Re: [Pnet-developers] Re: back to Emit problems

[Gopal V]

Hi,

> > might work.. But of course, we're trying to build
> a 
> > Re.Emit that generates valid code ;-)
> 
> I believe there are no guarantees about validity of
> code generated by
> S.R.E. It is even possible to insert some junky
> opcodes by accident
> by using wrong overload of Emit method (like
> Emit(Ldarg_s,
> 0xdeadf00ddeadf00dULL)). 

This is exactly why we can write an IL verifier test 
suite with this ... Since we can generate invalid 
opcodes we can test the verifier directly . I was
just joking that we are supposed to build a Re.Emit
that really works , not something that is tested
"extensively" with invalid cases .

> error during generation. And verification is far
> more complex then just valid opcodes.

It's fairly easy to do a check for some of the 
Re.Emit operations . I find it interesting that
neither of those actually do .

Of course, doing stack analysis to match types is
and should be out of the Re.Emit's job.

/me goes back to debugging good old reflection (not
emit)

Gopal


        
                
__________________________________
Do you Yahoo!?
Win a $20,000 Career Makeover at Yahoo! HotJobs  
http://hotjobs.sweepstakes.yahoo.com/careermakeover