Arc Forumnew | comments | leaders | submitlogin
1 point by stefano 4522 days ago | link | parent

Do you think to-pointers are expensive in terms of space or of runtime(i.e. scanning the entire heap to look for them)? In term of space they aren't really expensive because you can reuse the space allocated by the object: as an example if you move a cons cell (2 words) you can use its first word to hold the new address of the cell. I've implemented a simple copying GC in C, if you want to have a look at it:

2 points by almkglor 4522 days ago | link

Actually the other problem I'm thinking of is the "size" of an object in the heap under management. I'm currently using C++ for the memory management stuff, and the base type Generic requires a virtual getsize() member function which just returns sizeof(Type); most C++ implementations just allocate a single vtable for all virtuals, and of course I also have recursive marking as virtual functions of Generic. Basically I am overloading the vtable of the objects to also encode the size of individual types.

(the vtable also serves as my "type code", and is probably better thought of that way)

So I can't safely replace a Generic object on the heap with a ToPointer object, because I'd lose access to the size (i.e. I lose the type code, which I'm using to determine the size). This means I can't sweep and finalize afterwards. The alternative is to keep the size separate from the type code/vtable, but then for each object I end up storing a size_t, which on most systems is probably the same size as a Generic* - so I don't save it either.

So basically my options are:

  type code (automagically inserted by C++ as the vtable)

  type code
Reusing the space of the object is possible, but to differentiate a to-pointer from a "real" object, I need to replace the type-code anyway, so I lose the size information.

Edit: Okay, I've been trying to hack through your posted GC. I don't quite get how the type code is encoded - is the heap split into regions? You seem to be masking the pointer itself to get the type.


1 point by almkglor 4522 days ago | link

Okay, I've figured bits of your GC out after sleeping, it seems that the pointer is a typed pointer. A pointer with a type of extended_tag (?) includes the type with the memory area, and this type is overwritten with a broken_heart type to turn it into a to-pointer.

However I intend to use the same copy-algorithm for message passing, so I can't overwrite the tag/type code and the data just because I've copied it, since the original could be reused.

Or maybe I should actually use a different routine for message passing and GC.


1 point by stefano 4521 days ago | link

> However I intend to use the same copy-algorithm for message passing

In this case, as you've already said, you can't overwrite the orginal type and you have to reserve extra space for the to-pointer (this means one extra word for every allocated object). This way another problem arises: you have to be sure that the to-pointer is clear every time you start a copy this is automatic after a GC, but not after a copy for message passing. This means that after every message passing, the whole heap must be scanned.

> Or maybe I should actually use a different routine for message passing and GC.

I think this is the best solution: passing a message and copying an object during garbage collection are enough different to make them distinct.


1 point by almkglor 4521 days ago | link

Yes, you're right. Looks like I'll end up having the message-passing use a std::map<> instead (which has O(log N) for each lookup, averaging O(N log N) for all memory areas involved in the message), and leave the to-pointers for the semispace destructor to determine which objects need to be cleaned up. Overhead either way.

Or again use a mark and sweep with copying instead of sharing of objects ^^. Well, not much different.

Memory management is hard. Let's go bytecoding!