[Patch, Fortran] Fix error locations
Sun Dec 14 18:41:00 GMT 2008
Daniel Kraft wrote:
> Now the loop does one more iteration, but I guess this is ok from
> looking quickly at the code as because of your other changes now simply
> the value 0 is possible for c1/c2, while before it would have been
> changed to 1.
Not exactly, you are explaining why the other part is removed. :)
cmax is max(c1,c2): we finish the loop as soon as both 1 and 2 have been
printed. If we stopped at cmax-1, either 1 or 2 wouldn't be printed.
Another way to say it: If you want to offset by one character to the
right, you will have to do one more iteration as you will print one
> Ok if no regression failures.
No regression, thanks for the review
More information about the Gcc-patches