caml-list - the Caml user's mailing list
 help / color / mirror / Atom feed
* [Caml-list] ocaml vs. NetBSD/R3K
@ 2001-08-04 15:23 William Lee Irwin III
  2001-08-06  9:20 ` Xavier Leroy
  0 siblings, 1 reply; 2+ messages in thread
From: William Lee Irwin III @ 2001-08-04 15:23 UTC (permalink / raw)
  To: caml-list

I recently tried bootstrapping the native code on a DecStation running
NetBSD. I don't believe there should be too many hangups, but mips.s
appears not to suffice for this target.

The float register stuff shouldn't be too tough to produce hackish
workarounds for, but I'm not sure what .cpsetup and .cpreturn are
supposed to accomplish.

Any other expected hangups?

P.S.: Here are the error messages:

gcc -c -DSYS_netbsd -o mips.o mips.s
mips.s: Assembler messages:
mips.s:31: Error: Unknown pseudo-op:  `.cpsetup'
mips.s:66: Warning: Float register should be even, was 1
	[... similar error messages elided ...]
mips.s:152: Warning: Float register should be even, was 31
mips.s:163: Error: Unknown pseudo-op:  `.cpreturn'
	[... similar error messages elided ...]
mips.s:355: Error: Unknown pseudo-op:  `.cpsetup'
mips.s:364: Warning: .end directive names unknown symbol
mips.s:373: Error: Unknown pseudo-op:  `.cpsetup'
gmake[1]: *** [mips.o] Error 1
gmake[1]: Leaving directory `/usr/pkgsrc/lang/ocaml/work/ocaml-3.01/asmrun'
gmake: *** [runtimeopt] Error 2

-------------------
Bug reports: http://caml.inria.fr/bin/caml-bugs  FAQ: http://caml.inria.fr/FAQ/
To unsubscribe, mail caml-list-request@inria.fr  Archives: http://caml.inria.fr


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [Caml-list] ocaml vs. NetBSD/R3K
  2001-08-04 15:23 [Caml-list] ocaml vs. NetBSD/R3K William Lee Irwin III
@ 2001-08-06  9:20 ` Xavier Leroy
  0 siblings, 0 replies; 2+ messages in thread
From: Xavier Leroy @ 2001-08-06  9:20 UTC (permalink / raw)
  To: William Lee Irwin III; +Cc: caml-list

> I recently tried bootstrapping the native code on a DecStation running
> NetBSD. I don't believe there should be too many hangups, but mips.s
> appears not to suffice for this target.
> 
> The float register stuff shouldn't be too tough to produce hackish
> workarounds for, but I'm not sure what .cpsetup and .cpreturn are
> supposed to accomplish.
> 
> Any other expected hangups?

The current MIPS code generator is for the newer version of the MIPS
architecture as found on the R4000 and later processor, i.e. the one
that has 32 64-bit float registers and some 64-bit integer operations.
The R3000 in your Decstation implements an older version of the MIPS
instruction set, hence the errors you got.

Up to OCaml version 2.00, the MIPS back-end supported both the R4000
under IRIX and the R3000 under Ultrix.  The latter was removed in 2.01
because we didn't know of anyone still using an R3000 machine.
You could try to resurrect the R3000/Ultrix port as a R3000/NetBSD port,
but I'm not quite sure this is worth the effort...

- Xavier Leroy
-------------------
Bug reports: http://caml.inria.fr/bin/caml-bugs  FAQ: http://caml.inria.fr/FAQ/
To unsubscribe, mail caml-list-request@inria.fr  Archives: http://caml.inria.fr


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2001-08-06  9:20 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2001-08-04 15:23 [Caml-list] ocaml vs. NetBSD/R3K William Lee Irwin III
2001-08-06  9:20 ` Xavier Leroy

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).