[m-users.] Grade for profiling

[Peter Wang]

On Tue, 21 Mar 2023 18:43:36 +1100 Julien Fischer <jfischer at opturion.com> wrote:
> 
> Hi Peter,
> 
> On Tue, 21 Mar 2023, Peter Wang wrote:
> 
> > On Tue, 21 Mar 2023 08:08:19 +0100 Volker Wysk <post at volker-wysk.de> wrote:
> >> For the record (if someone stumbles into the same problem). With the help of
> >> Julien I was able to find the solution.
> >> 
> >> A (Linux) kernel feature called "Address space layout randomization" (ASLR)
> >> must be turned off, before running the program which is to be profiled. This
> >> can be accomplished with this command:
> >> 
> >> sudo bash -c "echo 0 > /proc/sys/kernel/randomize_va_space"
> >> 
> >> Then you get the right names for most predicates.
> >
> > Did you merge profiling data from different runs?
> > Otherwise, if the Prof.* files are generated in the same run
> > then I don't see why ASLR should make a difference.
> 
> Because there's more going on that just that. Volker's program forks
> off separate processes and I suspect the the Prof.* files end up being
> inconsistent -- ASLR just complicates matters because the addresses
> differ between each of them.

That makes sense. The Prof.Decl file would be written when the program
starts up, and the Prof.Counts file is probably overwritten by each
process when it shuts down. Obviously, you may not get usable profiling
information then.

Peter