|
|
1.1 root 1: Fri Jan 13 15:05:29 1984 by jkf
2: the 68k version will now record function argument
3: counts for lambdas. It still doesn't do lexprs (even those
4: with &optional and &rest forms).
5:
6: Wed Dec 7 16:13:08 1983 by sklower
7: changed profiling code to refer to mcnts instead of
8: mcounts so that the C variable could be so renamed
9: to make sure that all names are distinguishable
10: in the first 6 characters. This change is invisible
11: to users.
12:
13: Tue Nov 22 08:59:34 1983 by jkf
14: added the -W switch. After compilation (but before assembly)
15: it aborts the compilation with an error status if there were
16: any warnings.
17: (now: version 8.39)
18:
19: Tue Oct 25 14:45:48 1983 by jkf
20: fixed it so that if a lisp error occurs when compiling a function,
21: the name of the function being compiled will be printed
22: (now: version 8.38)
23:
24: Mon Oct 10 07:42:54 1983 by jkf
25: fixed bug in liszt (tlev)
26: (now: version 8.37)
27:
28: Sat Jul 30 15:37:40 1983 by layer
29: open code vset. Move all vector stuff to vector.l.
30: Files: func.l, decl.l, vector.l, and cmake.l
31: (now: version 8.33)
32:
33: Mon Jul 25 21:29:54 PDT 1983 by layer
34: fixed 'liszt' to mung object file header for the sun
35: File: tlev.l
36: (now: version 8.32)
37:
38: Mon Jul 25 07:05:49 1983 by jkf
39: fix bug whereby -mr would cause an 'comma not in backquote error'
40: because the escape character was / instead of \ when reading
41: /usr/lib/lisp/autorun.xxx
42: (now: version 8.31)
43:
44: Sat Jun 25 13:28:10 1983 by jkf
45: Turn off uctolc converstion before printing bindtab.
46: This is useful for liszt in opus 38.65 and greater since it
47: will eliminate unnecessary |'s in the bindtab if uctolc
48: is need to compile the file
49: (now: version 8.29)
50:
51: Thu Jun 16 21:52:59 1983 by jkf
52: fixed compilation of &aux for vax version. Now &aux (foo foo)
53: will work if foo is special.
54: (now: version 8.28)
55:
56: Wed May 4 18:32:46 1983 by layer
57: added sun autorun header to io.l
58: (now: version 8.27)
59:
60: Thu Mar 24 08:24:53 1983 by jkf
61: -x file will now be placed in the same place as the output
62: file (instead of the input file).
63: (now: version 8.25)
64:
65: Sat Mar 19 17:49:18 1983 by jkf
66: mixed in the 68k liszt source.
67: (now: version 8.24)
68:
69: Mon Mar 14 13:14:26 1983 by jkf
70: liszt used to open code 'times' if all operands were fixnums.
71: Thus (times 256 256 256 256) would open code to a 0 fixnum.
72: This is clearly wrong and liszt will now not convert times, add,
73: etc to their fixnum equivalents. If you want fixnum semantics
74: then you must use fixnums.
75: Another bug was fixed: d-functyp would return the wrong thing
76: give a name bound to a foreign function. now it returns 'lambda.
77: files affected: datab.l decl.l
78: (now: version 8.23)
79:
80: Tue Feb 22 08:43:44 1983 by jkf
81: force jump to vecindexerr to use 'jmp', since the object
82: file might be large and the assembler is stupid
83: (now: version 8.22)
84:
85: Thu Feb 10 20:39:42 1983 by jkf
86: open coded vectorp and vectorip, vsize-...
87: (now: version 8.21)
88:
89: Thu Jan 20 01:57:01 1983 by layer
90: Added three features to liszt command line processing:
91: -e <expr> evaluates <expr> before compilation.
92: -i <filename> loads <filename> before compilation.
93: -S -o filename names .s file.
94: file: tlev.l
95: (now: version 8.20)
96:
97: Mon Jan 17 09:41:12 1983 by jkf
98: added arg number checking and open coded &keywords.
99: (now: version 8.19)
100:
101: Sun Jan 16 10:05:01 1983 by jkf
102: fixed nasty bug in d-exp which would show up if a macro
103: returned a recursive call it itself
104: (now: version 8.18)
105:
106: Wed Jan 12 10:43:53 1983 by jkf
107: added open coding of vref functions
108: (now: version 8.17)
109:
110: Wed Oct 27 20:24:47 1982 by jkf
111: removed references to sys_ functions, replaced them with
112: sys: functions so code will be more portable.
