Gambit-list Septembre 2013

gambit-list@iro.umontreal.ca

24 participants
29 discussions

How to configure Gambit with maximum debugging?
by Bradley Lucier 29 Sep '13

29 Sep '13

I should know this. I want to configure and build gambit with maximum,maximum,maximum debugging info. For example, when I add -track-scheme -debug -debug-location -debug-source -debug-environments to the gsc-boot line for converting .scm to .c files in lib/makefile, and I hit an error called on the routine at the bottom, I see 1> ,y 0 naive-div "lib/_num.scm"@9184.3 (apply error 'crap2 (map unveil-bignum (list normalizing-bit... 1> ,e n == 2 normalizing-bit-shift == 5 u2 == 2097152 v2 == 3200000000 but there are also other variables---u, v, iota, etc.---that are in the routine and don't show up in the environment. Is it possible to get these? I'm trying to debug a problem where ... well, you might not believe me, but it seems to be nasty. Brad #<procedure #2 naive-div> = (##lambda (u v) (define (iota n) (if (##fx= n 0) '() (cons (fx- n 1) (iota (fx- n 1))))) (or (and (##bignum? u) (##bignum? v)) (error 'crap u v)) '(pp 'naive-div-orig) '(pp (cons '##bignum.mdigit-length (map ##bignum.mdigit-length (list u v)))) '(pp (cons '##bignum.adigit-length (map ##bignum.adigit-length (list u v)))) '(pp (map unveil-bignum (list u v))) (let ((n (let loop1 ((i (##fixnum.- (##bignum.mdigit-length v) 1))) (if (##fixnum.< 0 (##bignum.mdigit-ref v i)) (##fixnum.+ i 1) (loop1 (##fixnum.- i 1)))))) (let ((normalizing-bit-shift (##fixnum.- ##bignum.mdigit-width (##integer-length (##bignum.mdigit-ref v (##fixnum.- n 1)))))) (let ((u2 (##bignum.arithmetic-shift u normalizing-bit-shift)) (v2 (##bignum.arithmetic-shift v normalizing-bit-shift))) (or (and (##bignum? u2) (##bignum? v2)) (let () '(pp '(list normalizing-bit-shift u v u2 v2)) (apply error 'crap2 (map unveil-bignum (list normalizing-bit-shift u v u2 v2))))) (let* ((u u2) (v v2) (q (##bignum.make (##fixnum.+ (##fixnum.- (##bignum.adigit-length u) (##bignum.adigit-length v)) 2) #f #f)) (v_n-1 (##bignum.mdigit-ref v (##fixnum.- n 1))) (v_n-2 (##bignum.mdigit-ref v (##fixnum.- n 2)))) (let ((m (let loop2 ((i (##fixnum.- (##bignum.mdigit-length u) 1))) (let ((u_i (##bignum.mdigit-ref u i))) (if (##fixnum.< 0 u_i) (if (##not (##fixnum.< u_i v_n-1)) (##fixnum.- (##fixnum.+ i 1) n) (##fixnum.- i n)) (loop2 (##fixnum.- i 1))))))) (let loop3 ((j m)) (if (##not (##fixnum.< j 0)) (let ((q-hat (let ((q-hat (##bignum.mdigit-quotient u (##fixnum.+ j n) v_n-1)) (u_n+j-2 (##bignum.mdigit-ref u (##fixnum.+ (##fixnum.- j 2) n)))) (let ((r-hat (##bignum.mdigit-remainder u (##fixnum.+ j n) v_n-1 q-hat))) (if (or (##fixnum.= q-hat ##bignum.mdigit-base) (##bignum.mdigit-test? q-hat v_n-2 r-hat u_n+j-2)) (let ((q-hat (##fixnum.- q-hat 1)) (r-hat (##fixnum.+ r-hat v_n-1))) (if (and (##fixnum.< r-hat ##bignum.mdigit-base) (or (##fixnum.= q-hat ##bignum.mdigit-base) (##bignum.mdigit-test? q-hat v_n-2 r-hat u_n+j-2))) (##fixnum.- q-hat 1) q-hat)) q-hat))))) (##declare (not interrupts-enabled)) (let loop4 ((i j) (k 0) (borrow 0)) (if (##fixnum.< k n) (loop4 (##fixnum.+ i 2) (##fixnum.+ k 2) (##bignum.mdigit-div! u (##fixnum.+ i 1) v (##fixnum.+ k 1) q-hat (##bignum.mdigit-div! u i v k q-hat borrow))) (let ((borrow (if (##fixnum.< n k) borrow (##bignum.mdigit-div! u i v k q-hat borrow)))) (if (##fixnum.< borrow 0) (let loop5 ((i j) (l 0) (carry 0)) (if (##fixnum.< n l) (begin (##bignum.mdigit-set! q j (##fixnum.- q-hat 1)) (loop3 (##fixnum.- j 1))) (loop5 (##fixnum.+ i 1) (##fixnum.+ l 1) (##bignum.mdigit-mul! u i v l 1 carry)))) (begin (##bignum.mdigit-set! q j q-hat) (loop3 (##fixnum.- j 1)))))))) (##cons (##bignum.normalize! q) (##arithmetic-shift (##bignum.normalize! u) (##fixnum.- normalizing-bit-shift)))))))))))

