[3/3] bitops.h: set_mask_bits() to return old value
diff mbox series

Message ID 1547166387-19785-4-git-send-email-vgupta@synopsys.com
State New
Headers show
Series
  • Replace opencoded set_mask_bits
Related show

Commit Message

Vineet Gupta Jan. 11, 2019, 12:26 a.m. UTC
| > Also, set_mask_bits is used in fs quite a bit and we can possibly come up
| > with a generic llsc based implementation (w/o the cmpxchg loop)
|
| May I also suggest changing the return value of set_mask_bits() to old.
|
| You can compute the new value given old, but you cannot compute the old
| value given new, therefore old is the better return value. Also, no
| current user seems to use the return value, so changing it is without
| risk.

Link: http://lkml.kernel.org/g/20150807110955.GH16853@twins.programming.kicks-ass.net
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Cc: Miklos Szeredi <mszeredi@redhat.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Will Deacon <will.deacon@arm.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 include/linux/bitops.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Anthony Yznaga Jan. 11, 2019, 4:25 a.m. UTC | #1
On 1/10/19 4:26 PM, Vineet Gupta wrote:
> | > Also, set_mask_bits is used in fs quite a bit and we can possibly come up
> | > with a generic llsc based implementation (w/o the cmpxchg loop)
> |
> | May I also suggest changing the return value of set_mask_bits() to old.
> |
> | You can compute the new value given old, but you cannot compute the old
> | value given new, therefore old is the better return value. Also, no
> | current user seems to use the return value, so changing it is without
> | risk.
>
> Link: http://lkml.kernel.org/g/20150807110955.GH16853@twins.programming.kicks-ass.net
> Suggested-by: Peter Zijlstra <peterz@infradead.org>
> Cc: Miklos Szeredi <mszeredi@redhat.com>
> Cc: Ingo Molnar <mingo@kernel.org>
> Cc: Jani Nikula <jani.nikula@intel.com>
> Cc: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> Cc: Will Deacon <will.deacon@arm.com>
> Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
>

Reviewed-by: Anthony Yznaga <anthony.yznaga@oracle.com>
Peter Zijlstra Jan. 11, 2019, 9:24 a.m. UTC | #2
On Thu, Jan 10, 2019 at 04:26:27PM -0800, Vineet Gupta wrote:

> @@ -246,7 +246,7 @@ static __always_inline void __assign_bit(long nr, volatile unsigned long *addr,
>  		new__ = (old__ & ~mask__) | bits__;		\
>  	} while (cmpxchg(ptr, old__, new__) != old__);		\

diff --git a/include/linux/bitops.h b/include/linux/bitops.h
index 705f7c442691..2060d26a35f5 100644
--- a/include/linux/bitops.h
+++ b/include/linux/bitops.h
@@ -241,10 +241,10 @@ static __always_inline void __assign_bit(long nr, volatile unsigned long *addr,
 	const typeof(*(ptr)) mask__ = (mask), bits__ = (bits);	\
 	typeof(*(ptr)) old__, new__;				\
 								\
+	old__ = READ_ONCE(*(ptr));				\
 	do {							\
-		old__ = READ_ONCE(*(ptr));			\
 		new__ = (old__ & ~mask__) | bits__;		\
-	} while (cmpxchg(ptr, old__, new__) != old__);		\
+	} while (!try_cmpxchg(ptr, &old__, new__));		\
 								\
 	new__;							\
 })


While there you probably want something like the above... although,
looking at it now, we seem to have 'forgotten' to add try_cmpxchg to the
generic code :/
Vineet Gupta Jan. 11, 2019, 8:58 p.m. UTC | #3
On 1/11/19 1:24 AM, Peter Zijlstra wrote:
> diff --git a/include/linux/bitops.h b/include/linux/bitops.h
> index 705f7c442691..2060d26a35f5 100644
> --- a/include/linux/bitops.h
> +++ b/include/linux/bitops.h
> @@ -241,10 +241,10 @@ static __always_inline void __assign_bit(long nr, volatile unsigned long *addr,
>  	const typeof(*(ptr)) mask__ = (mask), bits__ = (bits);	\
>  	typeof(*(ptr)) old__, new__;				\
>  								\
> +	old__ = READ_ONCE(*(ptr));				\
>  	do {							\
> -		old__ = READ_ONCE(*(ptr));			\
>  		new__ = (old__ & ~mask__) | bits__;		\
> -	} while (cmpxchg(ptr, old__, new__) != old__);		\
> +	} while (!try_cmpxchg(ptr, &old__, new__));		\
>  								\
>  	new__;							\
>  })
> 
> 
> While there you probably want something like the above... 

