Odd 'dead' primitives
Tim Rowledge
tim at sumeru.stanford.edu
Thu Oct 17 22:01:03 UTC 2002
I'm working on removing the obsolete primitive tables etc from the VM,
ready for VI4 stuff. There are several prims that have odd comments
associated with them that I don't recognise, so I'm hoping some alert
reader can recall what might be intended by:-
(123 primitiveValueUninterruptably)"@@@: Remove this when all VMs have
support" - there is a method calling this prim in
BlockContext>ifProperUnwindSupportedElseSignalAboutToReturn and another
in BlockContext>valueUninterruptably
"NOTE: When removing the obsolete indexed primitives,
the following two should go become #primitiveIntegerAt/atPut"
(147 primitiveObsoloteIndexedPrimitive) "primitiveReadJoystick"
(147 primitiveObsoleteIndexedPrimitive) "primitiveWarpBits"
- which reads a bit oddly and makes little sense to me, as if it got
reformatted by prettyprint sometime.
In addition, there are a mere ten prims with no module name, eight
relating to the webbrowser plugin stuff plus #actualScreenDepth and
#disablePowerManager. Anyone have strong thoughts on which plugins might
be better homes for them?
If we can clean this up, move a couple of high number prims and so on we
could drop to only 9 bits for primitive index in the method header,
leaving some for other uses. Getting rid of the obsolete prim tables and
shrinking the prim table will save a few kb.
tim
--
Tim Rowledge, tim at sumeru.stanford.edu, http://sumeru.stanford.edu/tim
Mommy! The cursor's winking at me!
More information about the Squeak-dev
mailing list
|