matt@3am-software.com said:
> > A simpler way, is to have trapsignal assert that si_code > 0, 
> That won't work.  For arithmetic faults, both vax and powerpc use
> a ksi_code of 0

This would be inconsistent with the documentation (siginfo(5)):
     If si_code is less than or equal to zero, the signal was generated by a
     user process or a user requested service:
     SI_USER
     [...]
     SI_TIMER
     [...]
     SI_ASYNCIO

My proposed patch was motivated by exactly this sentence.

best regards
Matthias