Remove OMR JIT and dmr_C

[dibyendumajumdar]

There is no particular advantage in maintaining the OMRJIT backend as MIRJIT is much smaller and achieves almost similar performance. For high performance we already have LLVM.

[dibyendumajumdar]

Created omrjit branch to preserve the current implementation

[pfalcon]

Hey @dibyendumajumdar, great progress here! Is MIR really that good? Still didn't give it a try.

@itssamuelrowe: If you're interested why people first adopt Eclipse OMR, even maintain a fork, and then kill it out. (You would need to make up a rest of the story yourself of course ;-) ).

[itssamuelrowe]

@pfalcon I was thinking of using NJ and then I saw this thread. I should consider using LLVM JIT.

@dibyendumajumdar Is there any leaner version of OMR GC? Or some alternative?

[dibyendumajumdar]

If you need GC then you have to use the base OMR library. It is much smaller than LLVM. You can have a look at https://github.com/b9org/b9

[Udayraj123]

Hi @dibyendumajumdar,

I know I'm going off topic here(will delete later if needed), but this seemed like an easier way to reach you, so here I go -

I came across this project of yours while browsing on the 'omr' topic on github. I'm a maintainer of an open source repository called OMRChecker whose discovery relies on the omr tag(optical mark recognition). Currently it stands at #2 there (thanks to eclipse OMR removing the tag :D). It'd be great if you could help me out a bit by renaming your omr tag to something like eclipse-omr or omr-jit

Thanks and regards, Udayraj

[dibyendumajumdar]

@Udayraj123 removed the label as eclipse-OMR support was removed

[Udayraj123]

Thanks @dibyendumajumdar I appreciate it.