<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=US-ASCII">
<META NAME="Generator" CONTENT="MS Exchange Server version 5.5.2654.45">
<TITLE>RE: [PATCH] cl_timer</TITLE>
</HEAD>
<BODY>
<BR>
<P><FONT SIZE=2>> -----Original Message-----</FONT>
<BR><FONT SIZE=2>> From: Fab Tillier [<A HREF="mailto:ftillier@silverstorm.com">mailto:ftillier@silverstorm.com</A>] </FONT>
<BR><FONT SIZE=2>> Sent: Thursday, September 15, 2005 7:10 PM</FONT>
<BR><FONT SIZE=2>> To: 'Yossi Leybovich'</FONT>
<BR><FONT SIZE=2>> Cc: openib-windows@openib.org</FONT>
<BR><FONT SIZE=2>> Subject: RE: [PATCH] cl_timer</FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>> > From: Yossi Leybovich [<A HREF="mailto:sleybo@mellanox.co.il">mailto:sleybo@mellanox.co.il</A>]</FONT>
<BR><FONT SIZE=2>> > Sent: Thursday, September 15, 2005 3:14 AM</FONT>
<BR><FONT SIZE=2>> > </FONT>
<BR><FONT SIZE=2>> > Fab</FONT>
<BR><FONT SIZE=2>> > Attached patch for supporting nsec in timer stamp and </FONT>
<BR><FONT SIZE=2>> provide way to </FONT>
<BR><FONT SIZE=2>> > get the CPU freq We use this in our performance tests (I </FONT>
<BR><FONT SIZE=2>> hope that we </FONT>
<BR><FONT SIZE=2>> > upload them, to the OpenIB in the near future)</FONT>
<BR><FONT SIZE=2>> > (I left the cl_get_time_stamp_sec as is so we will not </FONT>
<BR><FONT SIZE=2>> require to change old</FONT>
<BR><FONT SIZE=2>> > code)</FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>> Why add the cl_get_time_stamp_usec function when </FONT>
<BR><FONT SIZE=2>> cl_get_time_stamp is identical? Why not just use the existing </FONT>
<BR><FONT SIZE=2>> function? At a minimum, why not call the existing function </FONT>
<BR><FONT SIZE=2>> rather than duplicating it?</FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>I did not want to change the interface of existing code , </FONT>
<BR><FONT SIZE=2>but you right the cl_get_time_stamp and cl_get_time_stamp_usec do the same</FONT>
<BR><FONT SIZE=2>I want new code to use cl_get_time_stamp_usec and later change existing code to use it.</FONT>
</P>
<P><FONT SIZE=2>> Also, for performance tests it would probably be better </FONT>
<BR><FONT SIZE=2>> (lower impact of</FONT>
<BR><FONT SIZE=2>> counters) to have two functions - one to return the raw tick </FONT>
<BR><FONT SIZE=2>> count, and the other to return the frequency. That way, you </FONT>
<BR><FONT SIZE=2>> only perform the conversion to human readable time units </FONT>
<BR><FONT SIZE=2>> after the test completes, rather than at runtime.</FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>> If this is OK, I can code this up - it will add two </FONT>
<BR><FONT SIZE=2>> functions, cl_get_tick_count, and cl_get_tick_freq.</FONT>
<BR><FONT SIZE=2>> </FONT>
</P>
<P><FONT SIZE=2>OK its fine with me , the moment you check it in I will change the perf_main.</FONT>
<BR><FONT SIZE=2>(BTW I think I already add function to get the freq)</FONT>
</P>
<P><FONT SIZE=2>> Let me know.</FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>> Thanks,</FONT>
<BR><FONT SIZE=2>> </FONT>
<BR><FONT SIZE=2>> - Fab </FONT>
<BR><FONT SIZE=2>> </FONT>
</P>
</BODY>
</HTML>