2 1

News from ICFP
by Marc Feeley 28 Sep '13

28 Sep '13

Today, Chris Petersen gave a very nice presentation at ICFP about the development of LambdaNative by the Pediatric Anesthesia Research Team at UBC. It was nice to see a very concrete application of Gambit to health care. The session chair said that a memorable quote from this year's conference is "Scheme saves lives". In addition to this, it seems that the Pulse Oximeter application developed by the team has been selected by the United Nations as one of the 10 innovations for 2015 to achieve the Millennium Development Goals to decrease maternal and child mortality (http://www.path.org/innovations2015/). Congratulations to the Pediatric Anesthesia Research Team at UBC! Marc

3 2

Help with a macro expansion.
by Bradley Lucier 28 Sep '13

28 Sep '13

Marc: I'm trying to write my own tracing routine for my new naive-div routine, and it's a little tricky. I wrote the following macro to (1) use naive-div-new for naive-div, (2) collect the args and results of all calls to naive-div-new when computing an expression, (3) resetting naive-div to naive-div-orig, and (4) pretty-printing some of the results. (define-macro (while-tracing-naive-div-new expression) (let ((result (gensym)) (args+result (gensym))) `(let ((,args+result '())) (set! naive-div (lambda (u v) (let ((result (naive-div-new u v))) (set! ,args+result (cons (list u v result) ,args+result)) result))) ;; calculate expression with naive-div-new (let ((result ,expression)) ;; set naive-div back to naive-div-orig before calling any more number-manipulating routines (set! naive-div naive-div-orig) (pp (map (lambda (args+result) (append args+result (list (naive-div-orig (car args+result) (cadr args+result))))) ,args+result)) ;; supposed to return result, but returns #!void ???????????????? result)))) I use it in the following way: (while-tracing-naive-div-new (string->number (number->string xx))) which expression is expanded by gsc -expansion -c test-naive-div.scm > ! crap.scm to (let ((#:g4 ('#<procedure #24 ##box> '()))) (let ((begin-temp.27 (set! naive-div (lambda (u v) (let ((result (naive-div-new u v))) (let ((begin-temp.24 ('#<procedure #31 ##set-box!> #:g4 ('#<procedure #5 cons> ('#<procedure #6 list> u v result) ('#<procedure #25 ##unbox> #:g4))))) result)))))) (let ((result ('#<procedure #28 string->number> ('#<procedure #29 number->string> xx)))) (let ((begin-temp.26 (set! naive-div naive-div-orig))) (let ((begin-temp.25 (pp ('#<procedure #32 map> (lambda (args+result) ('#<procedure #33 append> args+result ('#<procedure #6 list> (naive-div-orig ('#<procedure #9 car> args+result) ('#<procedure #34 cadr> args+result))))) ('#<procedure #25 ##unbox> #:g4))))) #!void))))) So where did the #!void at the end come from? Brad

2 5

Compiling programs that use syntax-case macros
by Panicz Maciej Godek 26 Sep '13

26 Sep '13

Hi, is there any way to use the Gambit compiler to compile programs that use syntax-case macros? I see that one needs to (load "~~lib/syntax-case.scm") in order to use them, but that only seems to work with the interpreter. Thanks in advance, M.

3 8

Returning a byte array from C
by Vijay Mathew 26 Sep '13

26 Sep '13

The following C function is a wrapper for the `read' system call. It returns a pair that contains the number of bytes read and the data. Is this the right way to initialize a u8vector from C? ___SCMOBJ read_bytes (int fd, size_t count) { void *data = malloc (count); ___SCMOBJ result; if (data == NULL) result = ___make_pair (___BOOLEAN (0), ___BOOLEAN (0), ___STILL); else { int r = read (fd, data, count); ___SCMOBJ retcode; ___SCMOBJ ___err; ___BEGIN_SFUN_INT_TO_SCMOBJ (r, retcode, 1); ___END_SFUN_INT_TO_SCMOBJ (r, retcode, 1); if (r == 0) result = ___make_pair (retcode, ___BOOLEAN (0), ___STILL); else if (r > 0) { ___SCMOBJ data_vect; data_vect = ___alloc_scmobj(___sU8VECTOR, count, ___STILL); ___EXT (___release_scmobj) (data_vect); memcpy (___BODY (data_vect), data, r); result = ___make_pair (retcode, data_vect, ___STILL); } else result = ___make_pair (retcode, ___BOOLEAN (0), ___STILL); free (data); ___release_scmobj (result); } return result; } Thanks, --Vijay

