[parsec-users] Precompiled blackscholes not showing time spent within ROI
cbienia at CS.Princeton.EDU
Thu Jul 2 12:29:13 EDT 2009
The precompiled binaries that are intended for simulations (the ones with a
magic instruction) have time measurement disabled. You are better off to use
the time measurement functionality of your simulator.
Concerning your post from yesterday, "gtar" is a common name of the GNU tar
utility on Solaris. The normal "tar" binary that comes with Solaris is
another version. Solaris tar has caused problems for me in the past, so I'm
usually using GNU tar.
From: parsec-users-bounces at lists.cs.princeton.edu
[mailto:parsec-users-bounces at lists.cs.princeton.edu] On Behalf Of Wael Kdouh
Sent: Thursday, July 02, 2009 12:21 PM
To: PARSEC Users
Subject: Re: [parsec-users] Precompiled blackscholes not showing time spent
I tried doing that, but no luck. As I mentioned before it did not have this
problem with the version that I compiled.
On Thu, Jul 2, 2009 at 11:04 AM, Kshitij Sudan <kshitij at cs.utah.edu> wrote:
You might want to run the Simics simulation for a few thousand steps
*after* it breaks off at the end of ROI. Sometimes, the print
statements are printed out on the simics terminal after breaking off
2009/7/2 Wael Kdouh <waelkdouh at gmail.com>:
> Chris another note about the precompiled version. When I
> try to run blackscholes it breaks before and after the ROI as expected
> magic break
> points, but it does not display the time spent inside the ROI. It used to
do it with the version that I compiled.
> The reason why I am insisting on using the precompiled version versus the
> one I compiled is running time. I noticed that the precompiled one is
> running faster. I am using sparc-solaris.gcc-hooks.pre-simics. Thanks.
> P.S I have enabled the magic break points and fixed the path to the
> magic-instruction file.
> Sincerely yours,
> Wael Kdouh
> parsec-users mailing list
> parsec-users at lists.cs.princeton.edu
parsec-users mailing list
parsec-users at lists.cs.princeton.edu
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the parsec-users