- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I know that I've asked this before but the problem persists.
on the use of constants like from DFWINTY
STATUS_ARRAY_BOUNDS_EXCEEDED
I get the Warning:
Case selection values must fit within a signed 32 bit integer.
Now I've had this problem before and following Jugoslav's suggestion I recompiled the f90 file (DFWINTY.f90) under CVF6.6B and it went away. However, this trick did not work this time, and the warnings persist.
I'd apprecaite any other hints on removing these warnings.
Duh...Tim
on the use of constants like from DFWINTY
STATUS_ARRAY_BOUNDS_EXCEEDED
I get the Warning:
Case selection values must fit within a signed 32 bit integer.
Now I've had this problem before and following Jugoslav's suggestion I recompiled the f90 file (DFWINTY.f90) under CVF6.6B and it went away. However, this trick did not work this time, and the warnings persist.
I'd apprecaite any other hints on removing these warnings.
Duh...Tim
Link Copied
2 Replies
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
OK, this time I wasn't lazy to try it myself. And I got the same result. Sorry for questioning your sanity :-(.
My earlier suggestion was pointed to backward .mod compatibility CVF bug. But it's something fishy going on even in 6.6B. The culprit seems to be the definition of
integer, parameter :: STATUS_INVALID_DISPOSITION = ( #C0000026)
This is actually a negative number (-1073741684) and CVF doesn't like something when it performs type conversion. It looks as if it internally treats #C0000026 as an unsigned number and warns you when it's converted to a signed value -- the only one that Fortran has.
The bit-pattern is preserved, so it seems harmless, i.e. the code works. Any comments Steve?
I'd guess it wouldn't happen if such PARAMETERs were defined using negative decimal numbers, not hex constants. But it's ugly.
Jugoslav
My earlier suggestion was pointed to backward .mod compatibility CVF bug. But it's something fishy going on even in 6.6B. The culprit seems to be the definition of
integer, parameter :: STATUS_INVALID_DISPOSITION = ( #C0000026)
This is actually a negative number (-1073741684) and CVF doesn't like something when it performs type conversion. It looks as if it internally treats #C0000026 as an unsigned number and warns you when it's converted to a signed value -- the only one that Fortran has.
The bit-pattern is preserved, so it seems harmless, i.e. the code works. Any comments Steve?
I'd guess it wouldn't happen if such PARAMETERs were defined using negative decimal numbers, not hex constants. But it's ugly.
Jugoslav
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Yes, if this was defined using a decimal number, it would be happier. I think the warning is the only problem here, but it is a nuisance...
Steve
Steve

Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page