bf23a8a1b7
* pretty-print now handles vectors
2007-01-13 20:37:54 -05:00
38082fc489
pretty-print now handles read-macros.
2007-01-13 18:51:23 -05:00
17c1b3ba10
added initial pretty-print file to lab
2007-01-13 18:34:35 -05:00
53b51e4088
* Tokenizer now recognizes "\r" in strings.
2007-01-13 14:39:58 -05:00
8b9dd5fd85
Added cgi.ss and cgi-errors-to-browser.ss to the lib directory.
2007-01-13 13:03:23 -05:00
1a7a4e1c2a
fixed bug in environ that causes substring to fail when one of the items is "foo=".
2007-01-13 13:02:03 -05:00
212224b4cf
* Added the primitive (environ):
...
returns an alist of (string x string) as obtained from the
system environ.
* Added the primitive env:
(env "key") retrieves the env value associated with key
(env "key" "val") sets the env value assocuated with key to val
(env "key" "val" overwrite?) same as above but it does not
overwrite an existing value of overwrite? is #f.
2007-01-13 00:42:37 -05:00
6359ca9ee4
Fixed copyright notice to be 2006-2007 instead of 2006.
2007-01-11 19:22:32 -05:00
09d9687fdd
* Cogen no longer generates (int x) where x is a fixnum.
2007-01-09 09:44:00 +03:00
ffc5ef557c
(int x) is no longer emitted by the cogen.
...
assembler handles (int x) and fixnums similarly.
2007-01-09 09:24:07 +03:00
63575db860
* makefile.ss is updated to use fork and waitpid in order to make
...
use of the dual-core capability of my macbook :-)
Compile-time is now 1.5secs real-time.
2007-01-01 02:48:17 +03:00
a5618ef877
* Fork, posix-fork, and waitpid are implemented.
2007-01-01 02:02:41 +03:00
eb24d17049
Fixed bug that caused mvcalls to lose their live masks.
2007-01-01 01:46:47 +03:00
a8e1b860bb
Live-masks are busted again.
...
Some closures that are live are not traced by the GC.
2006-12-30 22:52:37 +03:00
b10d954548
updated TODO
2006-12-29 13:47:24 +03:00
2ea03e0417
* Fixed bug in collecting bignums.
...
* time-it now displays the number of collections and bytes
allocated.
2006-12-29 13:45:30 +03:00
b507118f5d
Added libtimers.ss:
...
(time-it proc) runs proc and reports the time.
(time-it proc message) runs proc and reports the time, with message.
(time <expr>) evaluates expr and prints the running time.
2006-12-29 10:53:47 +03:00
d38880b777
Fixed bug in libfasl that caused reading a thunk to return the code
...
instead.
2006-12-26 12:03:43 +03:00
0b38b1ff30
ikarus-fasl.c plays along with the rtd-symbol game by setting and
...
referencing rtd symbols.
2006-12-26 11:46:23 +03:00
81b2f217a3
* reader can now read fasl objects.
2006-12-26 11:39:14 +03:00
f5f91712dd
* Fasl file reader almost complete with full record and code reader.
2006-12-26 11:25:48 +03:00
9c00592520
fasl-read handles negative fixnums.
2006-12-26 09:22:58 +03:00
3ddceaa164
fasl-read now handled pairs, positive-fixnums and the empty list.
2006-12-26 06:32:59 +03:00
d41a2c0096
fasl-write now returns void instead of the meaningless number it used to return
2006-12-26 06:25:35 +03:00
4fa2bdf0d9
tokenize recognizes #@ as a fasl header and calls $fasl-read (in
...
libfasl.ss).
2006-12-26 06:21:07 +03:00
903023592d
Updated todo.
2006-12-26 06:05:54 +03:00
cfe69af8f8
BUGS file added to keep track of some pending bugs.
2006-12-25 11:38:18 +03:00
f25cc3cb67
Pretty-printed gensyms:
...
* print-gensym now accepts the symbol "pretty" as a valid
argument.
* The writer prints #:pretty-name if the print-gensym value
is |pretty|.
* The reader accepts #:id as a valid syntax. The result is
a fresh gensym with id as a pretty-name.
2006-12-25 11:33:03 +03:00
e7d0a0a1cc
Tokenize now recognizes gensym syntax:
...
#{id0} => a fresh gensym with pretty-string id0
#{id0 id1} => an interned gensym with pretty-string id0
and unique-string id1.
id0 and id1 can have any identifier syntax including bar-quoted ids.
any number of whitespaces can be placed around id0 and id1.
2006-12-25 11:18:37 +03:00
158980aeea
The fasl-loader interns gensyms by default now.
2006-12-25 10:35:18 +03:00
b9a369197a
gensym->unique-string now interns gensyms in the gensym table.
2006-12-25 10:25:40 +03:00
b3a6744691
renamed oblist to symbol_table in pcb
2006-12-25 09:28:53 +03:00
711cabe8bb
removed all references to oblist from ss files
2006-12-25 09:20:09 +03:00
72653c5ef1
updated todo
2006-12-25 09:17:08 +03:00
4f5490039b
* (interrupt-handler) is added to handle INT signals.
...
* Cafes now customize the interrupt-handler to suppress the error
message during read that the default interrupt handler generates.
2006-12-24 12:53:01 +03:00
da9518cc49
added comment to libguardians about where the code came from.
2006-12-24 12:20:16 +03:00
06158d90e0
deleted Makefile
2006-12-24 12:09:21 +03:00
0a01d8b4b8
makefile.ss and ikarusdoc now use "/usr/bin/env ikarus" in the shebang
...
line.
2006-12-24 12:08:18 +03:00
1e572559b7
collect is added to the set of public primitives
2006-12-24 11:50:19 +03:00
f729a725d2
libinterpret is gone.
2006-12-24 11:27:53 +03:00
599f68aa84
* Ikarus now handles interrupts (e.g. ^C) by signalling an error
...
when an interrupt occurs.
2006-12-24 11:24:53 +03:00
7fed78fa9d
An alt-stack and interrupt-handler are added to ikarus-main.c. When
...
an interrupt occurs, the engine_ticks is set to 1 and the
interrupted field is set to 1 in the pcb.
2006-12-24 09:43:20 +03:00
fc609b4916
* lab/interupts program is added to experiment with sigaction and
...
sigaltstack.
added:
lab/interrupts/
lab/interrupts/Makefile
lab/interrupts/inf_loop.s
lab/interrupts/main
lab/interrupts/main.c
2006-12-24 09:25:01 +03:00
e69f0a5c80
* Assembler-output suppresses printing the closure-size; it prints
...
a newline instead.
2006-12-23 05:17:57 +03:00
b06b28f562
* Print-gensym is now set to off during assembler-output.
2006-12-21 18:11:43 +03:00
a41457ae66
* Heap and stack overflows now use interrupt-calls in order to push
...
the calls to the handlers outside of the main-line code.
2006-12-21 18:01:46 +03:00
0af5f6cc2a
* Engine handler is functional.
2006-12-21 17:49:30 +03:00
3aa22ea200
* Added 'add imm, mem' instruction format.
2006-12-21 17:16:33 +03:00
efae97a6e0
Actually, the profiling times show that memory-based tests can be
...
performed at around 3 secs per 1 billion checks (about 6 times
slower than register-based counters).
2006-12-21 16:32:45 +03:00
67a850e712
Profiling timing shows that we can do interrupt checks with cost:
...
* about 2.0 secs per 1 billion checks if the counter is kept in memory
* about 0.5 secs per 1 billion checks if the counter is kept in register.
2006-12-21 16:22:07 +03:00
Abdulaziz Ghuloum