[Bug target/11251] [3.3/3.4 Regression] g77 complex number program produces wrong code when optimising

[toon at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11251


toon at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |WAITING


------- Additional Comments From toon at gcc dot gnu dot org  2003-06-23 19:50 -------
The Fortran code is illegal.

You cannot call cross with the first two actual arguments being equal, because
the first argument is being written into.

I cannot duplicate this behaviour on powerpc-unknown-linux-gnu, but
perhaps one of you can test whether

g77 -O -fargument-alias <file.f>

fixes the problem.  In that case this report can be closed with INVALID.

Thanks, Toon.