Re: [bigloo] Memory profiling with 4.3a

Hi. "Ollagnon, Pierre-Francois", 2016-12-06: > I am trying to profile memory with bglmemrun but I can't get it work. > Here is my little test: > > test.scm: > --- > (module main > (main main)) > (define (main argv) > (print "Ok")) > > Compilation > --- > bigloo test

Re: [bigloo] Memory profiling with 4.3a

I wrote, 2016-12-07 08:10: > Hi. > > "Ollagnon, Pierre-Francois", 2016-12-06: >> I am trying to profile memory with bglmemrun but I can't get it work. >> Here is my little test: >> >> test.scm: >> --- >> (module main >> (main main)) >> (define (main argv) >> (print "Ok")) >> >>

Re: [bigloo] Memory profiling with 4.3a

Hello Joseph. You wrote, 2016-12-11 00:18: > Ok, after digging a bit more, it looks as though the issue is due to bmem > being linked to libbigloo_u and the profiled program being linked to > libbigloo_s. This results in two version of the runtime being loaded at > once, and unfortunately, on lin

[bigloo] bigloo benchmarks: i386 vs. x86-64 vs. x32

Hi. I have 21 benchmarks results for the current beta version (bigloo4.3a-beta07Dec16.tar.gz), with gcc 6.2.0 from Ubuntu 16.10: i386257016.0 430 760 1750 200 1090 1220 430 1690 870 950 5750 930 920 950 125010 830 1180 8850 1820 AVG x86-64 -17.7 -13.1 -

Re: [bigloo] bigloo benchmarks: i386 vs. x86-64 vs. x32

MS wrote, 2017-01-04: > Hi Sven, > > Happy new year! > >> I have 21 benchmarks results for the current beta version >> (bigloo4.3a-beta07Dec16.tar.gz), with gcc 6.2.0 from Ubuntu 16.10: >> >> i386257016.0 430 760 1750 200 1090 1220 430 1690 870 950 >> 5750 930 920 950

[bigloo] How to detect GC warnings?

Hi all. Is there an easy way to detect GC warnings like GC Warning: Failed to expand heap by ... GC Warning: Out of Memory! Heap size: ... Returning NULL! inside a running bigloo-compiled program? Background: I would like to cleanly shutdown the program and initiate a restart. Ciao Sven

Re: [bigloo] bug report: use srfi1 library - segmentation violation

Hi. "Fours, Dmitri" schrieb am Thu, 9 Mar 2017 23:59:21 +: > I have got the following error when > trying to use the srfi1 library in > the interpreter: > > (module foo (library srfi1)) > > *** ERROR:bigloo: > `segmentation violation' exception -- raised > > Compiled and tested on Linux D

[bigloo] Bug fixes of the reference implementation for SRFI-1

Hi all. The SRFI-1 reference implementation has received bug fixes for list= lset-xor lset-xor! Attached is a patch with theses fixes for srfi1.srfi as contained in bigloo. Ciao Sven --- api/srfi1/src/Llib/srfi1.srfi.orig 2017-05-03 11:03:12.0 +0200 +++ api/srfi1/src/Llib/srfi1.srfi 201

[bigloo] file-size for large files

Hi all. file-size for bigloo in 64-bit mode returns correct numbers for large files: > (file-size "n.db") #e154376286080 In i386 and x32 mode, the results for the same file are incorrect: #e-1 #e-242536576 Ciao Sven

[bigloo] Efficiency of reading cyclic structures (SRFI-38)

Hi Bigloo users. The attached patch speeds up reading large expressions that contain many instances of sharing (via the notation of SRFI-38). Read time for a 40 MB expression (180 MB without sharing) with around 500.000 sharings: before: 5.20 sec after: 0.37 sec (Times are averages measured on an

Re: [bigloo] Efficiency of reading cyclic structures (SRFI-38)

Hi Manuel. MS, 2017-09-23 09:41: > Hi Sven, > >> The attached patch speeds up reading large expressions that contain >> many instances of sharing (via the notation of SRFI-38). >> Read time for a 40 MB expression (180 MB without sharing) with >> around 500.000 sharings: >> before: 5.20 sec >> aft

[bigloo] Profiling bigloo programs with gcc 5 or newer

Hi all. Profiling bigloo compiled programs shows no counters at all if using gcc-5, gcc-6 or gcc-7 from Ubuntu (and probably Debian). The reason is that these gcc versions are configured with: --enable-default-pie You can check your gcc version as follows: # gcc -v |& grep pie The solution is

