profiling not working, mon.out file is not created

[migurus]

My setup is OSR 5.0.7 MP5 and SCO Development 5.2.0A
I am trying to run a profiling against my program, so I recompiled my
code and all archives used in my build with -p. When I run the program
the profile file is not being generated for some reason.The source
code consists of some 30 C files, the three archives used have around
60 object files in them. I made sure program ends with a legit exit
(0);

When I try to profile a simple couple source files program everything
is working. So, does size matter, as they say, or there could be some
other issues I need to be concerned with? Any pointers would be
appreciated.

[migurus]

Post by migurus
My setup is OSR 5.0.7 MP5 and SCO Development 5.2.0A
I am trying to run a profiling against my program, so I recompiled my
code and all archives used in my build with -p. When I run the program
the profile file is not being generated for some reason.The source
code consists of some 30 C files, the three archives used have around
60 object files in them. I made sure program ends with a legit exit
(0);
When I try to profile a simple couple source files program everything
is working. So, does size matter, as they say, or there could be some
other issues I need to be concerned with? Any pointers would be
appreciated.

I found the reason for mon.out not being generated.
My link step somehow omits the -p that I have specified in CFLAGS,
that is either my ignorance, or a compiler 'feature'

Here is my example makefile:
$ cat mkf
CFLAGS = -p
OBJS = a.o b.o

a: $(OBJS)
$(CC) -o $@ $(OBJS)

$ rm [ab].o; make -f mkf
cc -p -c a.c
cc -p -c b.c
cc -o a a.o b.o <<< here is where -p is needed, but missing
So, now a make question:
How do I ensure -p is passed to the last step?

[Brian K. White]

----- Original Message -----
From: "migurus" <***@yahoo.com>
Newsgroups: comp.unix.sco.programmer
To: <***@jpr.com>
Sent: Friday, December 19, 2008 6:56 PM
Subject: Re: profiling not working, mon.out file is not created

Post by migurus
My setup is OSR 5.0.7 MP5 and SCO Development 5.2.0A
I am trying to run a profiling against my program, so I recompiled my
code and all archives used in my build with -p. When I run the program
the profile file is not being generated for some reason.The source
code consists of some 30 C files, the three archives used have around
60 object files in them. I made sure program ends with a legit exit
(0);
When I try to profile a simple couple source files program everything
is working. So, does size matter, as they say, or there could be some
other issues I need to be concerned with? Any pointers would be
appreciated.

I found the reason for mon.out not being generated.
My link step somehow omits the -p that I have specified in CFLAGS,
that is either my ignorance, or a compiler 'feature'

Here is my example makefile:
$ cat mkf
CFLAGS = -p
OBJS = a.o b.o

a: $(OBJS)
$(CC) -o $@ $(OBJS)

$ rm [ab].o; make -f mkf
cc -p -c a.c
cc -p -c b.c
cc -o a a.o b.o <<< here is where -p is needed, but missing
So, now a make question:
How do I ensure -p is passed to the last step?

---------


a: $(OBJS)
$(CC) $(CFLAGS) -o $@ $(OBJS)
--
Brian K. White ***@aljex.com http://www.myspace.com/KEYofR
+++++[>+++[>+++++>+++++++<<-]<-]>>+.>.+++++.+++++++.-.[>+<---]>++.
filePro BBx Linux SCO FreeBSD #callahans Satriani Filk!