Re: [dylan] [archeology] locked vs atomic

Scott McKay <[email protected]>
Newsgroups gmane.comp.lang.dylan.gwydion.devel
Message-ID <CALMyEALe-PwMvLHP2VtFydbxFXSwkaJh3-OY5u+rQCH9UMvk-A@mail.gmail.com>
Ah, so 'locked' means 'atomic'.

If we *must* have an adjective, I'd deprecate 'locked'
and replace it with 'atomic', then keep the 'atomic-xxx'
primitives. Then when you've got 5-10 minutes free,
you can just get the memory model right so that all
things like inc! and dec! DTRT. :-)

On Sun, Nov 25, 2012 at 7:41 PM, Bruce Mitchener
<[email protected]>wrote:

> Scott,
>
> "define locked variable" is for use with "conditional-update!" ... which
> also (unfortunately) includes atomic-increment!, atomic-decrement!.
>
> Ingo and I are looking at various things where atomic operations would be
> handy / necessary, hence the interest...
>
> I do agree that it would be nice if both adjectives died and there were a
> good set of basic atomic primitives that just worked.
>
>  - Bruce
>
>
>
> On Sun, Nov 25, 2012 at 9:06 PM, Scott McKay <[email protected]> wrote:
>
>> Does 'define locked variable' actually take a lock?
>>
>> Personally I dislike the idea of making it easy to take
>> locks on tiny things like variables and slots. It's expensive,
>> prone to deadlock, composes poorly, etc etc.
>>
>> It seems to me that we should provide locks (we do)
>> and have a memory model that does atomic updates
>> to "atoms". That is, lose both adjectives!
>>
>> For what it's what, that's not what I thought 16 years ago.
>>
>> --Scott
>>
>>
>> On Sun, Nov 25, 2012 at 7:18 PM, Bruce Mitchener <
>> [email protected]> wrote:
>>
>>> This is going back into some ancient history ...
>>>
>>> OpenDylan supports "define locked variable" (for use with atomic
>>> operations). However, it also supports "define atomic variable", however,
>>> nothing actually uses that adjective.
>>>
>>> Similarly, there is the start of some support for "atomic slot", but not
>>> "locked slot".
>>>
>>> Does anyone recall what the "atomic variable" and "atomic slot" were
>>> intended to be and how they compared with "locked variable"?
>>>
>>> This goes back to 1996 or so ...
>>>
>>> Thanks!
>>>
>>>  - Bruce
>>>
>>>
>>> _______________________________________________
>>> hackers mailing list
>>> [email protected]
>>> https://lists.opendylan.org/mailman/listinfo/hackers
>>>
>>>
>>
>> _______________________________________________
>> hackers mailing list
>> [email protected]
>> https://lists.opendylan.org/mailman/listinfo/hackers
>>
>>
>
> _______________________________________________
> hackers mailing list
> [email protected]
> https://lists.opendylan.org/mailman/listinfo/hackers
>
>

_______________________________________________
hackers mailing list
[email protected]
https://lists.opendylan.org/mailman/listinfo/hackers
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.