[Info-vax] VMS Software needs to port VAX DIBOL to OpenVMS X86 platform

[Stephen Hoffman]

On 2021-01-06 01:45:16 +0000, Tim Sneddon said:

> Arne Vajh??j <arne at vajhoej.dk> wrote:
> 
>> As I remember it then VAX C was the only DEC compiler where it was 
>> common to turn off optimization to get code working.
> 
> I certainly experienced that when using builtins that accessed registers.
> However, beyond that I never had any huge problems.

The MD5 source code implementation from the RFC and the VGC optimizer 
are not... friendly.

Whether that VCG optimizer "fun" was limited to VAX C, or both VAX C 
and DEC C were afflicted, I don't recall.

There have been other bugs. I bumped into a gratuitous pointer 
dereference, though that did get fixed in a later VAX C.

VAX C V3.2 was decently stable, and DEC C with VCG worked reasonably 
well within its expected limits.

But then it's also all about VAX, and VAX is a quarter-century past 
much in the way of maintenance and updates.

Any compiler from a more recent era is just not going to be using VCG. 
LLVM, Microsoft de jour, JVM, etc.


-- 
Pure Personal Opinion | HoffmanLabs LLC