MMtk uses it  eg 

 

“I'm currently leaning towards building some kind of concurrent collector for 
MMTk” 

which Jeremie was referring to and is the only real competitor to the CLR at 
this point … yes you can build your own GCs  but these are not  really the LLVM 
one as its just a few very basic hooks then.  And note it is not just the 
shadow stack but  the whole easy to bolt in GC leads to one that is not very 
competitive. And  I don’t think Bitc  has time at the moment to write its own , 
when it does it will probably be a mostly pause less one …

 

 

 

And, regardless of whether you use compiled stack maps or the shadow stack, the 
actual collector -- the trickiest bit by far -- is still up to you. Which is as 
it should be. High-performance garbage collectors are strongly tied to a 
language implementation's object model and semantics. You might get decent 
performance from another language's GC, but you'll never get great performance, 
especially on pause times. And especially for BitC, it seems like pause times 
are much more critical than raw throughput...

 

 

Agree …but the poster was referring to MMtk…. If you want decent performance  
now on a Sandbox the CLR is better ( not just because of the GC)  ,  Bitc is 
also  likely to use  LLVM and its own GC later for fully compiled code ( I 
don’t think its practical to write your own compiler so your choice boils down 
to LLVM or compile via gcc )  .   If Jeremie wants to implement a BitC 
pauseless collector for LLVM that would be great but I don’t think its what he 
had in mind.

 

Ben

 

_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to