As a separate change perhaps so that a revert (unlikely as it might be) could be
done with less pain.

> although,
> looking at it now, we seem to have 'forgotten' to add try_cmpxchg to the
> generic code :/

So it _has_ to be a separate change ;-)

But can we even provide a sane generic try_cmpxchg. The asm-generic cmpxchg relies
on local irq save etc so it is clearly only to prevent a new arch from failing to
compile. atomic*_cmpxchg() is different story since atomics have to be provided by
arch.

Anyhow what is more interesting is the try_cmpxchg API itself. So commit
a9ebf306f52c756 introduced/use of try_cmpxchg(), which indeed makes the looping
"nicer" to read and obvious code gen improvements.

So,
        for (;;) {
                new = val $op $imm;
                old = cmpxchg(ptr, val, new);
                if (old == val)
                        break;
                val = old;
        }

becomes

        do {
        } while (!try_cmpxchg(ptr, &val, val $op $imm));


But on pure LL/SC retry based arches, we still end up with generated code having 2
loops. We discussed something similar a while back: see [1]

First loop is inside inline asm to retry LL/SC and the outer one due to code
above. Explicit return of try_cmpxchg() means setting up a register with a boolean
status of cmpxchg (AFAIKR ARMv7 already does that but ARC e.g. uses a CPU flag
thus requires an additional insn or two). We could arguably remove the inline asm
loop and retry LL/SC from the outer loop, but it seems cleaner to keep the retry
where it belongs.

Also under the hood, try_cmpxchg() would end up re-reading it for the issue fixed
by commit 44fe84459faf1a.

Heck, it would all be simpler if we could express this w/o use of cmpxchg.

	try_some_op(ptr, &val, val $op $imm);

P.S. the horrible API name is for indicative purposes only

This would remove the outer loop completely, also avoid any re-reads due to the
semantics of cmpxchg etc.

[1] https://www.spinics.net/lists/kernel/msg2029217.html
Will Deacon Jan. 14, 2019, 7 p.m. UTC | #4
On Thu, Jan 10, 2019 at 04:26:27PM -0800, Vineet Gupta wrote:
> | > Also, set_mask_bits is used in fs quite a bit and we can possibly come up
> | > with a generic llsc based implementation (w/o the cmpxchg loop)
> |
> | May I also suggest changing the return value of set_mask_bits() to old.
> |
> | You can compute the new value given old, but you cannot compute the old
> | value given new, therefore old is the better return value. Also, no
> | current user seems to use the return value, so changing it is without
> | risk.
> 
> Link: http://lkml.kernel.org/g/20150807110955.GH16853@twins.programming.kicks-ass.net
> Suggested-by: Peter Zijlstra <peterz@infradead.org>
> Cc: Miklos Szeredi <mszeredi@redhat.com>
> Cc: Ingo Molnar <mingo@kernel.org>
> Cc: Jani Nikula <jani.nikula@intel.com>
> Cc: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> Cc: Will Deacon <will.deacon@arm.com>
> Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
> ---
>  include/linux/bitops.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/include/linux/bitops.h b/include/linux/bitops.h
> index 705f7c442691..602af23b98c7 100644
> --- a/include/linux/bitops.h
> +++ b/include/linux/bitops.h
> @@ -246,7 +246,7 @@ static __always_inline void __assign_bit(long nr, volatile unsigned long *addr,
>  		new__ = (old__ & ~mask__) | bits__;		\
>  	} while (cmpxchg(ptr, old__, new__) != old__);		\
>  								\
> -	new__;							\
> +	old__;							\
>  })
>  #endif

Acked-by: Will Deacon <will.deacon@arm.com>

May also explain why no in-tree users appear to use the return value!

Will

Patch
diff mbox series

diff --git a/include/linux/bitops.h b/include/linux/bitops.h
index 705f7c442691..602af23b98c7 100644
--- a/include/linux/bitops.h
+++ b/include/linux/bitops.h
@@ -246,7 +246,7 @@  static __always_inline void __assign_bit(long nr, volatile unsigned long *addr,
 		new__ = (old__ & ~mask__) | bits__;		\
 	} while (cmpxchg(ptr, old__, new__) != old__);		\
 								\
-	new__;							\
+	old__;							\
 })
 #endif