Re: Structure of large link libraries in f95
- From: Charles Russell <SPAMworFREEwor@xxxxxxxxxxxxx>
- Date: Wed, 29 Mar 2006 23:13:35 -0600
robert.corbett@xxxxxxx wrote:
Charles Russell wrote:
Is there any way in f95 to structure a large link library such as LAPACK
without either 1) having to link the whole library when calling only a
few routines or 2) having to create and maintain a large collection of
explicit interfaces?
LAPACK is structured as separately compileable routines. If you
put the routines in an archive, programs linked against the archive
will extract only the routines the program needs from the archive.
Yes, that is what I am accustomed to with f77. But f90 modules build to object files and not archives. That is exactly the problem.
I cited LAPACK as an example that might be of general interest, and for which LAPACK90 shows just how ugly option 2 can get. What I am really interested in is what to do with my big library of old f77 math routines, which I like to keep at my fingertips for model building, should I switch to f95. The compiler would then have to do the kinds of type checking that I now obtain with ftnchek.
I asked approximately this same question about five years ago, but perhaps less clearly, and perhaps there has been some progress in the interim.
.
- Follow-Ups:
- Re: Structure of large link libraries in f95
- From: Walter Spector
- Re: Structure of large link libraries in f95
- From: robert . corbett
- Re: Structure of large link libraries in f95
- References:
- Structure of large link libraries in f95
- From: Charles Russell
- Re: Structure of large link libraries in f95
- From: robert . corbett
- Structure of large link libraries in f95
- Prev by Date: Re: Enum not found in PNNL FLIB library
- Next by Date: Re: Structure of large link libraries in f95
- Previous by thread: Re: Structure of large link libraries in f95
- Next by thread: Re: Structure of large link libraries in f95
- Index(es):
Relevant Pages
|
|