search for: ticket_slowpath_flag

...rt spinlock clears slowpath flag after doing unlock. As explained by Linus currently it does: prev = *lock; add_smp(&lock->tickets.head, TICKET_LOCK_INC); /* add_smp() is a full mb() */ if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) __ticket_unlock_slowpath(lock, prev); which is *exactly* the kind of things you cannot do with spinlocks, because after you've done the "add_smp()" and released the spinlock for the fast-path, you can't access the spinlock any more. Exactly because a f...

...rt spinlock clears slowpath flag after doing unlock. As explained by Linus currently it does: prev = *lock; add_smp(&lock->tickets.head, TICKET_LOCK_INC); /* add_smp() is a full mb() */ if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) __ticket_unlock_slowpath(lock, prev); which is *exactly* the kind of things you cannot do with spinlocks, because after you've done the "add_smp()" and released the spinlock for the fast-path, you can't access the spinlock any more. Exactly because a f...

On Tue, Feb 10, 2015 at 10:30 AM, Raghavendra K T <raghavendra.kt at linux.vnet.ibm.com> wrote: > On 02/10/2015 06:23 AM, Linus Torvalds wrote: >> add_smp(&lock->tickets.head, TICKET_LOCK_INC); >> if (READ_ONCE(lock->tickets.tail) & TICKET_SLOWPATH_FLAG) .. >> >> into something like >> >> val = xadd((&lock->ticket.head_tail, TICKET_LOCK_INC << >> TICKET_SHIFT); >> if (unlikely(val & TICKET_SLOWPATH_FLAG)) ... >> >> would be the right thing to do. Somebody should jus...

On Tue, Feb 10, 2015 at 10:30 AM, Raghavendra K T <raghavendra.kt at linux.vnet.ibm.com> wrote: > On 02/10/2015 06:23 AM, Linus Torvalds wrote: >> add_smp(&lock->tickets.head, TICKET_LOCK_INC); >> if (READ_ONCE(lock->tickets.tail) & TICKET_SLOWPATH_FLAG) .. >> >> into something like >> >> val = xadd((&lock->ticket.head_tail, TICKET_LOCK_INC << >> TICKET_SHIFT); >> if (unlikely(val & TICKET_SLOWPATH_FLAG)) ... >> >> would be the right thing to do. Somebody should jus...

...rt spinlock clears slowpath flag after doing unlock. As explained by Linus currently it does: prev = *lock; add_smp(&lock->tickets.head, TICKET_LOCK_INC); /* add_smp() is a full mb() */ if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) __ticket_unlock_slowpath(lock, prev); which is *exactly* the kind of things you cannot do with spinlocks, because after you've done the "add_smp()" and released the spinlock for the fast-path, you can't access the spinlock any more. Exactly because a f...

...rt spinlock clears slowpath flag after doing unlock. As explained by Linus currently it does: prev = *lock; add_smp(&lock->tickets.head, TICKET_LOCK_INC); /* add_smp() is a full mb() */ if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) __ticket_unlock_slowpath(lock, prev); which is *exactly* the kind of things you cannot do with spinlocks, because after you've done the "add_smp()" and released the spinlock for the fast-path, you can't access the spinlock any more. Exactly because a f...

...rt spinlock clears slowpath flag after doing unlock. As explained by Linus currently it does: prev = *lock; add_smp(&lock->tickets.head, TICKET_LOCK_INC); /* add_smp() is a full mb() */ if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) __ticket_unlock_slowpath(lock, prev); which is *exactly* the kind of things you cannot do with spinlocks, because after you've done the "add_smp()" and released the spinlock for the fast-path, you can't access the spinlock any more. Exactly because a f...

...rt spinlock clears slowpath flag after doing unlock. As explained by Linus currently it does: prev = *lock; add_smp(&lock->tickets.head, TICKET_LOCK_INC); /* add_smp() is a full mb() */ if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) __ticket_unlock_slowpath(lock, prev); which is *exactly* the kind of things you cannot do with spinlocks, because after you've done the "add_smp()" and released the spinlock for the fast-path, you can't access the spinlock any more. Exactly because a f...

...the reasons we dropped support for original 80386. So I think Raghavendra's last version (which hopefully fixes the lockup problem that Sasha reported) together with changing that add_smp(&lock->tickets.head, TICKET_LOCK_INC); if (READ_ONCE(lock->tickets.tail) & TICKET_SLOWPATH_FLAG) .. into something like val = xadd((&lock->ticket.head_tail, TICKET_LOCK_INC << TICKET_SHIFT); if (unlikely(val & TICKET_SLOWPATH_FLAG)) ... would be the right thing to do. Somebody should just check that I got that shift right, and that the tail is in the high b...

...the reasons we dropped support for original 80386. So I think Raghavendra's last version (which hopefully fixes the lockup problem that Sasha reported) together with changing that add_smp(&lock->tickets.head, TICKET_LOCK_INC); if (READ_ONCE(lock->tickets.tail) & TICKET_SLOWPATH_FLAG) .. into something like val = xadd((&lock->ticket.head_tail, TICKET_LOCK_INC << TICKET_SHIFT); if (unlikely(val & TICKET_SLOWPATH_FLAG)) ... would be the right thing to do. Somebody should just check that I got that shift right, and that the tail is in the high b...

