- Oct 18, 1997
-
-
dtc authored
-
- Oct 15, 1997
-
-
dtc authored
and tan, that provide the result type and work with union types. The result range which is problematic is not calculated. Removed the old versions.
-
- Oct 09, 1997
- Oct 08, 1997
-
-
dtc authored
though it's not used and could just as well have been conditionalised out.
-
dtc authored
results.
-
dtc authored
o Fixed bug in merge-types-aux where incorrect results were returned. o Enhanced interval-range-info to allow the split point to be specified instead of always using 0. (Used by the expt optimizer.)
-
- Oct 05, 1997
- Oct 02, 1997
- Sep 29, 1997
-
-
dtc authored
-
dtc authored
and can thus use the fast locking; adjust feature accordingly.
-
dtc authored
-
dtc authored
-
dtc authored
Add fixnum and unsigned truncate VOPs, and constant cases to a avoid unnecessary checking for division by zero: fast-truncate/fixnum=>fixnum fast-truncate-c/fixnum=>fixnum fast-truncate/unsigned=>unsigned fast-truncate-c/unsigned=>unsigned fast-truncate-c/signed=>signed Add fixnum ASH VOPs: fast-ash-c/fixnum=>fixnum fast-ash-left/fixnum=>fixnum
-
dtc authored
are not being check when choosing cached prefilter functions. This lead to the same prefilter being used for the near-cond-jump and near-jump instructions which broke much of the jump location labelling. The workaround uses different filter functions for each, forcing the creation of separate prefilter functions.
-
dtc authored
for compat. with clim-sys. Rework the with-lock-held macro; bug fixes; faster version for the Pentium using new %instance-set-conditional inline function.
-
dtc authored
New VOP slot-set-conditional to atomically test and set a slot. New VOPs instance-set-conditional and instance-set-conditional-c to atomically test and set instance slots; translations of new function %instance-set-conditional.
-
dtc authored
-
- Sep 24, 1997
- Sep 22, 1997
-
-
dtc authored
life-time analysis and targeting which usually gives better code.
-
- Sep 20, 1997
-
-
dtc authored
object file but byte compiling so should have been looking for a byte-fasl object file.
-
dtc authored
Elfun-derive-type-union didn't properly handle the case of elfun-derive-type-1 returning a union type like (or float (complex float)) - it does now. Elfun-derive-type-1 is slightly enhanced so that if the input is a complex type, the result is a complex type of the same. Before, we just returned default-type. Also, it's a bit smarter now so that if the result is supposed to be (or float (complex float)), we try to refine float to single-float or double-float if possible.
-
dtc authored
case where the numeric-type list contained complex numbers. No merging is done in this case, and that complex type is simply added.
-
- Sep 16, 1997
- Sep 15, 1997
- Sep 13, 1997