- Gambit-list - mailman.iro.umontreal.ca

Gambit bugzilla
by Marc Feeley 24 Jul '06

24 Jul '06

I am pleased to announce that Gambit now has a bug tracking system (which is powered by bugzilla). If you want to report a bug, suggest a feature, or search for reported bugs please do so here: http://www.iro.umontreal.ca/~gambit/bugzilla/ This should be easier to manage that searching the mailing list. The mailing list should still be used to get general help. The bug tracking system should be used when you know, or strongly suspect, that Gambit is not behaving properly, or if you have suggestions for improvements. To submit new bug reports and to get email when new bugs are reported, you need to create an account at: http://www.iro.umontreal.ca/~gambit/bugzilla/createaccount.cgi Creating an account is a 1 minute process (give email address and real name and then retrieve password in your email). Marc

1 0

Re: [gambit-list] Hakmem bitcount.
by Bradley Lucier 24 Jul '06

24 Jul '06

(I'm replying to the mailing list too, since they're CC'ed on the original.) On Jul 23, 2006, at 10:01 PM, Marc Feeley wrote: > I could add (fxbit-count n), as per R6RS, and inline it to > ___FIXBITCOUNT which can implement hakmem. The bignum case could > be build on top of fxbit-count. Does that make sense? Sure. Sounds like the GVM is getting closer and closer to milli-code ... ;-) Brad

1 0

Hakmem bitcount.
by Bradley Lucier 24 Jul '06

24 Jul '06

Marc: This comp.lang.lisp thread http://groups.google.com/group/comp.lang.lisp/browse_thread/thread/ b1cdf3a5106abe59/545e852f2be87053#545e852f2be87053 discusses various fixnum bitcount algorithms, so I tried these three: (declare (standard-bindings)(extended-bindings)(block)(fixnum)(not safe)(not interrupts-enabled)) (define (bits n) ;; remove one bit at a time (let loop ((i 0) (n n)) (if (zero? n) i (loop (+ i 1) (##fixnum.bitwise-and n (- n 1)))))) (define (my-bitcount x) ;; do 6 at a time, from _num.scm (define (at-most-6 x) (##u8vector-ref '#u8(0 1 1 2 1 2 2 3 1 2 2 3 2 3 3 4 1 2 2 3 2 3 3 4 2 3 3 4 3 4 4 5 1 2 2 3 2 3 3 4 2 3 3 4 3 4 4 5 2 3 3 4 3 4 4 5 3 4 4 5 4 5 5 6) x)) (define (count x n) (if (##fixnum.< x 64) (##fixnum.+ n (at-most-6 x)) (count (##fixnum.arithmetic-shift-right x 6) (##fixnum.+ n (at-most-6 (##fixnum.bitwise-and x 63)))))) (count x 0)) (define (hakmem-bitcount n) (let* ((i (+ (##fixnum.bitwise-and n #x5555555) (##fixnum.bitwise-and (##fixnum.arithmetic-shift-right n 1) #x5555555))) (i (+ (##fixnum.bitwise-and i #x3333333) (##fixnum.bitwise-and (##fixnum.arithmetic-shift-right i 2) #x3333333))) (i (+ (##fixnum.bitwise-and i #xf0f0f0f) (##fixnum.bitwise-and (##fixnum.arithmetic-shift-right i 4) #xf0f0f0f))) (i (+ i (##fixnum.arithmetic-shift-right i 8))) (i (+ i (##fixnum.arithmetic-shift-right i 16) (##fixnum.arithmetic-shift-right n 28)))) (##fixnum.bitwise-and i #xff))) ;; time them, including calling the system routine (time (do ((i 0 (+ i 1))) ((= i 20000000)) (bits i))) (time (do ((i 0 (+ i 1))) ((= i 20000000)) (bit-count i))) (time (do ((i 0 (+ i 1))) ((= i 20000000)) (my-bitcount i))) (time (do ((i 0 (+ i 1))) ((= i 20000000)) (hakmem-bitcount i))) which yields [brad:~/Desktop] lucier% gsi bittest (time (do ((i 0 (+ i 1))) ((= i 20000000)) (bits i))) 927 ms real time 572 ms cpu time (565 user, 7 system) no collections no bytes allocated no minor faults no major faults (time (do ((i 0 (+ i 1))) ((= i 20000000)) (bit-count i))) 8225 ms real time 4446 ms cpu time (4394 user, 52 system) no collections no bytes allocated no minor faults no major faults (time (do ((i 0 (+ i 1))) ((= i 20000000)) (my-bitcount i))) 1067 ms real time 588 ms cpu time (582 user, 6 system) no collections no bytes allocated no minor faults no major faults (time (do ((i 0 (+ i 1))) ((= i 20000000)) (hakmem-bitcount i))) 638 ms real time 352 ms cpu time (349 user, 3 system) no collections no bytes allocated no minor faults no major faults From which I conclude that (a) doing 6 at a time really isn't better than doing one at a time, (b) as usual, the trampoline takes an enormous amount of time, (c) the hakmem approach doesn't help too much on average, and (d) it would be best to find a way to specialize bit-count (and other small routines like it that aren't single GVM instructions) for fixnum arguments. Brad

