PQI Coalescing timer

David Geddes dgeddes at marvell.com
Mon Dec 12 16:51:57 PST 2011


Formatted message: <a href="http://www.t10.org/cgi-bin/ac.pl?t=r&f=r1112120_f.htm">HTML-formatted message</a>
Attachment #1: <a href="http://www.t10.org/cgi-bin/ac.pl?t=r&f=r1112120_image002.gif">image002.gif</a>
Attachment #2: <a href="http://www.t10.org/cgi-bin/ac.pl?t=r&f=r1112120_image001.gif">image001.gif</a>

Re: 11-383r5
I am confused by an apparent contradiction between a comment in Figure 2
(page-11) and the text that explains the rules for starting the coalescing
timer (also page-11).
The red and blue arrows at the right end of the timeline indicate that, if
the coalescing timer has reached the maximum coalescing time, an interrupt
may be sent immediately when count transitions to 1 (count=1).	 This makes
sense and I would support this behavior, but the text does not provide for a
way for the coalescing timer to start when the count =0
If the behavior in Figure 2 is the intention, then the spec should provide an
explicit way to re-start the coalescing timer when the count=0 (OQ PI=CI). 
Otherwise, the figure needs to be amended.
I'm curious what others think.
Best Regards,
David
[cid:image001.gif at 01CCB8EE.44B658F0]
...
[cid:image002.gif at 01CCB8EE.44B658F0]



More information about the T10 mailing list