1509 | } |
| 1510 | |
Steven Rostedt | 992caf7 | 2021-08-15 23:29:25 +0200 | [diff] [blame] | 1511 | if (waiter == rt_mutex_top_waiter(lock)) |
| 1512 | owner = rt_mutex_owner(lock); |
| 1513 | else |
| 1514 | owner = NULL; |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1515 | raw_spin_unlock_irq(&lock->wait_lock); |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1516 | |
Steven Rostedt | 992caf7 | 2021-08-15 23:29:25 +0200 | [diff] [blame] | 1517 | if (!owner || !rtmutex_spin_on_owner(lock, waiter, owner)) |
| 1518 | schedule(); |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1519 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1520 | raw_spin_lock_irq(&lock->wait_lock); |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1521 | set_current_state(state); |
| 1522 | } |
| 1523 | |
Davidlohr Bueso | afffc6c | 2015-02-01 22:16:24 -0800 | [diff] [blame] | 1524 | __set_current_state(TASK_RUNNING); |
Darren Hart | 8dac456 | 2009-04-03 13:40:12 -0700 | [diff] [blame] | 1525 | return ret; |
| 1526 | } |
| 1527 | |
Thomas Gleixner | d7a2edb | 2021-03-26 16:29:40 +0100 | [diff] [blame] | 1528 | static void __sched rt_mutex_handle_deadlock(int res, int detect_deadlock, |
| 1529 | struct rt_mutex_waiter *w) |
Thomas Gleixner | 3d5c934 | 2014-06-05 12:34:23 +0200 | [diff] [blame] | 1530 | { |
| 1531 | /* |
| 1532 | * If the result is not -EDEADLOCK or the caller requested |
| 1533 | * deadlock detection, nothing to do here. |
| 1534 | */ |
| 1535 | if (res != -EDEADLOCK || detect_deadlock) |
| 1536 | return; |
| 1537 | |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1538 | if (build_ww_mutex() && w->ww_ctx) |
| 1539 | return; |
| 1540 | |
Thomas Gleixner | 3d5c934 | 2014-06-05 12:34:23 +0200 | [diff] [blame] | 1541 | /* |
Ingo Molnar | e2db759 | 2021-03-22 02:35:05 +0100 | [diff] [blame] | 1542 | * Yell loudly and stop the task right here. |
Thomas Gleixner | 3d5c934 | 2014-06-05 12:34:23 +0200 | [diff] [blame] | 1543 | */ |
Sebastian Andrzej Siewior | 6d41c67 | 2021-03-26 16:29:32 +0100 | [diff] [blame] | 1544 | WARN(1, "rtmutex deadlock detected\n"); |
Thomas Gleixner | 3d5c934 | 2014-06-05 12:34:23 +0200 | [diff] [blame] | 1545 | while (1) { |
| 1546 | set_current_state(TASK_INTERRUPTIBLE); |
| 1547 | schedule(); |
| 1548 | } |
| 1549 | } |
| 1550 | |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1551 | /** |
| 1552 | * __rt_mutex_slowlock - Locking slowpath invoked with lock::wait_lock held |
| 1553 | * @lock: The rtmutex to block lock |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1554 | * @ww_ctx: WW mutex context pointer |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1555 | * @state: The task state for sleeping |
| 1556 | * @chwalk: Indicator whether full or partial chainwalk is requested |
| 1557 | * @waiter: Initializer waiter for blocking |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1558 | */ |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1559 | static int __sched __rt_mutex_slowlock(struct rt_mutex_base *lock, |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1560 | struct ww_acquire_ctx *ww_ctx, |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1561 | unsigned int state, |
| 1562 | enum rtmutex_chainwalk chwalk, |
| 1563 | struct rt_mutex_waiter *waiter) |
| 1564 | { |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1565 | struct rt_mutex *rtm = container_of(lock, struct rt_mutex, rtmutex); |
| 1566 | struct ww_mutex *ww = ww_container_of(rtm); |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1567 | int ret; |
| 1568 | |
| 1569 | lockdep_assert_held(&lock->wait_lock); |
| 1570 | |
| 1571 | /* Try to acquire the lock again: */ |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1572 | if (try_to_take_rt_mutex(lock, current, NULL)) { |
| 1573 | if (build_ww_mutex() && ww_ctx) { |
| 1574 | __ww_mutex_check_waiters(rtm, ww_ctx); |
