Thanks for the MacOSX x86 build. Unfortunately there is a
regression in my full application that I haven't been able to
isolate into a test case. I get a nonsensical error message:
~/gfc/geloe$ gfortran -c readlens.f90
readlens.f90:407.103:
2','MODE SEQ','NAME '//trim(titles),'UNIT '//UpperCase(UNAME(kunit
(0))(:2))
1
Error: Variable 'str' is used at (1) before the ENTRY statement in
which it is a parameter
The is no variable 'str' in the statement pointed to. My
application continues to compile and execute fine with g95, ivf,
lf95 and xlf (including with full bounds and runtime
uninitialization checking).
Every time there is a new build of gfortran, I end up fighting at
least one thing that used to work but has been broken since the
last build.