Re: [bigloo] Dataflow+ debuging

Hello. Pierre-Francois wrote, 2018-03-09: > I have an issue when compiling: The Dataflow+ fail and "remove" a part of the > code: > Unfortunatly I did not success to make a testcase Issue is probably > coming by the fact that several things are coming and used in several modules. > For infor

[bigloo] For a long time, Bigloo does not compile

Hi Eduardo. Eduardo Costa wrote on 2019-04-16 23:35: > I believe that I did not use Bigloo since 2017. The reason: It does not > compile out of the box anymore, and I am not a hacker to analyse the > problem. I tried to install version 4.3e. Here is the last lines of the > make output: > > /Users/

Re: [bigloo] How to compile Bigloo on Debian Linux

Hi Lassi. You wrote, 2019-08-17 19:01: > How does one compile the latest Bigloo releases (4.3e and 4.3f) on Debian? > I get the following output at the end of `./configure && make`. On Mac the > same commands succeed. > > make[1]: Entering directory '/bigloo/runtime' > make[1]: '.afile' is up to d

Re: [bigloo] How to compile Bigloo on Debian Linux

Hi Lassi. You wrote, 2019-08-17 23:43: > Thank you for helping out :) I managed to switch from dash to bash but get > the same error. OK, I am sorry it was not that easy. > I'm trying to build a Debian-based Docker container out of Bigloo; I have no Docker experiences for bigloo; anyone else?

Re: [bigloo] How to compile Bigloo on Debian Linux

Hi Lassi, hi Manuel. You wrote, 2019-08-18 11:59: >> OK, I am sorry it was not that easy. > > No problem. I appreciate the help. If we figure this out, maybe we'll learn > something to make it easier :) > >>> RUN tar -xf bigloo4.3f.tar.gz && mv bigloo4.3f bigloo >> Why do you need the renaming? >>

Re: [bigloo] How to compile Bigloo on Debian Linux

Hi Lasse. >> Manuel, >> can you please fix this in github? > > What is the official Bigloo repo by the way - is it > ? Correct. > We could add that information > to . Yes, you could open a pull request.

Re: [bigloo] How to compile Bigloo on Debian Linux

Hi Lassi. You wrote, 2019-08-18 20:46: > Is also the main issue > tracker used by Bigloo nowadays Yes, see Manuel's announcement on this list: https://sympa.inria.fr/sympa/arc/bigloo/2018-09/msg1.html Ciao Sven

[bigloo] Using valgrind for profiling bigloo-compiled binaries

Hi Bigloo users. I just want to share a positive experience with valgrind (version 3.15) for profiling bigloo-compiled binaries. Here is a step-by-step guide: 1 Compile your binary with debug options (-g2); my binary name is foo. 2 Run your binary; you can stop it when you collected enough data:

Re: [bigloo] Building Bigloo on Ubuntu with JVM support

Julián Herrera wrote, 2020-04-30 21:01: > I ran > > ./configure --jvm=yes > > configure runs fine and spits out a summary that looks accurate. > > I run make, that runs for a couple seconds but then stops at the following: > > /home/jules/Code/bigloo4.3g/bin/bgljfile.sh: 19: exec: > /home/jules/Cod

Re: [bigloo] How do you perceive Bigloo?

Hello. Joseph wrote, 2020-06-19 23:32: > How do you perceive Bigloo? Do you see it as just an implementation of > scheme or as its own language, albeit heavily based on/inspired by scheme? > I ask because, personally, I am increasingly seeing it as a separate > language and am wondering if it woul

[bigloo] futures and parallel forms in Guile and Bigloo

Dear Bigloo users. I read about futures in Guile (see https://www.gnu.org/software/guile/manual/html_node/Futures.html) and the parallel forms built on-top, e.g. parallel, par-map and par-for-each (see https://www.gnu.org/software/guile/manual/html_node/Parallel-Forms.html). They look very promi

Re: [bigloo] futures and parallel forms in Guile and Bigloo

no/hop/blob/master/src/queue_scheduler.scm. > Taking a look at Java's ThreadPoolExecutor abstraction may also be helpful. > > I think this would be an interesting project and would be willing to > contribute to it. > > Best Regards, > Joseph Donaldson. > > On Wednesday,

