Date: Mon, 14 Nov 1994 12:28:21 -0500 (CDT)
From: Aaron Ucko <UCKO AT VAX1 DOT ROCKHURST DOT EDU>
Subject: Re: Shared libraries
To: dj AT stealth DOT ctron DOT com
Cc: djgpp AT sun DOT soe DOT clarkson DOT edu
Organization: Rockhurst College; Kansas City, MO

>> Ouch.  Hmmm.  How about some sort of (don't tell the FSF :-) ) #pragma?  How
>> do other systems handle this sort of thing?
>
>The linker just knows to put that data somwhere that strip won't
>touch.
That should be easy enough to fix.
>
>> Why not build something like that into DJGPP 2.0?  Have the stub launch a
>> loader which exits after loading the libraries?
>
>There's still the problem of where to put the library.  Most systems
>that support shared libraries can load the library at a predictable
>address, so that no relocation is needed.  In DPMI 0.9, there is no
>way of doing this.
>
Good point...but isn't the place x86 segmentation might _finally_ come in
handy?  Just give each library its own segment and tell the linker which one
each owns.  Even under DPMI, it should at least be possible to get a segment
to point to the appropriate location, shouldn't it?  (If I sound incoherent,
it's because I don't know a whole lot about protected-mode programming. 
Sorry.)

--- Aaron Ucko (ucko AT vax1 DOT rockhurst DOT edu; finger for PGP public key) -=- httyp!
-=*=-Just because you're paranoid doesn't mean they aren't out to get you.-=*=-
Geek code 2.1 [finger hayden AT vax1 DOT mankato DOT msus DOT edu for explanation]: 
 GCS/M/S d(-) H s g+ p? !au a-- w+ v+ C++(+++)>++++ U-(S+)>++++ P+ L>++ 3(-) 
 E-(----) !N>++ K- W(--) M-(--) V(--) po-(--) Y+(++) t(+) !5 j R G tv--(-) 
 b+++ !D(--) B--(---) e>++++(*) u++(@) h!() f(+) r-(--)>+++ n+(-) y?