X-Authentication-Warning:	acp3bf.physik.rwth-aachen.de: broeker owned process doing -bs
Date:	Fri, 25 May 2001 12:53:42 +0200 (MET DST)
From:	Hans-Bernhard Broeker <broeker AT physik DOT rwth-aachen DOT de>
X-Sender:	broeker AT acp3bf
To:	djgpp-workers AT delorie DOT com
Subject:	Re: More control for profiling
In-Reply-To:	<7704-Wed23May2001220209+0300-eliz@is.elta.co.il>
Message-ID:	<Pine.LNX.4.10.10105251231190.21714-100000@acp3bf>
MIME-Version:	1.0
Reply-To:	djgpp-workers AT delorie DOT com
Errors-To:	nobody AT delorie DOT com
X-Mailing-List:	djgpp-workers AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

On Wed, 23 May 2001, Eli Zaretskii wrote:

> The following changes implement a couple of new functions (by mostly
> reusing existing code ;-) which let you control profiling
> programmatically from within the program being profiled.  

Sounds like a good idea. Gprof can do a large part of this, if you really
know how to use it (static callgraph deduction, limiting the displayed
data to a specific part of the call-graph, etc.), but it can be quite 
cumbersome, to put it mildly.

> The interface is modeled on functions which exist on other platforms
> (GNU/Linux and FreeBSD at the very least).  Could someone please tell
> what header should I put the function prototypes, though?  

I checked on Digital Unix: they have monitor() and friends <stdio.h>. 
But I wouldn't say that's a particularly good place to put them in.

SysV Unix (i.e. SCO, now Caldera) seems to have put monitor() into a
headerfile <mon.h>. I don't know what else is in there.

> One GNU/Linux box has monstartup in sys/gmon.h, but I wonder if this
> is standard, and if it makes sense to add a brand new header just for
> this?

I would say it does. One could argue that these function should/could
go into <sys/gprof.h>, but I don't like that idea any better than a 
separate headerfile. 



-- 
Hans-Bernhard Broeker (broeker AT physik DOT rwth-aachen DOT de)
Even if all the snow were burnt, ashes would remain.

- Raw text -