Re: [PATCH] tools/memory-model: update: remove rb-dep, smp_read_barrier_depends, and lockless_dereference
From: Akira Yokosawa
Date: Sat Feb 24 2018 - 17:47:33 EST
On 2018/02/24 10:08:14 -0800, Paul E. McKenney wrote:
> On Sat, Feb 24, 2018 at 11:49:20AM -0500, Alan Stern wrote:
>> On Sat, 24 Feb 2018, Andrea Parri wrote:
>>
>>> On Fri, Feb 23, 2018 at 07:30:13PM -0800, Paul E. McKenney wrote:
>>>> On Sat, Feb 24, 2018 at 12:22:24PM +0900, Akira Yokosawa wrote:
>>>>> On 2018/02/22 07:29:02 +0900, Akira Yokosawa wrote:
>>>>>> On 2018/02/22 2:15, Alan Stern wrote:
>>>
>>> [...]
>>>
>>>>>>>
>>>>>>> Akira pointed out some typos in the original patch, and he noted that
>>>>>>> cheatsheet.txt should be updated to indicate how unsuccessful RMW
>>>>>>> operations relate to address dependencies.
>>>>>>
>>>>>> My point was to separate unannotated loads from READ_ONCE(), if the
>>>>>> cheatsheet should concern such accesses as well.
>>>>>> Unsuccessful RMW operations were brought up by Andrea.
>>>>>>
>>>>>
>>>>> Paul, can you amend above paragraph in the change log to something like:
>>>>>
>>>>> Akira pointed out some typos in the original patch, and he noted that
>>>>> cheatsheet.txt should be updated to indicate READ_ONCE() implies
>>>>> address dependency, which invited Andrea's observation that it should
>>>>> also be updated to indicate how unsuccessful RMW operations relate to
>>>>> address dependencies.
>>>>>
>>>>> , if Alan and Andrea are OK with the amendment.
>>>>>
>>>>> Also, please append my Acked-by.
>>>>>
>>>>> Acked-by: Akira Yokosawa <akiyks@xxxxxxxxx>
>>>>
>>>> I can still amend this, and have added your Acked-by. If Alan and Andrea
>>>> OK with your change, I will apply that also.
>>>
>>> LGTM. Thanks,
>>
>> Me too.
>
> Very good, how about this for the new version?
>
> Thanx, Paul
>
> ------------------------------------------------------------------------
>
> commit 21ede43970e50b7397420f17ed08bb02c187e2eb
> Author: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx>
> Date: Wed Feb 21 12:15:56 2018 -0500
>
> tools/memory-model: Update: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference
>
> Commit bf28ae562744 ("tools/memory-model: Remove rb-dep,
> smp_read_barrier_depends, and lockless_dereference") was accidentally
> merged too early, while it was still in RFC form. This patch adds in
> the missing pieces.
>
> Akira pointed out some typos in the original patch, and he noted that
> cheatsheet.txt should indicate that READ_ONCE() now implies an address
> dependency. Andrea suggested documenting the relationship betwwen
> unsuccessful RMW operations and address dependencies.
Looks good. But I've found a remaining typo in the patch. See below.
> > Andrea pointed out that the macro for rcu_dereference() in linux.def
> should now use the "once" annotation instead of "deref". He also
> suggested that the comments should mention commit 5a8897cc7631
> ("locking/atomics/alpha: Add smp_read_barrier_depends() to
> _release()/_relaxed() atomics") as an important precursor, and he
> contributed commit cb13b424e986 ("locking/xchg/alpha: Add
> unconditional memory barrier to cmpxchg()"), another prerequisite.
>
> Signed-off-by: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx>
> Suggested-by: Akira Yokosawa <akiyks@xxxxxxxxx>
> Suggested-by: Andrea Parri <parri.andrea@xxxxxxxxx>
> Fixes: bf28ae562744 ("tools/memory-model: Remove rb-dep, smp_read_barrier_depends, and lockless_dereference")
> Acked-by: Andrea Parri <parri.andrea@xxxxxxxxx>
> Acked-by: Akira Yokosawa <akiyks@xxxxxxxxx>
> Signed-off-by: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>
>
> diff --git a/tools/memory-model/Documentation/cheatsheet.txt b/tools/memory-model/Documentation/cheatsheet.txt
> index 04e458acd6d4..956b1ae4aafb 100644
> --- a/tools/memory-model/Documentation/cheatsheet.txt
> +++ b/tools/memory-model/Documentation/cheatsheet.txt
> @@ -1,11 +1,11 @@
> Prior Operation Subsequent Operation
> --------------- ---------------------------
> C Self R W RWM Self R W DR DW RMW SV
> - __ ---- - - --- ---- - - -- -- --- --
> + -- ---- - - --- ---- - - -- -- --- --
>
> Store, e.g., WRITE_ONCE() Y Y
> -Load, e.g., READ_ONCE() Y Y Y
> -Unsuccessful RMW operation Y Y Y
> +Load, e.g., READ_ONCE() Y Y Y Y
> +Unsuccessful RMW operation Y Y Y Y
> rcu_dereference() Y Y Y Y
> Successful *_acquire() R Y Y Y Y Y Y
> Successful *_release() C Y Y Y W Y
> diff --git a/tools/memory-model/Documentation/explanation.txt b/tools/memory-model/Documentation/explanation.txt
> index dae8b8cb2ad3..889fabef7d83 100644
> --- a/tools/memory-model/Documentation/explanation.txt
> +++ b/tools/memory-model/Documentation/explanation.txt
> @@ -826,7 +826,7 @@ A-cumulative; they only affect the propagation of stores that are
> executed on C before the fence (i.e., those which precede the fence in
> program order).
>
> -read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have
> +read_read_lock(), rcu_read_unlock(), and synchronize_rcu() fences have
rcu_read_lock()
Don't ask why I missed this in the first place...
Paul, can you fix this directly?
Thanks, Akira
> other properties which we discuss later.
>
>
> @@ -1138,7 +1138,7 @@ final effect is that even though the two loads really are executed in
> program order, it appears that they aren't.
>
> This could not have happened if the local cache had processed the
> -incoming stores in FIFO order. In constrast, other architectures
> +incoming stores in FIFO order. By contrast, other architectures
> maintain at least the appearance of FIFO order.
>
> In practice, this difficulty is solved by inserting a special fence
> diff --git a/tools/memory-model/linux-kernel.def b/tools/memory-model/linux-kernel.def
> index 5dfb9c7f3462..397e4e67e8c8 100644
> --- a/tools/memory-model/linux-kernel.def
> +++ b/tools/memory-model/linux-kernel.def
> @@ -13,7 +13,7 @@ WRITE_ONCE(X,V) { __store{once}(X,V); }
> smp_store_release(X,V) { __store{release}(*X,V); }
> smp_load_acquire(X) __load{acquire}(*X)
> rcu_assign_pointer(X,V) { __store{release}(X,V); }
> -rcu_dereference(X) __load{deref}(X)
> +rcu_dereference(X) __load{once}(X)
>
> // Fences
> smp_mb() { __fence{mb} ; }
>