...after doing unlock. > As explained by Linus currently it does: > prev = *lock; > add_smp(&lock->tickets.head, TICKET_LOCK_INC); > > /* add_smp() is a full mb() */ > > if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) > __ticket_unlock_slowpath(lock, prev); > > > which is *exactly* the kind of things you cannot do with spinlocks, > because after you've done the "add_smp()" and released the spinlock > for the fast-path, you can't access the spinlock an...

...after doing unlock. > As explained by Linus currently it does: > prev = *lock; > add_smp(&lock->tickets.head, TICKET_LOCK_INC); > > /* add_smp() is a full mb() */ > > if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) > __ticket_unlock_slowpath(lock, prev); > > > which is *exactly* the kind of things you cannot do with spinlocks, > because after you've done the "add_smp()" and released the spinlock > for the fast-path, you can't access the spinlock an...

...include/asm/spinlock_types.h index 5f9d757..579409a 100644 --- a/arch/x86/include/asm/spinlock_types.h +++ b/arch/x86/include/asm/spinlock_types.h @@ -4,8 +4,9 @@ #include <linux/types.h> #ifdef CONFIG_PARAVIRT_SPINLOCKS +#define TICKET_SLOWPATH_BIT 0 #define __TICKET_LOCK_INC 2 -#define TICKET_SLOWPATH_FLAG ((__ticket_t)1) +#define TICKET_SLOWPATH_FLAG ((__ticket_t)(1 << TICKET_SLOWPATH_BIT)) #else #define __TICKET_LOCK_INC 1 #define TICKET_SLOWPATH_FLAG ((__ticket_t)0) -- 2.1.4

...rt spinlock clears slowpath flag after doing unlock. As explained by Linus currently it does: prev = *lock; add_smp(&lock->tickets.head, TICKET_LOCK_INC); /* add_smp() is a full mb() */ if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) __ticket_unlock_slowpath(lock, prev); which is *exactly* the kind of things you cannot do with spinlocks, because after you've done the "add_smp()" and released the spinlock for the fast-path, you can't access the spinlock any more. Exactly because a fa...

...rt spinlock clears slowpath flag after doing unlock. As explained by Linus currently it does: prev = *lock; add_smp(&lock->tickets.head, TICKET_LOCK_INC); /* add_smp() is a full mb() */ if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) __ticket_unlock_slowpath(lock, prev); which is *exactly* the kind of things you cannot do with spinlocks, because after you've done the "add_smp()" and released the spinlock for the fast-path, you can't access the spinlock any more. Exactly because a fa...

On 02/10, Raghavendra K T wrote: > > On 02/10/2015 06:23 AM, Linus Torvalds wrote: > >> add_smp(&lock->tickets.head, TICKET_LOCK_INC); >> if (READ_ONCE(lock->tickets.tail) & TICKET_SLOWPATH_FLAG) .. >> >> into something like >> >> val = xadd((&lock->ticket.head_tail, TICKET_LOCK_INC << TICKET_SHIFT); >> if (unlikely(val & TICKET_SLOWPATH_FLAG)) ... >> >> would be the right thing to do. Somebody should just check t...

On 02/10, Raghavendra K T wrote: > > On 02/10/2015 06:23 AM, Linus Torvalds wrote: > >> add_smp(&lock->tickets.head, TICKET_LOCK_INC); >> if (READ_ONCE(lock->tickets.tail) & TICKET_SLOWPATH_FLAG) .. >> >> into something like >> >> val = xadd((&lock->ticket.head_tail, TICKET_LOCK_INC << TICKET_SHIFT); >> if (unlikely(val & TICKET_SLOWPATH_FLAG)) ... >> >> would be the right thing to do. Somebody should just check t...

...rt spinlock clears slowpath flag after doing unlock. As explained by Linus currently it does: prev = *lock; add_smp(&lock->tickets.head, TICKET_LOCK_INC); /* add_smp() is a full mb() */ if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) __ticket_unlock_slowpath(lock, prev); which is *exactly* the kind of things you cannot do with spinlocks, because after you've done the "add_smp()" and released the spinlock for the fast-path, you can't access the spinlock any more. Exactly because a fa...

...rt spinlock clears slowpath flag after doing unlock. As explained by Linus currently it does: prev = *lock; add_smp(&lock->tickets.head, TICKET_LOCK_INC); /* add_smp() is a full mb() */ if (unlikely(lock->tickets.tail & TICKET_SLOWPATH_FLAG)) __ticket_unlock_slowpath(lock, prev); which is *exactly* the kind of things you cannot do with spinlocks, because after you've done the "add_smp()" and released the spinlock for the fast-path, you can't access the spinlock any more. Exactly because a fa...

...rect, but apart from this, before doing xadd in unlock, > we would have to make sure lsb bit is cleared so that we can live with 1 > bit overflow to tail which is unused. now either or both of head,tail > lsb bit may be set after unlock. Sorry, can't understand... could you spell? If TICKET_SLOWPATH_FLAG lives in .head arch_spin_unlock() could simply do head = xadd(&lock->tickets.head, TICKET_LOCK_INC); if (head & TICKET_SLOWPATH_FLAG) __ticket_unlock_kick(head); so it can't overflow to .tail? But probably I missed your concern. And we we do this, probably it makes sense t...