[erlang-questions] Deforesting tuple updates

Richard Carlsson richardc@REDACTED
Thu Feb 22 16:11:08 CET 2007


James Hague wrote:
>> you are likely to end up with some pointers from the old generation into
>> the new generation, and this breaks the main invariant that the GC is
>> based on. The result will be an emulator crash.
> 
> That's only if you put a pointer into a tuple, right?  If the VM
> verified new value was a fixnum or atom, then destructive updates
> should be okay (after the compiler verifies that the array isn't
> shared, as Joel originally mentioned).

Quite.

     /Richard




More information about the erlang-questions mailing list