Thanks, that might help. I've found that for the 2.3.0compiler, the table is printed via "ibmfefca", got the "brilliant" idie of looking for "grand total" in the library routines. Haven't yet figured out how to disassemble it, and even if I could, I'm not sure what I would be looking for, i.e. the instruction that skips the printing of zero, or, as negative counts come out as zero, zeror-or-negative counts, if anyone here could help with that, I'd be grateful.
Robert On Sat, 18 Jun 2022 at 18:08, Tony Harminc <[email protected]> wrote: > On Sat, 18 Jun 2022 at 08:53, Robert Prins <[email protected]> > wrote: > > > Does anyone know which library modules are responsible for the > > implementation of these two compiler options, and more precisely, for the > > reporting of statement counts, as I need to (try to) patch the latter, > > given that it treats negative counts as zero, and I now have a program > > where four of the counts have turned negative, because the statements are > > executed more than 2^31-1 times. > > According to > http://bitsavers.org/pdf/ibm/370/pli/SC33-0025-3_PLI_Optimizing_Compiler_Execution_Logic_Sep85.pdf > you want module IBMBEFL. > > In that book, look at page 141 onwards for the description of how > things work, and for the Flow data area, page 369. I'm not seeing > right away where the Count info is stored, but I'm sure you can figure > it out from there... > > Tony H. > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN > -- Robert AH Prins robert(a)prino(d)org The hitchhiking grandfather <https://prino.neocities.org/index.html> Some REXX code for use on z/OS <https://prino.neocities.org/zOS/zOS-Tools.html> ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