| 1575 | ww_mutex_lock_acquired(ww, ww_ctx); |
| 1576 | } |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1577 | return 0; |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1578 | } |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1579 | |
| 1580 | set_current_state(state); |
| 1581 | |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1582 | ret = task_blocks_on_rt_mutex(lock, waiter, current, ww_ctx, chwalk); |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1583 | if (likely(!ret)) |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1584 | ret = rt_mutex_slowlock_block(lock, ww_ctx, state, NULL, waiter); |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1585 | |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1586 | if (likely(!ret)) { |
| 1587 | /* acquired the lock */ |
| 1588 | if (build_ww_mutex() && ww_ctx) { |
| 1589 | if (!ww_ctx->is_wait_die) |
| 1590 | __ww_mutex_check_waiters(rtm, ww_ctx); |
| 1591 | ww_mutex_lock_acquired(ww, ww_ctx); |
| 1592 | } |
| 1593 | } else { |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1594 | __set_current_state(TASK_RUNNING); |
| 1595 | remove_waiter(lock, waiter); |
| 1596 | rt_mutex_handle_deadlock(ret, chwalk, waiter); |
| 1597 | } |
| 1598 | |
| 1599 | /* |
| 1600 | * try_to_take_rt_mutex() sets the waiter bit |
| 1601 | * unconditionally. We might have to fix that up. |
| 1602 | */ |
| 1603 | fixup_rt_mutex_waiters(lock); |
| 1604 | return ret; |
| 1605 | } |
| 1606 | |
| 1607 | static inline int __rt_mutex_slowlock_locked(struct rt_mutex_base *lock, |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1608 | struct ww_acquire_ctx *ww_ctx, |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1609 | unsigned int state) |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1610 | { |
| 1611 | struct rt_mutex_waiter waiter; |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1612 | int ret; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1613 | |
Peter Zijlstra | 5080935 | 2017-03-22 11:35:56 +0100 | [diff] [blame] | 1614 | rt_mutex_init_waiter(&waiter); |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1615 | waiter.ww_ctx = ww_ctx; |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1616 | |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1617 | ret = __rt_mutex_slowlock(lock, ww_ctx, state, RT_MUTEX_MIN_CHAINWALK, |
| 1618 | &waiter); |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1619 | |
| 1620 | debug_rt_mutex_free_waiter(&waiter); |
| 1621 | return ret; |
| 1622 | } |
| 1623 | |
| 1624 | /* |
| 1625 | * rt_mutex_slowlock - Locking slowpath invoked when fast path fails |
| 1626 | * @lock: The rtmutex to block lock |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1627 | * @ww_ctx: WW mutex context pointer |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1628 | * @state: The task state for sleeping |
| 1629 | */ |
| 1630 | static int __sched rt_mutex_slowlock(struct rt_mutex_base *lock, |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1631 | struct ww_acquire_ctx *ww_ctx, |
Thomas Gleixner | ebbdc41 | 2021-08-15 23:28:00 +0200 | [diff] [blame] | 1632 | unsigned int state) |
| 1633 | { |
| 1634 | unsigned long flags; |
| 1635 | int ret; |
| 1636 | |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1637 | /* |
| 1638 | * Technically we could use raw_spin_[un]lock_irq() here, but this can |
| 1639 | * be called in early boot if the cmpxchg() fast path is disabled |
| 1640 | * (debug, no architecture support). In this case we will acquire the |
| 1641 | * rtmutex with lock->wait_lock held. But we cannot unconditionally |
| 1642 | * enable interrupts in that early boot case. So we need to use the |
| 1643 | * irqsave/restore variants. |
| 1644 | */ |
| 1645 | raw_spin_lock_irqsave(&lock->wait_lock, flags); |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1646 | ret = __rt_mutex_slowlock_locked(lock, ww_ctx, state); |
