Why these original FORTRAN quirks?
in [Fortran]
|
From: Steve O'Hara-Smith on 1 Nov 2006 09:26 On Wed, 1 Nov 2006 14:17:41 +0000 (UTC) pa(a)see.signature.invalid (Pierre Asselin) wrote: > In comp.lang.fortran Steve O'Hara-Smith <steveo(a)eircom.net> wrote: > > > That's true, I was thinking of load time relocation and linking > > loaders which AFAIK are completely gone. It's been a while since I last > > looked at the detailed behaviour of a linker but they seem to talk about > > symbol resolution these days instead of relocation. > > Probably because of the move to shared libraries. You can't do relocation > if different processes map the same code to different addresses. Yes shared libraries use PIC. > AFAIK relocation is still used when linking statically. This is where the descriptions seem to be in terms of symbol resolution instead of relocation these days. -- C:>WIN | Directable Mirror Arrays The computer obeys and wins. | A better way to focus the sun You lose and Bill collects. | licences available see | http://www.sohara.org/
From: jmfbahciv on 2 Nov 2006 08:48 In article <eia5kh$nvr$1(a)newslocal.mitre.org>, Joe Morris <jcmorris(a)mitre.org> wrote: >jmfbahciv(a)aol.com writes: >> Joe Morris <jcmorris(a)mitre.org> wrote: >>>Charles Richmond <richchas(a)comcast.net> writes: > >>>>I knew several engineering graduate students in the late 70's, >>>>and understand that programming work for their theses *had* to >>>>be done in FORTRAN. The thesis would be rejected if the software >>>>was done in another language. > >>>Ouch. What was the justification for that policy? And at what university? > >>Remember that Charles is an auld fart. My best guesses are: >>1. That's the only lanugage the advisors knew or >>2. Interdeparmental politics (a.k.a budgeting games) required >>that FORTRAN to be used for past and future expenditures. > >Maybe I was just lucky...but my MS thesis (a program simiulating >digital circuitry, including a compiler to define the circuit) >was presented to a committee that was quite up-front at being >unqualified to analyze the underlying code. What interested them >was whether the code actually did what it claimed to do. > >One thing that I included in both the thesis and my oral defense of >it was extensive test scripts and their output...starting with >trivial primatives ("does this inverter actually emit NOT(input)?") >and building up to a complex mesh, showing the propagation of pulses >through the various elements. KEWL. One of course grad courses offered when I went to school was to write a compiler. There was a guy who worked on his for at least four years' worth of that class. I don't remember if he ever got it to work. > >I'll admit that I was fortunate to be working for the computer center >at that time, so I didn't have to worry about paying for the >computer time. <grin> Of course! Why do you think we all started working for the people who made them. Free toys and 7x24 playtime. And, instead of getting kicked off, you're getting begged to use it. /BAH
From: jmfbahciv on 2 Nov 2006 08:50 In article <eiaaa5$b7p$1(a)reader2.panix.com>, pa(a)see.signature.invalid (Pierre Asselin) wrote: >In comp.lang.fortran Steve O'Hara-Smith <steveo(a)eircom.net> wrote: > >> That's true, I was thinking of load time relocation and linking >> loaders which AFAIK are completely gone. It's been a while since I last >> looked at the detailed behaviour of a linker but they seem to talk about >> symbol resolution these days instead of relocation. > >Probably because of the move to shared libraries. You can't do relocation >if different processes map the same code to different addresses. Of course you can do this. As TW would remark, "All it takes is a small matter of programming." > >AFAIK relocation is still used when linking statically. You can do it at runtime or at load time. There are tradeoffs between choosing which one to use. /BAH
From: Pierre Asselin on 2 Nov 2006 09:49 In comp.lang.fortran Steve O'Hara-Smith <steveo(a)eircom.net> wrote: > On Wed, 1 Nov 2006 14:17:41 +0000 (UTC) > pa(a)see.signature.invalid (Pierre Asselin) wrote: > > AFAIK relocation is still used when linking statically. > This is where the descriptions seem to be in terms of symbol > resolution instead of relocation these days. I dunno, on Linux the man pages for ld(1) and elf(5) still use the word "relocation" in lots of places. It seems to mean what I remember it to mean, but I don't have time to check in detail. -- pa at panix dot com
From: Anne & Lynn Wheeler on 2 Nov 2006 11:33
jmfbahciv(a)aol.com writes: > Of course you can do this. As TW would remark, "All it takes > is a small matter of programming." or lots of SMOP ... misc. posts mentioning SMOP for doing relocation http://www.garlic.com/~lynn/subtopic.html#adcon |