| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 206.1 | compilable, yes; executable as is, no | COOKIE::DOUCETTE | Chuck, DBS/CCAS/Vortex/Babelfish | Thu Oct 20 1988 13:41 | 6 |
| You could probably compile it; but, you couldn't run it as is since
it generates Ultrix VAX assembler (which isn't the same as MACRO on VMS)
with symbol information for dbx (the Ultrix debugger) and not VMS object code
with symbol information for VAX DEBUG.
Chuck
|
| 206.2 | Will it work for a non-Vax? | JEREMY::DAN | Dan Frommer | Thu Oct 20 1988 14:09 | 10 |
| Let me be more specific:
I'm interested in generating code for another architecture, for
which I have a cross assembler under VMS.
Is it possible to take
PCC, make the necessary modifications to its source, compile it
using VAX C, and then use the output image as a cross-compiler?
Dan
|
| 206.3 | Try the DECUS C compiler | 3D::VESPER | A bug is a feature with no known use | Mon Oct 24 1988 10:05 | 17 |
| I did a similar exersize a few years ago, changing the DECUS C compiler
code generator to output 68000 assembly source. This may be a better
choice than pcc because you don't have to worry about license issues
and it already runs under VMS. I used a switch to get the compiler
to dump out an intermediate form, read that file and generated the
assembler source as a separate pass. I based my program on the code
generation in DECUS C. Of course, since it was for a
college course, I didn't get everything done to my satisfaction.
In particular, I never found out how the DECUS C compiler re-used
registers in a complex expression and so it was easy to get my compiler
to fail.
The notes file MAY14::DECUSC will tell you how to get the sources.
Martin Minow, the principal author, reads that file regularly and
is willing to answer questions.
Andy V
|