- Apr 21, 1992
-
-
wlott authored
-
wlott authored
some random subtype of block-annotation. People who want to use control-analyze can :include block-annotation in their info structure, pass control-analyze the constructor function, and be happy campers.
-
wlott authored
block-annotation struct which is :included by ir2-blocks. This is so the byte-code stuff can use control-analyze also.
-
wlott authored
back ends :include in their block-info annotations.
-
wlott authored
instead of assuming that it's going to be an ir2-component. Pass #'make-ir2-block to control-analyze in native-compile-component now that control-analyze takes the constructor function as an argument instead of assuming it should make ir2-blocks.
-
wlott authored
longer exist.
-
- Apr 19, 1992
-
-
wlott authored
Changed the places fop-fset and fop-fdefinition are dumped to just dump the name instead of using fop-normal-load/fop-maybe-cold-load around them. Otherwise, uninterned symbols as names will not be EQ preserved.
-
- Apr 18, 1992
-
-
ram authored
invariant that :LET lambdas always have a combination and bind node. We fix the problem we were running into by being more careful about deleting LETs when we see a combination, since the REF for the combination might have already been deleted. We might have been the 2n'd to last ref.
-
- Apr 15, 1992
- Apr 14, 1992
-
-
ram authored
callee are in different components.
-
ram authored
:NOTINLINE functional. The inlinep information must be retained, since we count on :NOTLININE calls never being local call converted.
-
ram authored
the argument is a FUNCTION form.
-
ram authored
-
wlott authored
-
wlott authored
-
wlott authored
-
wlott authored
than function-header-type should go though closure-tramps instead of worrying about the exact header type.
-
wlott authored
-
wlott authored
-
wlott authored
-
wlott authored
lowtag instead of a list-pointer lowtag.
-
- Apr 12, 1992
- Apr 09, 1992
-
-
ram authored
longer TR when we convert a LET in a non-tail position. Fixed comment in MAYBE-CONVERT-TO-ASSIGNMENT to make clear that we only attempt conversion when there is <= 1 non-recursive call. This ensures that we will never need to back out of assignment conversion.
-
ram authored
Deletion of the REF should suffice, and will avoid problems with deleting twice.
-
- Apr 08, 1992
-
-
wlott authored
try to run when the 1 is real. Otherwise it does fun things like call minusp on #C(0 1).
-
- Apr 07, 1992
-
-
wlott authored
compiler-macroexpand{,-1}, and compiler-macro-function.
-
- Apr 04, 1992
- Apr 02, 1992
-
-
wlott authored
decls from (or function null list) to (or function null) because lists were only in there to allow bootstrapping off of the old rt system.
-
ram authored
Fixed function consistency checking to work better on deleted functions. Added a condition handler in PRINT-ALL-BLOCKS.
-
ram authored
continuation type assertions conflict.
-
wlott authored
MACROLET-DECLARE:ALLOW. When declaring things about symbol macros, type declarations just wrap (the type ...) around the expansion, special declarations signal an error, and ignore/ignorable declarations are ignored.
-
wlott authored
SHADOW-ALREADY-PRESENT:WORKS.
-
- Apr 01, 1992
-
-
wlott authored
if between 2 and 3 (inclusive) values were expected. Also, spiffed up the case where > 3 values were expected.
-
wlott authored
if between 2 and 6 (inclusive) values were expected. Also, spiffed up the case where > 6 values were expected.
-
wlott authored
if between 2 and 6 (inclusive) values were expected. Also, spiffed up the case where > 6 values were expected.
-
ram authored
tail set have no RETURN.
-