- Feb 20, 1990
-
-
ram authored
initially making it.)
-
ram authored
compiled-debug-blocks. Changed compiled-location to compiled-debug-variable. Added DEBUG-FUNCTION-TLF-NUMBER and DEBUG-FUNCTION-START-PC slots. Allow the DEBUG-FUNCTION-ARGUMENTS to be any kind of vector so that it can be an i-vector in fixed-arg lambdas with no deleted arguments.
-
ram authored
localify so that GGS doesn't get confused by objects being trashed with GC forwards. In TRANSPORT-FUNCTION, we must be prepared to find closure-entry objects (and treat them the same as normal entries.) Also, we must check if the entry for a closure is pure before transporting the entry. In various places, I had to flush uses of INFO, since this accesses global variables (*info-environment*).
-
- Feb 18, 1990
- Feb 14, 1990
- Feb 12, 1990
-
-
ram authored
-
ram authored
-
ram authored
-
ram authored
DEFSTRUCT-SLOT-DESCRIPTION to use INFO TYPE DEFINED-STRUCTURE-INFO.
-
ram authored
-
ram authored
-
ram authored
-
ram authored
TYPE PRINTER. These get around problems with the printer &c thinking that the DD from compilation of a structure definition is the actual loaded one.
-
ram authored
-
- Feb 11, 1990
- Feb 10, 1990
- Feb 09, 1990
-
-
ram authored
that it will preserve locality. With the new function representation, we need to transport both the function entry and function constants in LOCALIFY to ensure that they are near each other. Doing this transporting also prevents the recursion in TRANSPORT-FUNCTION from losing now that there can be circular references. Also changed INLINEP to know about (INFO FUNCTION INFO ...). Blew away plist hackery, as that info is all maintained by gloabldb now. Hacked on SAVE-STAND-ALONE-LISP so that it should work, but it needs to blow away debug info, etc.
-
ram authored
print the lambda with bounded depth.
-
- Feb 08, 1990
- Feb 07, 1990
-
-
ram authored
them.
-
ram authored
%TIME with the form wrapped into a lambda. This helps interpreted uses of TIME, mainly by reducing the spurious elapsed time for preprocessing the huge form, but should also produce somewhat more accurate results, since the code in %TIME always runs compiled. Theoretically this could have a slight negative effect on the accuracy of compiled uses by introducing a call, but this should be well down in the noise, since call overhead should be well less than the clock resolution.
-
ram authored
reason is that it lets us load the miscops which currently still have "fasl" type, but this also has the nice side-effect of giving a better error if someone tries to load an old "fasl" file.
-
- Feb 06, 1990
- Feb 05, 1990
-
-
wlott authored
-
- Feb 03, 1990
-
-
wlott authored
Added lisp register names.
-
- Feb 01, 1990
-
-
ch authored
-