🚀 go-pugleaf

Sequent Symmetry S81 running DYNIX V3.0.17.10.  Gcc-2.1, gnu-ld (binutils-1.9)

	Hi folks.  I have a problem, and I hope that someone can help.  My
    code, compiled with our Sequent's cc compiler, produces an executable
    which appears to use about 1/3 more memory than that produced by
    gcc-2.1/ld-1.9.  (At least, according to the SIZE field of ps(1).)  The
    code does a heck of a lot of mallocing, but does free all memory that it
    no longer needs.  (The host cc is not too hot anyway.  Strangely enough,
    the code produced by gcc/ld also performs better than that of the Sequent's
    own cc with respect to swapping and paging.)

	Unfortunately, I need a compiler which will link in the Sequent
    parallel libs (libpps.a), and gcc-2.1/ld-1.9 does not like the look of
    the lib:

% gcc ... -lpps
ld: relocation address out of range in /usr/lib/libpps.a(shbrk.o)

	As I will be shortly producing forked processes, each mallocing
    chunks of memory, the memory consumption difference between
    gcc-2.1/ld-1.9 and the Sequent cc becomes even more important.

	Is there, anywhere, any version of gcc/ld which supports Sequent
    parallel libs, or has anyone hacked a version of gcc/ld so that it will?

	Please please email/post (with good news!).  Many thanks in advance,
    Simon.
_______________________________________________________________________________
Simon Marshall, Dept. of Computer Science, University of Hull, Hull HU6 7RX, UK
  "Football isn't about life and death.  It's more important than that." Bill
Email: S.Marshall@Hull.ac.uk    Phone: +44 482 465181    Fax: 466666   Shankley