belatedly 'cvs remove src/pcl/iterate.lisp'
deleted dead code that I noticed when looking at NJF patch...
...DEFCONSTRUCTOR
...MAKE-CONSTRUCTOR
...LOAD-CONSTRUCTOR
...ENABLE-CONSTRUCTORS
...DISABLE-CONSTRUCTORS
...RESET-CONSTRUCTORS
...MAP-CONSTRUCTORS
...INSTALL-LAZY-CONSTRUCTOR-INSTALLER
...COMPUTE-CONSTRUCTOR-CODE
...SET-CONSTRUCTOR-CODE
...ADD-CONSTRUCTOR, REMOVE-CONSTRUCTOR, GET-CONSTRUCTOR
...CLASS-CONSTRUCTORS
added note that more (all?) of construct.lisp might be dead
(never checked into CVS separately, since I was travelling)
gave up on fixing BACKTRACE function name nastiness for 0.7.0,
wrote it up as bug 137 instead
experimented with bootstrapping under CLISP (which still
doesn't work, but at least fails at a new place now)
0.pre7.96:
(never checked into CVS separately, since I was travelling)
commented out code which uses SETF %FUN-NAME, since not
only do I know now that it doesn't work on the other
parallel representation of function names, I also know
now that it's insanely tricky to modify function names
correctly in the current representation (and somewhat
dangerous to try)
0.pre7.97:
merged NJF "No more SB-ITERATE" patch (sbcl-devel 2001-12-20)
made function names (as returned by FUNCTION-LAMBDA-EXPRESSION
and used by BACKTRACE) for DEFUNed functions mostly
reasonable again, e.g. PRINT instead of
"top level locall PRINT"
...made SETFable %FUN-NAME to encapsulate this
...renamed FUNCTION-DOC to %FUN-DOC for consistency
...got rid of NAMED-LAMBDA, since cold-load issues make it
awkward to use it in DEFUN where I intended to use it,
and there's no urgent other place to use it
...made %DEFUN use SETF %FUN-NAME
(There are still function-name infelicities left, especially
for closures where there's no good implementation of
SETFable FUN-NAME until weak hash tables work, but at
least most BACKTRACE entries look better now.)
changed publicly-visible DEF-FOO-style names to DEFINE-FOO:
DEF-ALIEN-ROUTINE, DEF-ALIEN-TYPE, DEF-ALIEN-VARIABLE,
DEF-SOURCE-CONTEXT (and a few other lexically-similar
DEF-FOO-style names as well)
defined deprecated wrappers using the old names
defined DEPRECATION-WARNING to support this
(gave in to the temptation to check in a version which had
only been "rebuilt" with slam.sh, not full make.sh)
changed default CHAR-NAME of unprintable ASCII characters not
specified in the ANSI standard to #\Nul, #\Soh, etc.
(Now DISASSEMBLE 'PRINT produces reasonable-looking output.
In fact, AFAIK it's not flaky any more, so it should
(finally!) be mergeable back into the main branch.)
made GET-FUNCTION-SEGMENTS return a list (as its comments
claim it does, and as its callers expect) even in the
(NULL SEGMENTS) case
renamed DEFUN REQUIRED-ARG to be consistent with all the
previously-renamed MISSING-ARG calls
Duh. WITH-SIMPLE-RESTART doesn't restart its operation. So it's
not a compiler bug, just the lack of an outer LOOP
wrapper in TOPLEVEL-REPL, which has been blowing the
system up. Gads.
deleted unused FAST-SYMBOL-FUNCTION and
OBJECT-NOT-FUNCTION-OR-SYMBOL-ERROR
replaced %COERCE-NAME-TO-FUNCTION and RAW-DEFINITION
with %COERCE-NAME-TO-FUN
s/coerce-callable-to-function/coerce-callable-to-fun/
removed FSET in favor of (SETF SYMBOL-FUNCTION)
DEFSTRUCT cleanups..
..got rid of old SLOT-ACCESSOR-FORM in favor of
%ACCESSOR-PLACE-FORM, so that now information about
the different types of raw slots is more nearly
centralized in *RAW-SLOT-DATA-LIST*
..moved target-only stuff from near %TARGET-DEFSTRUCT call in
%DEFSTRUCT into %TARGET-DEFSTRUCT
miscellaneous cleanups..
..added (SETQ *PRINT-CIRCLE* T) to the before-proper-printing
hacks in make-target-2.sh
fixed stupid *PRINT-CIRCLE*-related bug in OUTPUT-OBJECT
(introduced in the 0.pre7.76 changes)
miscellaneous DEFSTRUCT-related cleanups..
..There are no longer DEFSTRUCTs in defstruct.lisp itself,
so DEFSTRUCT-parsing stuff no longer needs to be
wrapped in EVAL-WHEN.
..moved DEFUN SLOT-ACCESSOR-FUNS to target-defstruct.lisp
..removed lotso /SHOW stuff from defstruct.lisp
..removed REMOVEMEs
"We release no software before its time.":-| (Hopefully the
interval between 0.6.13 and 0.7.0 will be shorter
than the interval between 0.6.12 and 0.6.13..)
error handling
The MAKE-SPECIALIZABLE-ARRAY call in DEFUN INTERNAL-ERROR-ARGUMENTS
can become MAKE-ARRAY, since M-S-A is something to make it easier to
build parts of the cross-compiler under the cross-compilation host,
and DEFUN INTERNAL-ERROR-ARGUMENTS is in x86-vm.lisp, which is not
part of the cross-compiler, and so is never compiled by the
cross-compilation host.
changed MAKE-VALID-LISP-OBJ from a MACROLET macro to a global
function, for clarity and easier debugging
deleted unused SET-VALUE macro from MACROLET in DEFERR
deleted code marked REMOVEME, accidentally left over from previous
debugging exercises
changed CONTEXT-PC-ADDR, CONTEXT-PC, CONTEXT-REGISTER-ADDR, and
CONTEXT-REGISTER functions to use unsigned representations instead of
signed representations, to conform to implicit assumptions in the
debug-int code inherited from CMU CL. (Without this, new type errors
are generated in infinite regress when we try to handle errors
involving negative fixnums, e.g. (BUTLAST NIL -1).)
tweaked stuff in test/ directory a little bit in anticipation of
setting up real regression tests
