Re: AW: [DotGNU] Oberon.net compat

[Gopal V]

If memory serves me right, Patrick Pletscher wrote:
> > I think I've fixed all the ilasm-related and engine-related bugs
> > that caused
> > this test case to fail.  There were a number of issues:

Well, I'm still debugging the compiler .... it seems that our verifier is
a lot stricter (with respect to ECMA) compared to any other .NET implementation
flying around ... (which is a good thing for runtime safety :) .. So I think
we're being faced with these minor inconveniences... Should mature as more
people test this ...

> I think the developer of the compiler will perhaps change this, because he
> wants to change all the array-stuff, so that also ARRAY OF SET is supported...

Hmm... I suppose the Oberon.net guy started bootstrapping off the native 
Oberon and worked his way into Oberon.net .... That seems a cool way to
bootstrap directly from source :)

> Thanks again, sorry, if I can not help you very much common with
> ILAsm-stuff, but keep on your good work

Reporting bugs is important work ... I would have never looked at this
stuff if you hadn't started this thread ..

                ldarg      2
                ldarg      2
                ldobj      int32
                ldc.i4     1
                add
                stobj      int32

Seems to be code generated by the Oberon.exe ... (found inside oberon.exe,
in Obparse.Optarg) This is not being verified by ilrun and my simplified 
test case didn't work either ..

I am not sure this is a bug ... because IMHO stind.i4 seems to be the
right (or fast ?) instruction to use here ? .. I'm not reporting that as
I'm not really sure about all that ...

Gopal
-- 
The difference between insanity and genius is measured by success

stobj.il
Description: Text document