Commit Graph

296 Commits (f8e2ddd53f162c76a21379c5307de278dc71b04c)

Author SHA1 Message Date
Joe Groff 8687a7e283 Merge branch 'master' of git://factorcode.org/git/factor into row-polymorphism 2010-03-12 14:36:13 -08:00
Joe Groff dfbe1211f7 pre-chew the bite-sized morsels of stack-checker.row-polymorphism so they're easy for old people to digest 2010-03-11 14:09:18 -08:00
Slava Pestov 16ddd015d3 typed: forgetting a final class would throw an error while recompiling typed words depending on it 2010-03-12 10:22:35 +13:00
Joe Groff 26ff757de4 carve the tough, gamey steak of stack-checker.polymorphism into chewable morsels 2010-03-11 00:53:40 -08:00
Joe Groff b9bced9a5e update docs for unbalanced-branches-error 2010-03-09 18:42:10 -08:00
Joe Groff 34b29af245 non-polymorphic input parameter check was too strict: wouldn't allow ( x -- ) for ( x x -- x ), for example 2010-03-09 15:57:00 -08:00
Joe Groff 191ac353fd generalize stack effects so we can bootstrap with the stricter stack effect checking 2010-03-08 23:38:10 -08:00
Joe Groff 3abf1f1ef7 stack-checker.row-polymorphism: modify check-variables to enforce non-polymorphic stack effects 2010-03-08 19:44:01 -08:00
Joe Groff 21aa4632c8 refactor stack-checker.row-polymorphism a little 2010-03-08 19:32:25 -08:00
Joe Groff 5597ee691f add documentation about stack effect variable syntax and the effect on the stack checker 2010-03-08 17:17:47 -08:00
Joe Groff 8159a19187 combine unbalanced-branches-error and invalid-quotation-input into one error 2010-03-07 22:23:24 -08:00
Joe Groff 9e6f84bc24 declare effect on inputs to infer-if in non-literal case, so we get a better error than "unbalanced drop call/nip call" 2010-03-07 21:37:24 -08:00
Joe Groff 6689113519 unify inner-d-index when unifying branches; close a few other leaks where meta-d could have been popped without updating inner-d-index 2010-03-07 20:44:50 -08:00
Joe Groff 66687d6b21 reuse @ and _ from fry in placeholders for invalid-quotation-input errors rather than making our own redundant placeholder symbols 2010-03-07 20:08:35 -08:00
Joe Groff f353b13bbc add unit test from quotation that fails to infer during bootstrap.compiler 2010-03-07 20:07:30 -08:00
Joe Groff 4cd6ad840d add tests from old row-polymorphism implementation to stack-checker unit tests 2010-03-07 19:13:41 -08:00
Joe Groff 1512ed12f3 further improve error message for invalid quotation inputs using a table display with the actual quotation effects determined so far 2010-03-07 18:40:58 -08:00
Joe Groff ea4545e366 improve error message for invalid quotation inputs by referencing each declared-effect to its other references 2010-03-07 18:27:55 -08:00
Joe Groff 011a39457e clear away some rebar 2010-03-07 18:11:44 -08:00
Joe Groff 6b9a79159d tweak declared-effect to wrap the existing known instead of introducing a new value, so we don't confuse the compiler 2010-03-07 18:07:42 -08:00
Joe Groff b14d59030f add stack variable unification to M\ declared-effect infer-call* 2010-03-07 17:51:41 -08:00
Joe Groff bbbda64ee7 add a with-effect-here combinator that determines the effect of a scoped subset of the stack checker 2010-03-07 16:54:25 -08:00
Joe Groff 63ad397cc1 tweak recursive call site checking to consider declared-effects equivalent to their wrapped values 2010-03-07 11:55:47 -08:00
Joe Groff 339cc8f04e row polymorphism new approach: wrap polymorphic quotation inputs in a "declared-effect" value. M\ declared-effect infer-call* will then assert the effect of declared-effect values during the normal course of stack inference 2010-03-07 11:54:57 -08:00
