BEBO@SLACVM.SLAC.STANFORD.EDU (Bebo White) (12/01/90)
I am interested in tools for "profiling" REXX programs. By this I mean
monitor how much execution time is spent in each loop, procedure, etc.
(Similiar to FORTRAN hot spot analysis).
Thanks in advance for any leads.
Bebo White
---> the opinions expressed are mine alone and not necessarily those
of SLAC, Stanford University or the DOEmwm@raven.relay.pa.dec.com (Mike (My Watch Has Windows) Meyer) (12/01/90)
In article <90334.131137BEBO@SLACVM.SLAC.STANFORD.EDU> BEBO@SLACVM.SLAC.STANFORD.EDU (Bebo White) writes:
I am interested in tools for "profiling" REXX programs. By this I mean
monitor how much execution time is spent in each loop, procedure, etc.
(Similiar to FORTRAN hot spot analysis).
I did a statement count profiler as an exercise. It's got some (known)
problems, but converting it to a time profiler (with problems intact)
would be straightforward. Fixing the problems would be more complicated.
It's written in ARexx, and may require some work to convert to another
dialect. Ask for a copy if you want it.
<mike
--KXH105@psuvm.psu.edu (Ken Hornstein) (12/01/90)
In article <90334.131137BEBO@SLACVM.SLAC.STANFORD.EDU>, BEBO@SLACVM.SLAC.STANFORD.EDU (Bebo White) says: > >I am interested in tools for "profiling" REXX programs. By this I mean >monitor how much execution time is spent in each loop, procedure, etc. >(Similiar to FORTRAN hot spot analysis). Get the HISTOREX PACKAGE from a VM-UTIL LISTSERV site (One is at UBVM, I believe). It's the closest thing I know of. It tells what percentage of execution time is spent on each line. It's not done by loop, procedure, etc .. but it's better than nothing. --- Suitably interpreted, this is an exact value. Ken Hornstein kxh105@psuvm.psu.edu Phone: 814/862-7007