113: file: tlev.l
114:
115: Wed Oct 27 08:15:14 1982 by jkf
116: discovered that making 'declare' a function is a bad idea because
117: if the compiler executes an interpreted function with local
118: declarations, those local declarations will seem like declarations
119: for the function begin compiled. Thus declare is no longer a
120: special function in the compiler: it must be seen by the compiler
121: to have an effect. Added the user callable function liszt-declare
122: which will have an effect just like declare used to: the compiler
123: will recognize it when compiling and it can be evaluated by a
124: user function.
125: (now: version 8.16)
126:
127: Mon Oct 25 22:55:37 1982 by jkf
128: removed If macro definition from cmacros.l . It is now in
129: the default franz.
130:
131: Mon Oct 25 09:15:18 1982 by jkf
132: catch the sigterm signal and die after removing the /tmp file.
133: (now: version 8.15)
134:
135: Tue Oct 19 15:56:07 1982 by jkf
136: fixed bugin cc-arg wherein a (arg (foo)) would be compiled
137: incorrectly (it was calling zerop on a non number).
138: Added checks for cmacro and macro-autoload properties.
139: When doing macro expansion, had it stop if the same car
140: was returned.
141: (now: version 8.14)
142:
143: Mon Oct 18 23:27:14 1982 by layer
144: Fixed bug in function cc-quote. Parens in an If statement
145: were messed up.
146:
147: Sat Oct 9 04:35:23 PDT 1982 by layer
148: Np-reg, Lbot-reg, oLbot-reg are now constants. Changes
149: made to expr.l and io.l.
150:
151: Thu Oct 7 01:02:26 1982 by jkf
152: fixed bug in the compiler. The compiler had neglected to enforce
153: the rule that 'macros' not be defined as local functions.
154: Now that is checked (modification to tlev.l).
155:
156: Wed Oct 6 22:53:36 1982 by jkf
157: added check to make sure that a file doesn't declare a function
158: to be local that has already had a compiled call made to it.
159: This is done by placing a t under indicator g-stdref for all
160: symbols called in the 'standard' way.
161: (now: version 8.13)
162:
163: Wed Oct 6 13:29:27 PDT 1982 by layer
164: put the loading of chead.l (in array.l datab.l decl.l expr.l
165: fixnum.l funa.l funb.l func.l io.l tlev.l util.l) under
166: include-if control. Also changed chead.l.
167:
168: Tue Oct 5 23:36:09 PDT 1982 by layer
169: removed function d-call from expr.l, since it isn't
170: used anymore.
171:
172: Tue Oct 5 23:00:09 PDT 1982 by layer
173: added sccs'ed file scmake.l, so that an all interp'd
174: version of the compiler can be made without deleting
175: all the .o files. Changes were also make to the Makefile.
176: snliszt (formerly slownliszt) now depends on source
177: files (like nliszt depends on .o files).
178:
179: Tue Oct 5 22:22:59 1982 by jkf
180: added argument checking. This involved mods to expr.l to
181: do the checking and to tlev.l to save argument number
182: info while compiling. The file /usr/lib/lisp/fcninfo.l
183: contains a description of what is known about C functions.
184: (now: version 8.12)
185:
186: Mon Oct 4 14:23:29 1982 by jkf
187: turn off [load ...] message when loading in lisztrc file
188:
189: Tue Sep 28 09:30:36 1982 by jkf
190: change tmp file name back to jkfnnnnn
191: (now: version 8.11)
192:
193: Wed Jul 21 13:43:56 PDT 1982 by layer
194: function "liszt" in tlev.l now returns the true assembler exit
195: status (if the compile goes that far).
196:
197: Sat Jul 3 18:01:25 1982 by jkf
198: open code getaux and getaccess
199: (now: version 8.10)
200:
201: Wed Jun 30 18:58:32 1982 by jkf
202: look for liszt rc files without extensions. now we look in
203: 12 places for the lisp init file:
204: { . , $HOME } { .lisztrc , lisztrc } { .o , .l , }
205: (now: version 8.09)
206:
207: Thu May 27 08:19:00 1982 by jkf
208: fixed bug in d-fixnump which caused liszt to assume too often that
209: arguments are all fixnums
210: (now: version 8.08)
211:
212: Wed May 12 13:46:03 1982 by jkf
213: new declaration scheme died when the object being declared
214: was a list. e.g (declare (fixnum (foo fixnum flonum)).
215: Fixed declare-handler so that cases like this are ignored.
216: (now: version 8.07)
217:
218: Fri May 7 19:28:04 1982 by jkf
219: allow (function (nlambda ()...)) and (function (lambda () ..))
220: (now: version 8.06)
221:
222: Thu May 6 15:03:50 1982 by jkf
223: support local declarations. They are stacked on g-decls.
224: localf and macarray don't fit in too well, they should
225: be fixed.
226: fixnum declarations now have meaning. They will allow the
227: compiler to convert from functions like 'add1' to 1+.