Thomas Gleixner | b4abf91 | 2016-01-13 11:25:38 +0100 | [diff] [blame] | 1647 | raw_spin_unlock_irqrestore(&lock->wait_lock, flags); |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1648 | |
Ingo Molnar | 23f78d4a | 2006-06-27 02:54:53 -0700 | [diff] [blame] | 1649 | return ret; |
| 1650 | } |
| 1651 | |
Peter Zijlstra | 830e6ac | 2021-08-15 23:27:58 +0200 | [diff] [blame] | 1652 | static __always_inline int __rt_mutex_lock(struct rt_mutex_base *lock, |
Thomas Gleixner | 531ae4b | 2021-08-15 23:27:57 +0200 | [diff] [blame] | 1653 | unsigned int state) |
| 1654 | { |
| 1655 | if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current))) |
| 1656 | return 0; |
| 1657 | |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1658 | return rt_mutex_slowlock(lock, NULL, state); |
Thomas Gleixner | 531ae4b | 2021-08-15 23:27:57 +0200 | [diff] [blame] | 1659 | } |
Thomas Gleixner | e17ba59b | 2021-08-15 23:28:12 +0200 | [diff] [blame] | 1660 | #endif /* RT_MUTEX_BUILD_MUTEX */ |
Thomas Gleixner | 1c143c4 | 2021-08-15 23:28:25 +0200 | [diff] [blame] | 1661 | |
| 1662 | #ifdef RT_MUTEX_BUILD_SPINLOCKS |
| 1663 | /* |
| 1664 | * Functions required for spin/rw_lock substitution on RT kernels |
| 1665 | */ |
| 1666 | |
| 1667 | /** |
| 1668 | * rtlock_slowlock_locked - Slow path lock acquisition for RT locks |
| 1669 | * @lock: The underlying RT mutex |
| 1670 | */ |
| 1671 | static void __sched rtlock_slowlock_locked(struct rt_mutex_base *lock) |
| 1672 | { |
| 1673 | struct rt_mutex_waiter waiter; |
Steven Rostedt | 992caf7 | 2021-08-15 23:29:25 +0200 | [diff] [blame] | 1674 | struct task_struct *owner; |
Thomas Gleixner | 1c143c4 | 2021-08-15 23:28:25 +0200 | [diff] [blame] | 1675 | |
| 1676 | lockdep_assert_held(&lock->wait_lock); |
| 1677 | |
| 1678 | if (try_to_take_rt_mutex(lock, current, NULL)) |
| 1679 | return; |
| 1680 | |
| 1681 | rt_mutex_init_rtlock_waiter(&waiter); |
| 1682 | |
| 1683 | /* Save current state and set state to TASK_RTLOCK_WAIT */ |
| 1684 | current_save_and_set_rtlock_wait_state(); |
| 1685 | |
Peter Zijlstra | add4613 | 2021-08-15 23:28:58 +0200 | [diff] [blame] | 1686 | task_blocks_on_rt_mutex(lock, &waiter, current, NULL, RT_MUTEX_MIN_CHAINWALK); |
Thomas Gleixner | 1c143c4 | 2021-08-15 23:28:25 +0200 | [diff] [blame] | 1687 | |
| 1688 | for (;;) { |
| 1689 | /* Try to acquire the lock again */ |
| 1690 | if (try_to_take_rt_mutex(lock, current, &waiter)) |
| 1691 | break; |
| 1692 | |
Steven Rostedt | 992caf7 | 2021-08-15 23:29:25 +0200 | [diff] [blame] | 1693 | if (&waiter == rt_mutex_top_waiter(lock)) |
| 1694 | owner = rt_mutex_owner(lock); |
| 1695 | else |
| 1696 | owner = NULL; |
Thomas Gleixner | 1c143c4 | 2021-08-15 23:28:25 +0200 | [diff] [blame] | 1697 | raw_spin_unlock_irq(&lock->wait_lock); |
| 1698 | |
Steven Rostedt | 992caf7 | 2021-08-15 23:29:25 +0200 | [diff] [blame] | 1699 | if (!owner || !rtmutex_spin_on_owner(lock, &waiter, owner)) |
| 1700 | schedule_rtlock(); |
Thomas Gleixner | 1c143c4 | 2021-08-15 23:28:25 +0200 | [diff] [blame] | 1701 | |
| 1702 | raw_spin_lock_irq(&lock->wait_lock); |
| 1703 | set_current_state(TASK_RTLOCK_WAIT); |
| 1704 | } |
| 1705 | |
| 1706 | /* Restore the task state */ |
| 1707 | current_restore_rtlock_saved_state(); |
| 1708 | |
| 1709 | /* |
| 1710 | * try_to_take_rt_mutex() sets the waiter bit unconditionally. |
| 1711 | * We might have to fix that up: |
| 1712 | */ |
| 1713 | fixup_rt_mutex_waiters(lock); |
| 1714 | debug_rt_mutex_free_waiter(&waiter); |
| 1715 | } |
| 1716 | |
| 1717 | static __always_inline void __sched rtlock_slowlock(struct rt_mutex_base *lock) |
| 1718 | { |
| 1719 | unsigned long flags; |
| 1720 | |
| 1721 | raw_spin_lock_irqsave(&lock->wait_lock, flags); |
| 1722 | rtlock_slowlock_locked(lock); |
| 1723 | raw_spin_unlock_irqrestore(&lock->wait_lock, flags); |
| 1724 | } |
| 1725 | |
| 1726 | #endif /* RT_MUTEX_BUILD_SPINLOCKS */ |