dynamic library cost (was RE: libtool, java woes)

Jonathan P. Olson olson@mmsi.com
Fri Apr 13 08:32:00 GMT 2001


Even in Java, it's often not possible. Consider how Java AWT 
implementations
typically store native peer objects in `long' variables. Yeah, this is 
totally evil but unfortunately Java doesn't provide any way to declare 
`native' object references.
On Friday, April 13, 2001, at 08:09 AM, dewar@gnat.com wrote:
> <<Seems like the compiler would have to emit a ton of debugging 
> information
> to specify the contents of each word of the stack frame at each PC
> location
> within the program. Wading through all this mess in a GC would likely 
> be
> more costly than just conservatively scanning the stacks.
>>>>> Well there is also the concern that in a language like C, type accurate
> scanning is not possible (because of free unions), so you still end up
> with conservative guesses in any case, and there is some question as to
> whether there is enough gain. Yes, it is definitely a gain to go to
> a 100% type accurate solution (as would be possible with a language like
> Java or Ada), since then you can count on no memory leaks, but if you
> only go part way, it is not clear that it is worth it.



More information about the Java mailing list

AltStyle によって変換されたページ (->オリジナル) /