Re: [Xen-devel] [PATCH 4/4] xen/x86: Correct mandatory and SMP barrier definitions

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

On 05/12/16 11:51, Jan Beulich wrote:
>>>> On 05.12.16 at 11:05, <andrew.cooper3@xxxxxxxxxx> wrote:
>> Barriers are a complicated topic, a common source of confusion in submitted
>> code, and their incorrect use is a common cause of bugs.  It *really* doesn't
>> help when Xen's API is the same as Linux, but its ABI different.
>>
>> Bring the two back in line, so programmers stand a chance of actually getting
>> their use correct.
>>
>> As Xen has no current need for mandatory barriers, leave them commented out 
>> to
>> avoid accidential misue.
> I'm not convinced of this part - common driver code could very well
> need to use such barriers, and I would find it rather odd if I had to
> first re-introduce them when adding any such. In fact I'm surprised
> there's no such use anywhere. (Depending on the discussion on
> earlier patches, there may in fact remain a few such uses.)

I'm not surprised that there is no common use the mandatory barriers. 
Xen has no interesting device drivers using mappings other than WB or UC.

As for hiding, I am betting that it is far more likely that an
introduction is (mis)use of a barrier, rather than a genuine real use.

~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel