Clasp catchup #6
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This makes it so that call arguments are on the stack for both MV and regular calls. It also skips messing about with vm-values, which is really slow and dumb. It also makes the bytecode more regular for IR purposes. See also the older Clasp commit where I implemented this first: clasp-developers/clasp@fad16ca
It's special, and changing it when we add new opcodes is silly.
This simplifies both the compilation of and generated bytecode for SETQ, and is pretty trivial to implement. Implemented earlier in Clasp: clasp-developers/clasp@0ce7eae
This lets VM functions create closures and stuff without consulting the mutable m:*client* dynamic variable. This ssssshould make it possible for cross and native VM functions to coexist smoothly, but I haven't tried it.
This is like fdefinition, but can be optimized since the VM knows that the result is only called and not otherwise stored. So for example you can skip an fboundp check if your environment works such that unfbound function names are bound to functions that immediately signal UNDEFINED-FUNCTION. Since neither VM here can actually use this they just do FDEFINITION. I'm not totally sure about this instruction, honestly, but at least for now it's not like we're lacking opcode space. Added in Clasp earlier: clasp-developers/clasp@56dab78
This lets us compile MULTIPLE-VALUE-CALL without requiring that the loader environment give if, type predicates, or CL:FDEFINITION their usual definitions. We could also use it to compile code like (CL:FUNCALL FOO ...) more efficiently, but that would entail the compiler have a principled way to learn that it's allowed to optimize CL:FUNCALL calls, so I held off for now. When I implemented this in Clasp earlier I neglected to consider first class environments. Therefore this version gives FDESIGNATOR a constant parameter representing the loader environment, which Clasp's doesn't yet. Without this parameter, a name has to be looked up in "the current global environment", but that may differ from the loader environment if it even exists, which could lead to weird situations like (let ((x 'foo)) (multiple-value-call x)) doing something different from (multiple-value-call 'foo). Original clasp commit, which also added CALLED-FDEFINITION: clasp-developers/clasp@56dab78
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Brings cvm up to the 0.12 format I already hacked up in Clasp, and then a little farther to pending 0.13 since I realized there was a problem with the
fdesignatorinstruction.