[Rivet] Dangerous casting to FinalState

[Chris Pollard]

Hi Frank,

I tried out your test analysis and hepmc file last night after making a
local change to remainingFinalState(), and I found the same issues as you
have. Unfortunately.

"Error in MY_RFSTEST2c::init method: No projections registered for
parent 0x7ffe07be8840"

This only seems to crash when "remainder" is a VetoedFinalState. When it is
declared a "const VetoedFinalState&", "FinalState", or "const FinalState&",
the bookkeeping works (although the physics fails in analyze() for
FinalState).

I am booked for the next few hours but will hopefully be able to take
another look in the afternoon. Curse c++.

Chris

On Thu, Jul 9, 2015 at 8:49 AM, Frank Siegert <frank.siegert at cern.ch> wrote:

> >> We both agree that the best thing to do is to have remainingFinalState()
> >> return "const VetoedFinalState&", which is the actual type of the
> underlying
> >> returned object. Currently it returns "const FinalState&".
> >
> > I agree, and this is part of the patch I posted in this thread.
> >
> >> This change
> >> *should* generate compile-time warnings when a user tries something like
> >>
> >> FinalState fs = zfinder.remainingFinalState();
> >>
> >> because fs is being copy-constructed from a more complex class. @Frank,
> I
> >> think this will also make your attached broken example work as expected.
> >
> > Is this a compile-time warning or error, and are you sure it will
> > really complain? I can test it myself later, but maybe you have
> > already tested it.
>
> I just tested compiling my example containing a
> FinalState remainder = zfinder.remainingFinalState();
> with the current tip, to which David has pushed those changes already.
> There is no warning, let alone a fatal error, when compiling with
> this.
>
> So, do we want to go ahead with my proposed changes to protect the
> user from the slicing mistake? Consider that even experienced Riveters
> like Holger and Christian have tripped over this!
>
> In that case we would need to understand why
>   VetoedFinalState remainder = zfinder.remainingFinalState();
>   addProjection(remainder, "RFS");
> (i.e. making a full copy, not just by reference)
> makes Rivet's projection book-keeping fall over:
>
>   Error in MY_RFSTEST2c::init method: No projections registered for
> parent 0x7ffe07be8840
>
> I am again attaching the patch (updated to remove the changes David
> has already committed, needs tip), and the problematic use case
> (MY_RFSTEST2c.cc) and a HepMC file to test this with:
>
> $ cd rivet; patch -p1 < hidden-copy.patch; make install
> $ rivet-buildplugin MY_RFSTEST2c.cc
> $ rivet --pwd -a MY_RFSTEST2c file.hepmc
>
> Cheers,
> Frank
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.hepforge.org/lists-archive/rivet/attachments/20150709/8b977628/attachment.html>