The problem with
2.0**-3*5
is whether it should be interpreted as "2.0**(-3) * 5" or as "2.0**(-3*5)".
gfortran does the former, ifort the latter.
g95, openf95, sunf95 and (of cause) NAG f95 reject it always.
ifort: Accepts it, unless -stand f* is used.
gfortran: Warning for -pedantic, Error for -std=f*. Nothing for -Wall.
I would suggest to print at least a warning by default; or to shift it to
legacy, but having it without diagnostics by default can lead to wrong results.
--
Summary: Make support of 2.0**-3*5 tighter: Print warning by
default
Product: gcc
Version: 4.3.0
Status: UNCONFIRMED
Keywords: diagnostic
Severity: normal
Priority: P3
Component: fortran
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: burnus at gcc dot gnu dot org
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34536
