[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <1480335612-12069-7-git-send-email-nhaehnle@gmail.com>
Date: Mon, 28 Nov 2016 13:20:07 +0100
From: Nicolai Hähnle <nhaehnle@...il.com>
To: linux-kernel@...r.kernel.org
Cc: Nicolai Hähnle <Nicolai.Haehnle@....com>,
Peter Zijlstra <peterz@...radead.org>,
Ingo Molnar <mingo@...hat.com>,
Maarten Lankhorst <dev@...ankhorst.nl>,
Daniel Vetter <daniel@...ll.ch>,
Chris Wilson <chris@...is-wilson.co.uk>,
dri-devel@...ts.freedesktop.org
Subject: [PATCH 06/11] locking/ww_mutex: Notify waiters that have to back off while adding tasks to wait list
From: Nicolai Hähnle <Nicolai.Haehnle@....com>
While adding our task as a waiter, detect if another task should back off
because of us.
With this patch, we establish the invariant that the wait list contains
at most one (sleeping) waiter with ww_ctx->acquired > 0, and this waiter
will be the first waiter with a context.
Since only waiters with ww_ctx->acquired > 0 have to back off, this allows
us to be much more economical with wakeups.
Cc: Peter Zijlstra <peterz@...radead.org>
Cc: Ingo Molnar <mingo@...hat.com>
Cc: Maarten Lankhorst <dev@...ankhorst.nl>
Cc: Daniel Vetter <daniel@...ll.ch>
Cc: Chris Wilson <chris@...is-wilson.co.uk>
Cc: dri-devel@...ts.freedesktop.org
Signed-off-by: Nicolai Hähnle <Nicolai.Haehnle@....com>
---
kernel/locking/mutex.c | 42 ++++++++++++++++++++++++++++++++----------
1 file changed, 32 insertions(+), 10 deletions(-)
diff --git a/kernel/locking/mutex.c b/kernel/locking/mutex.c
index 01dcae7..d310703e 100644
--- a/kernel/locking/mutex.c
+++ b/kernel/locking/mutex.c
@@ -609,23 +609,34 @@ void __sched ww_mutex_unlock(struct ww_mutex *lock)
EXPORT_SYMBOL(ww_mutex_unlock);
static inline int __sched
-__ww_mutex_lock_check_stamp(struct mutex *lock, struct ww_acquire_ctx *ctx)
+__ww_mutex_lock_check_stamp(struct mutex *lock, struct mutex_waiter *waiter,
+ struct ww_acquire_ctx *ctx)
{
struct ww_mutex *ww = container_of(lock, struct ww_mutex, base);
struct ww_acquire_ctx *hold_ctx = READ_ONCE(ww->ctx);
+ struct mutex_waiter *cur;
- if (!hold_ctx)
- return 0;
+ if (hold_ctx && __ww_mutex_stamp_after(ctx, hold_ctx))
+ goto deadlock;
- if (__ww_mutex_stamp_after(ctx, hold_ctx)) {
-#ifdef CONFIG_DEBUG_MUTEXES
- DEBUG_LOCKS_WARN_ON(ctx->contending_lock);
- ctx->contending_lock = ww;
-#endif
- return -EDEADLK;
+ /*
+ * If there is a waiter in front of us that has a context, then its
+ * stamp is earlier than ours and we must back off.
+ */
+ cur = waiter;
+ list_for_each_entry_continue_reverse(cur, &lock->wait_list, list) {
+ if (cur->ww_ctx)
+ goto deadlock;
}
return 0;
+
+deadlock:
+#ifdef CONFIG_DEBUG_MUTEXES
+ DEBUG_LOCKS_WARN_ON(ctx->contending_lock);
+ ctx->contending_lock = ww;
+#endif
+ return -EDEADLK;
}
static inline int __sched
@@ -654,6 +665,16 @@ __ww_mutex_add_waiter(struct mutex_waiter *waiter,
}
list_add_tail(&waiter->list, &cur->list);
+
+ /*
+ * Wake up the waiter so that it gets a chance to back
+ * off.
+ */
+ if (cur->ww_ctx->acquired > 0) {
+ debug_mutex_wake_waiter(lock, cur);
+ wake_up_process(cur->task);
+ }
+
return 0;
}
}
@@ -756,7 +777,8 @@ __mutex_lock_common(struct mutex *lock, long state, unsigned int subclass,
}
if (use_ww_ctx && ww_ctx && ww_ctx->acquired > 0) {
- ret = __ww_mutex_lock_check_stamp(lock, ww_ctx);
+ ret = __ww_mutex_lock_check_stamp(lock, &waiter,
+ ww_ctx);
if (ret)
goto err;
}
--
2.7.4
Powered by blists - more mailing lists