On Tue, 1 Aug 2017 19:37:31 +0200, Rob Van Der Heij ([email protected])
wrote about "Re: Question about CPUs" (in
<CAJCc=kjODGW3hgNtwpZS49AfdGgCFaz3CR0q48==h=g2ryg...@mail.gmail.com>):

>> The term you are groping for here is "memory interlock". This was coined
>> by IBM in regard to the TS instruction in that it imposes a lock on its
>> target byte to prevent any other processor in the SMP configuration from
>> manipulating that byte until its operation is complete.
> 
> I believe the thread has been ignoring the earlier reference to "block
> concurrent" that the Principles of Operation uses to explain the issue.

The term "memory interlock" is much older than the concept of block
concurrency -- by some decades. The memory management units of newer
CPUs implement integrity in more modern ways. You should bear in mind
that memory interlocks were just that: they applied directly to CSTOR,
as the older processors were not usually cached.

> To redefine the concept of interrupts does not make it more clear in my
> opinion.

I did not see any redefinition of interrupt in this thread.

The semantical shift in my message was about atomicity. I expanded it to
include concurrency as well as interruption.
-- 
Regards,

Dave  [RLU #314465]
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
[email protected] (David W Noon)
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*

 

Reply via email to