[Sbcl-devel] MOP and mmap'd manardb

[John F. <jf...@ms...>]

We (MSI) have just released a development version of our memory mapped
MOP database. There's a lot that can be improved with it but it has a
test suite and has been bashing stuff out for a while.

http://github.com/ilitirit/manardb/tree/master

Object instantiation seems a little slow on SBCL (just my impression
compared to Allegro 8.1 -- could be completely wrong) and I wonder if
there is something I'm doing wrong?

Basically each mm-object of a class descended from mm-metaclass has a
normal instance %ptr slot which gives an index into our mmap'd
regions. I guess should play more with allocate-instance and avoid doing
initialize-instance. Any wisdom appreciated!

Unfortunately as all mmap slots are always boundp, we have to munge the
general initialize-instance to set default slot values. This is quite
unfortunate. Is there a more sensible way of doing this (along the lines
of the way structure slots are handled)?

The normal slots are mmap'd, which are persistently stored in the mapped
regions and are accessed via overridden slot-value-using-class. This is
just

(defmethod slot-value-using-class ((class mm-metaclass) (object mm-object) (slotd mm-effective-slot-definition))
  (declare (ignorable class))
  (funcall (the mm-slot-definition-reader (slot-definition-reader-function slotd)) object))

(defmethod (setf slot-value-using-class) (new-value (class mm-metaclass) (object mm-object) (slotd mm-effective-slot-definition))
  (declare (ignorable class))
  (funcall (the mm-slot-definition-writer (slot-definition-writer-function slotd)) new-value object))

I notice that SBCL does a similar thing with its normal slots,
but when I tried to hook my compiled lambdas into
sb-pcl::slot-definition-writer-function, it crashes. Is that because I
am doing it wrong(tm) or is there some magic to follow?