228: 'declare' is now a function in the compiler so declarations
229: may be fasl'ed in. [this was removed Oct 27, 1982]
230:
231: Thu Apr 22 20:48:25 1982 by jkf
232: #' returns a symbol now, instead of the getd of the symbol.
233: This was required for maclisp compatibility.
234: (now: version 8.05)
235:
236: Wed Apr 21 07:50:41 1982 by jkf
237: load in init file before compiling. look in . first, then $HOME.
238: look for .liszt.o, .lisztrc.l, lisztrc.o, lisztrc.l
239:
240: Upon interrupt, remove temp file and exit
241: (now: version 8.04)
242:
243: Mon Apr 12 23:22:37 1982 by jkf
244: open code 'function'. Now it will cause another function to be
245: generated and compiled and the function call returns the
246: bcd header for the newly created function
247: liszt-process-forms may be filled with forms to compile while
248: compiling a form. Unlike liszt-eof-forms, liszt-process-forms
249: are compiled at the next opportunity.
250: (now: version 8.03)
251:
252: Wed Mar 31 08:47:46 1982 by jkf
253: fixed bug in which (> (or nil 9) 0) would return true. cc-eq
254: now rebinds g-trueop and g-falseop. (now: version 8.02)
255:
256: Wed Mar 31 08:24:27 1982 by jkf
257: added a new file to liszt: lversion.l It is not sccsed and it
258: just contains the version number. This will allow the version
259: number to be changed with ease whenever a modification is made.
260: start it at 8.01 (now: version 8.01)
261:
262: Fri Mar 19 11:17:12 1982 by jkf
263: did distribution. (now: version 8.00)
264:
265: Fri Feb 19 09:56:50 1982 by jkf
266: fixed e-docomment so that newlines in the comment will not
267: cause assembler problems [io.l]
268:
269: Wed Feb 17 12:46:24 1982 by jkf
270: to version 8.0 (to correspond with Opus 38)
271:
272: Wed Feb 10 21:28:41 1982 by jkf
273: fixed a bug in the return function and a possible bug in go.
274: both bugs have to do with returning or going through a
275: catch or errset, which is a pretty strange thing to do anyway.
276: The fix was to keep track of the number of catches and errsets
277: before the prog to be returned from or gone into.
278: I modified d-pushframe to always push 3 args, so we can know
279: how long each frame we pop will be.
280: Sat Dec 5 11:58:36 1981 by jkf
281: open coded boole (mods to decl.l and funa.l). used the open
282: coding capability of fixnums.
283: Added three internal functions fixnum-BitXor, fixnum-BitAndNot
284: fixnum-BitOr, which are the three functions which the vax
285: can do.
286: Mon Oct 26 21:15:36 1981 by jkf
287: fixed bug in d-supercxr (funa.l) which made this function
288: (cond ((setq y (cxr n zip)) (print 'hi))
289: (t (print 'lo)))
290: always prints 'hi'. The problem was that only jump on true was
291: checked if the value of a cxr was stored somewhere. In the case
292: above we want to jump on nil.
293:
294: Sat Oct 24 16:41:03 1981 by jkf
295: -*- to version 7.1 -*-
296: Added the 'a' option to liszt, which if set will cause a
297: special form to be output after a function is defined. that special
298: form will put a 'fcn-info' property on the function's property
299: list. the form of the fcn-info property is
300: (argdesc compileinfo)
301: argdesc may be nil or (x . y), the later meaning that there are a
302: minumum of x args required and a maximum of y. Currently lexprs
303: just put 'nil' there, but soon I will introduce a way to declare
304: min and max args for lexprs via a declare. The compileinfo
305: is a string saying what file this came from and when it was
306: compiled.
307:
308: Wed Oct 21 20:19:53 1981 by jkf
309: added functions <&, =&, and >& which are fixnum only
310: versions of <,= and >.
311:
312: Tue Oct 20 22:14:41 1981 by jkf
313: fix < and > compiling so that they are only open coded if
314: we can be sure that both operands are fixnums. For now this
315: amounts to checking that one of the operands is a fixnum.
316: We should augment to this check also for type declarations.
317:
318: Sat Oct 17 11:47:50 1981 by jkf
319: since most fixnums are in the range 0 to 1023 we can probably
320: win by doing inline checking in that range. Thus I've added
321: the function d-fixnumbox which does an inline reboxing if
322: the number is between 0 and 1023 and otherwise calls qnewint.
323: The file changed was fixnum.l
324:
325: Thu Oct 8 23:35:34 1981 by jkf
326: added cc-= to open code = if possible. It looks for a fixnum
327: as one of the arguments and converts the = expression to
328: (eq (cdr ..) (cdr ..)) which should do the trick.
329:
This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.