Joe Groff 68dd644233 add unit test to show that polymorphic check breaks inference of inline recursive words 2010-03-06 12:28:09 -08:00
Joe Groff 9571bf6d4b give terminating stack effects a pass in the polymorphic checker 2010-03-05 21:51:38 -08:00
Joe Groff 15e6a7fa5e move error summaries for effect parsing errors to debugger vocab 2010-03-05 21:50:40 -08:00
Joe Groff 26e4bb818a reuse infer-call* for polymorphic inference, which handles curried quotations. tests can pass with real "each" etc. now 2010-03-05 15:20:55 -08:00
Joe Groff c9162c5e31 fix stack error in stack-checker.row-polymorphism 2010-03-05 14:34:50 -08:00
Joe Groff 4892bf1d47 refactor stack-checker slightly to fix circular dependency 2010-03-05 14:27:36 -08:00
Joe Groff a3033e885a make effect variables part of effect syntax, stored out of band in effect tuple 2010-03-05 13:30:10 -08:00
Joe Groff 1b1ccb71a4 make tests pass 2010-03-04 22:43:01 -08:00
Joe Groff 053ba583fc improve error reporting 2010-03-04 21:51:49 -08:00
Joe Groff 31640ea9c6 nested scopes were messing up stack checker state 2010-03-04 20:15:26 -08:00
Joe Groff 23de281186 initial implementation of row-polymorphism check 2010-03-04 19:30:08 -08:00
Slava Pestov af0ddd5985 Specialized arrays, structs and other objects responding to the >c-ptr / byte-length protocol can now be written to binary streams 2010-02-24 20:18:41 +13:00
Joe Groff 31a0d98e02 Merge branch 'master' into new-alien-pointers
Conflicts:
	basis/alien/c-types/c-types.factor
	basis/stack-checker/dependencies/dependencies.factor
2010-02-22 10:32:59 -08:00
Slava Pestov 23a1f0ed8c alien: some code cleanups and fixes 2010-02-23 01:28:56 +13:00
Slava Pestov eb3f8632dd stack-checker.alien: now that C types are words, the compiler can add dependencies on them when compiling alien words. This triggers the necessary recompilation when C types are redefined 2010-02-22 21:32:41 +13:00
Joe Groff c4cc70b92c stack-checker.dependencies: extend c-type-word method for depends-on-c-type to all words (so it works for non-c-types like void) 2010-02-21 23:13:12 -08:00
Joe Groff 310b3df2ec stack-checker.dependencies: add method for pointers to depends-on-c-type 2010-02-21 22:31:32 -08:00
Slava Pestov 2c34ecbdb5 stack-checker.dependencies: add depends-on-c-type 2010-02-22 19:25:09 +13:00
Slava Pestov c5a62b14e7 stack-checker.dependencies: fix load error 2010-02-20 12:08:14 +13:00
Slava Pestov 6501480a0e Fix two problems with recompilation: predicate constant folding was recording unsatisfied dependencies in some cases, and literal tuple instances of forgotten classes would cause problems for method inlining 2010-02-20 12:01:47 +13:00
Slava Pestov 60296be964 typed: only unbox final classes. Fixes bug reported by littledan 2010-02-18 02:39:12 +13:00
Slava Pestov f01487eb83 A few documentation fixes 2010-02-17 00:05:36 +13:00
Slava Pestov 747974f4ae vm: nuke unused (string>float) primitive 2010-02-08 08:16:56 +13:00
Slava Pestov 26cae00eeb Add code heap introspection primitive to VM, and make a code-blocks word in tools.memory that wraps a useful interface around this. Add a lookup-return-address word and use this in tools.disassembler.utils 2010-02-03 23:11:32 +13:00
Slava Pestov 0b278adb5a Test fixes 2010-02-03 23:11:31 +13:00
Slava Pestov 7de81976a5 Revert "macros: macro body is now defined in its own subword, for compile-time stack effect checking"
This reverts commit 24de7c52f0c3f21cfcdb80235cac7296b0401c85.
2010-02-03 23:11:29 +13:00