4 10

debian packaging
by Jackson Doak 25 Sep '13

25 Sep '13

I'm planning to adopt gambit-c in debian, but the previous maintainer has been gone for many years now. I've got some questions 1. How has the copyright changed since version 4.2.8 2. How have the dependencies changed since 4.2.8 3. Have any of the existing debian patches been fixed? Once i get this packaged, there should be an increase in users. Jackson

2 1

Patch to correct treatment of exact 0
by Bradley Lucier 24 Sep '13

24 Sep '13

Marc: The numerical library currently returns > (expt 0 1.) 0 > (/ 0 1.) 0 > (/ 0 1.+1.i) 0 This small patch has it return > (expt 0 1.) 0. > (/ 0 1.) 0. > (/ 0 1.+1.i) 0.+0.i which is what I believe R6RS does. Gambit now passes most of the arithmetic and trigonometric and transcendental function tests in the "Test Programs" part of http://www.cesura17.net/~will/Professional/Standards/Scheme/R6RS/ The remaining differences are accuracy issues. I'll need to recode the complex inverse hyperbolic and inverse trigonometric routines as in Kahan, W. "Branch Cuts for Complex Elementary Functions, or Much Ado About Nothing's Sign Bit". Brad

2 1

Gambit JS backend
by Alexander Shendi 24 Sep '13

24 Sep '13

Dear List, I am trying to use the Gambit-Scheme JS-backend. I am using the 4.7.0 release. I compiled the following program (obtained from this list) with: gsc -c -target js fib.scm ;; ===================== fib.scm =============================== (declare (standard-bindings) (extended-bindings) (not safe) (fixnum) (block)) (define (fib n) (define (fib n) (if (< n 2) n (+ (fib (- n 1)) (fib (- n 2))))) (fib n)) (println (fib 30)) ;; EOF I tried to run it: node fib.js /home/alexshendi/sources/gambit/fib.js:42 print(obj); ^ ReferenceError: print is not defined at Gambit_println (/home/alexshendi/sources/gambit/fib.js:42:3) at Object.Gambit_bb1_println [as println] (/home/alexshendi/sources/gambit/fib.js:66:3) at Gambit_bb4__20_fib (/home/alexshendi/sources/gambit/fib.js:119:31) at Gambit_bb3__20_fib (/home/alexshendi/sources/gambit/fib.js:104:12) at Gambit_trampoline (/home/alexshendi/sources/gambit/fib.js:21:10) at Object.<anonymous> (/home/alexshendi/sources/gambit/fib.js:252:1) at Module._compile (module.js:456:26) at Object.Module._extensions..js (module.js:474:10) at Module.load (module.js:356:32) at Function.Module._load (module.js:312:12) Now I have the following questions: * Apparently the generated code is for another JS implementation (Spidermonkey?)? * How do I access e.g. console.log() from scheme? * I tried to use the standard Scheme functions "display" and "newline" from fib.scm, but that didn't work. How do I use them? * Where do I find more info on the JS backend? Many thanks in advance for your help. Best Regrads, Alexander

2 1

Operator is not a PROCEDURE, unbound
by Sven Hartrumpf 22 Sep '13

22 Sep '13

Hi. I have a large program that runs fine in the interpreter. But when compiled with these options (latest gambit from git) (declare (r5rs-scheme) (constant-fold) (inline) (inlining-limit 500) (standard-bindings) (extended-bindings) (safe) (mostly-generic)) the compiled program crashes as follows: *** ERROR IN | n| -- Operator is not a PROCEDURE (#!unbound '#(syntax-object change-directory ((top) #(ribcage #(change-directory) #((top)) #(change-directory)))) '(global . change-directory) '*top*) change-directory is from a simple Scheme portability layer: (define (change-directory path) (current-directory path)) Sven

3 4

Generating symbols from C code
by Vijay Mathew 18 Sep '13

18 Sep '13

What is the correct way to generate symbols in C code? What is the difference between DEF_SYM1 and DEF_SYM2 macros? Thanks, --Vijay

2 1

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Gambit-list Septembre 2013