forums.silverfrost.com

Paul, I think that it is more an FTN95 issue. I ran the test program with NAG and Lahey compilers with their checking flags turned on, and there were no subscript errors reported. The Lapack code has ...

I had occasion to compile a program that used Lapack/BLAS routines, and needed to compile the needed BLAS routines with /check. FTN95 detects a subscript error in a situation where (I think) that it s ...

Thanks for the additional findings, John C. Did you also see "TL): Too many closing brackets" when you looked at the variable TL?

In both the reproducers, if the subroutine is converted t ...

It turns out that replacing COMPLEX by REAL in the reproducer causes the SDBG64 bug to remain. The larger code from which I derived the reproducer contains lots of COMPLEX variables and I originally t ...

Please compile and link the test program below with /64 /check or /64 /checkmate.

Open the EXE in SDBG64, and place a breakpoint on the indicated line. Run until the breakpoint is hit.

Hover th ...

Thanks, Robert!

The new 8.64 version of SDBG64 works correctly not only with the two reproducers but also with the unabridged program, which spans 142 source files + 22 include files.

These days, it is more common to use dynamically allocated arrays -- typically, arrays whose sizes are computed from earlier portions of input data. There is no simple, reliable way of keeping a tally ...

I remembered this discussion regarding /save and /zeroise for 64-bit compilations: http://forums.silverfrost.com/viewtopic.php?t=4153 . On 20 Feb 2020 (02/20/2020 in US format!), Paul wrote:

/ZEROI ...

Here is a very short reproducer. I think that the presence of BLOCK DATA is causing a malfunction in SDBG64.

File main.f:
program mainprg
common /xx/x(5)
call sub(x ...

The following zip file contains three source files and three include files, all of which together amount to less than a hundred lines.

https://www.dropbox.com/s/5a849rsjvn0uwty/smpdbg.7z?dl=0 ...

For instructional purposes, that code for Qsort is fine. However, it is neither stable (in the sense of preserving the order of items with equal sort keys), nor efficient.

It is not stable since i ...

Fortran 2008 provides a feature that enables one to avoid repetitions (three in JohnC's code, one of which can be avoided) of the expression pack(mine,selection), since in a large program such repeate ...

This is interesting... With Mecej4's suggestion both alternative compilers return identical non-zero values for the remaining elements!

Rather, those remaining elements simply retain their previous ...

I think that the program is in error, and FTN95 /checkmate does not catch the error.

The error is in the assignment statement

mine = pack(mine,selection)

The expression on the right o ...

John C: EQUIVALENCE was used in old codes, and compilers continue to support it even though it is now marked "obsolescent" along with COMMON. What is not allowed by the standard is declaring ...