RE: my action item for explaining semantics of notify callbacks in SCE-MI pipes

Hi Per,

>-----Original Message-----
>From: owner-itc@eda.org [mailto:owner-itc@eda.org] On Behalf Of Per
Bojsen
>Sent: Friday, April 20, 2007 6:16 PM
>To: itc@eda.org
>Subject: Re: my action item for explaining semantics of notify
callbacks in
>SCE-MI pipes
>
>Hi Shabtay,
>
>> I assume that when in_pipe1 HDL-side consumer yielded to the C-side
>> producer on in_pipe1, also C-side ok_to_send callback will be called
on
>> in_pipe2. Furthermore, the order in which ok_to_send on both pipes
will
>> be called will be totally arbitrary.
>
>The ok_to_send() callback for in_pipe2 should only occur if in_pipe2
>was (partially or fully) drained.  
[Shabtay] Correct. That's what I meant.

I'm just making sure you're not
>saying that the activity on in_pipe1 is causing the ok_to_send()
>callback on in_pipe2.
>
>> I also assume that in the general case, if you have n input pipes and
m
>> output pipes, each time that the HDL side yields to the SW side,
>> ok_to_send/receive callbacks will be called for all pipes for which
the
>> #elements in their pipes have changed prior to when the yield
occurred.
>
>Yes, this is correct.  And as you said, the ordering of the callback
>invocations is undefined.
[Shabtay] Ok. Thanks.
>
>Per
>
>
>
>--
>This message has been scanned for viruses and
>dangerous content by MailScanner, and is
>believed to be clean.


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.