1 0

multiple main functions!
by Ryan Prescott 21 Jul '06

21 Jul '06

Hi, I've perused the archives of the Gambit-C mailing list looking for info on this issue so I apologize if this is a well known issue. I've got a C application that I want to use a gambit thread (actually a termite process). I have a main function defined in C and a library defined in scheme that I want to init/call from C. Unfortunately, I get multiple main functions when building and the msvc linker doesn't like that too much. Being brave, I instructed the linker to ignore the multiple main error condition and "just use" my C main func. This does NOT seem to result in a stable situation for the scheme world, as when I call __setup() I get an exception. Note that the exception is handled by the gambit runtime, but there definitely seems to be some init time badness. The termite process does not respond to messages, etc. I also tried defining ___LIBRARY during C builds to no avail. This did get rid of the multiple main error, but I still get the exception. Is there any way to get gambit to not generate a main function? I guess I was looking for something like chicken's -embedded option. I had some amount of hope when perusing gambit.h at the -flat option, but I think this requires me to resolve functions at startup manually. (chicken doc reference http://galinha.ucpel.tche.br/Embedding) FYI I'm building the scheme code with: gsc -f -check myfile.scm Thanks, Ryan

2 2

stepper hook
by Christian 21 Jul '06

21 Jul '06

Hello I want to add code to store the current continuation of each step in the single-stepper (to be able to feed it to ##repl-within or similar). How can I do it? I have found the ##main-stepper hook and the (##step-handler leapable? $code rte execute-body . other) definition. I've tried: (define conts '()) (define (stepper-init) (set! ##main-stepper (vector (make-vector 7 (lambda args (apply (lambda (leapable? $code rte execute-body . other) (set! conts (cons $code conts)) (apply ##step-handler args)) args))) #f #f #f #f #f #f #f))) But I can't manage to extract the continuation out of either |$code|, |rte|, |execute-body| or |other|. How is |$code| being structured? (I'm seeing that $code is being extracted out of cont in several places, so maybe it's not possible to the the continuation this way.) I've also tried: (define (stepper-init) (set! ##main-stepper (vector (make-vector 7 (lambda args (##continuation-capture (lambda (cont) (set! conts (cons cont conts)) (let ((res (apply ##step-handler args))) (step);; attempt against stepping being turned off res))))) #f #f #f #f #f #f #f))) but somehow single-stepping seems to be turned off after a few steps (even if my above code is compiled). I don't get any interesting continuation this way. Could you please guide me a bit? Thanks Christian.

2 2

favicon.ico on gambit web site
by Bradley Lucier 19 Jul '06

19 Jul '06

Marc: Your favicon.ico on the gambit web site (the horse head) is black and white. Unless you think that the white square is part of the logo, it should be black and transparent, so the white part takes the color of the background pattern. (See the icon for the math department at www.math.purdue.edu for what I mean.) Brad

1 0

Have you fixed the code sequence for JUMPGLOSAFE?
by Bradley Lucier 19 Jul '06

19 Jul '06

Marc: I hope you fixed the code for JUMPGLOSAFE and friends as I suggested in http://mailman.iro.umontreal.ca/pipermail/gambit-list/2006-March/ 000662.html Here are the sizes of the macro-expanded .i files for the libraries I use in my homework-on-the-web system: zuse-138% wc all-libs*.i 280652 563890 14846379 all-libs-2.i 280652 692371 17223267 all-libs.i 561304 1256261 32069646 total all-libs.i uses the default ___OPTIMIZE_TIME, all-libs-2.i uses ___OPTIMIZE_SPACE. That's with the default (declare (mostly- fixnum)); it will be larger with (declare (mostly-fixnum-flonun)), which you said you'd set as the default for the next beta. gcc-4.1.1 has been cranking on all-libs.c now for over 40 minutes on my (admittedly lame) ultrasparc, but that's the machine that will run the system, so if I compile it, I need to compile it there. Brad

1 0

Newby question on compiling Gambit output
by Kenneth Kellum 12 Jul '06

12 Jul '06

If this is a FAQ, sorry. I'm trying to use Gambit-C to get an executable version of a Scheme program. I'm comfortable with Scheme, but haven't used C in many years. I'm running Windows XP. I downloaded Gambit-C, version 4.0 beta 9. I've succeeded in using Gambit to translate foo.scm into foo.c. After quite a while looking for documentation I'm still clueless as to how to compile and link foo.c. I've tried both MS Visual Studio and Watcom-1.3. Any help will be greatly appreciated.

3 4

port-name
by Christian 12 Jul '06

12 Jul '06

Hello I've been looking for a way to get the file path of a file port, and have found ##port-name. I would find it worthwhile to add an official port-name function (and document it). (Should I just send a patch for such requests in the future?) Thanks Christian.

1 0

read-line behaviour changed between beta15 and beta17 (MinGW)
by François Magnan 10 Jul '06

10 Jul '06

Hi, When I compile the following program (define (command-loop) (let ((command (read-line))) (serve-command command) (command-loop))) (command-loop) in Gambit-C4.0b15 with the MinGW (win32) platform it works as expected when I execute it in a DOS window (I can enter some lines and they get echoed by the pp). However, when I compile the same program using beta17, only the first line is echoed and after that, I get only an infinite serie of #!eof. This only happens in a DOS window. If I execute the program from a MinGW shell it works as expected. How can I change this behavior? Is this a port-setting problem? Thank you, Francois Magnan

1 1