Re: [bigloo] futures and parallel forms in Guile and Bigloo

Hello Joseph. You wrote, 2020-09-06 19:25: > I took a look at porting scheme-concurrent to Bigloo. The results can be > found at https://github.com/donaldsonjw/bigloo-concurrent. The native > back-end passes all the tests but there are still some issues with the jvm > back-end; I will continue to

Re: [bigloo] futures and parallel forms in Guile and Bigloo

ssues with the jvm > back-end; I will continue to work on it. While doing the port, I made a few > changes to Bigloo. > These are linked from the bigloo-concurrent README. I have made a pull > request to have the changes included in the Bigloo mainline. > > Let me know what you th

Re: [bigloo] futures and parallel forms in Guile and Bigloo

ses all the tests but there are still some issues with the jvm >> back-end; I will continue to work on it. While doing the port, I made a few >> changes to Bigloo. >> These are linked from the bigloo-concurrent README. I have made a pull >> request to have the changes includ

Re: [bigloo] futures and parallel forms in Guile and Bigloo

Best Regards, > Joe > > On Friday, September 25, 2020, 7:54:18 AM PDT, Sven Hartrumpf > wrote: > > Hi again. > > Some more feedback. > > SH, 2020-09-21 19:35: >> Hello Joseph. >> >> You wrote, 2020-09-20 18:43: >>> Hello, Sven, >>&g

Re: [bigloo] futures and parallel forms in Guile and Bigloo

> On Monday, October 12, 2020, 2:45:54 AM PDT, Sven Hartrumpf > wrote: > > Hello Joseph. > > Thanks for your explanations. > I now have a small helper function par-map (parallel map) that does not impose > any thread limit, > while para-map/cores limits it to the num

Re: [bigloo] Preparing 4.4a

Hi. Jerry James wrote, 2020-11-17 09:29: > Building for Fedora turned up two Java-related problems. First, > Fedora now builds with OpenJDK 11 by default, but some bigloo files > don't know about JDKs larger than 9: > api/ssl/src/Makefile > configure > runtime/Makefile > > I can force a build wit

[bigloo] Hyphen in bigloo modules with Java backend?

Hello. Normally, I use the C backend, but for some legacy application I need the Java backend. With newer versions of Java (e.g. 11.0.10), I cannot run my stable code anymore, probably because the hyphen in class names is now a hard error: Exception in thread "main" java.lang.ClassFormatError: I

Re: [bigloo] Hyphen in bigloo modules with Java backend?

Hello Joseph. You wrote, 2021-02-12 01:27: > I am seeing the same issue as you with newer versions of Java. > Although not > ideal, if you had not already noted it, the contents of the modules do not > need to change only the file name. For example, I have a module named > sorted-bag in file name

[bigloo] Building bigloo without pcre2

Hello. I saw that bigloo after 4.4b uses pcre2 by default. As I want/have to avoid heavy dependencies, I tried to add --disable-pcre2 to my standard --disable-pcre in my configure call. But the make step fails as follows: /usr/bin/ld: Cgen/cgen.o: in function `BGl_modulezd2initializa7ationz75zzc

[bigloo] Building bigloo without pcre2

I wrote, 2021-07-14 20:57: > Hello. > > I saw that bigloo after 4.4b uses pcre2 by default. > As I want/have to avoid heavy dependencies, I tried > to add --disable-pcre2 to my standard --disable-pcre in > my configure call. > But the make step fails as follows: > > /usr/bin/ld: Cgen/cgen.o: in fun

Re: [bigloo] Building bigloo without pcre2

manuel.serr...@inria.fr schrieb am 2021-09-06 16:23: > Hi Sven, > >> Hello. >> >> I saw that bigloo after 4.4b uses pcre2 by default. >> As I want/have to avoid heavy dependencies, I tried >> to add --disable-pcre2 to my standard --disable-pcre in >> my configure call. >> But the make step fails as

Re: [bigloo] Building bigloo without pcre2

Hi Manuel. You wrote, 2021-09-07 14:02: > I think I have now fixed the dependency problem. Please let me know if you > are still facing problems. Thanks. You have fixed the problem, indeed. Ciao Sven

Re: [bigloo] Preparing 4.4c

Hi. Building (without Java backend) works for x86-64, but i586 and x32 are broken at the same position: gcc -I/var/tmp/bigloo/lib/bigloo/4.4c -O2 -fcf-protection=none -fno-stack-protector -fno-strict-aliasing -msse2 -mfpmath=sse -Wno-unused-value -Wno-parentheses-equality -Wno-parentheses -Wno

Re: [bigloo] 4.4c, second endeavor

Manuel wrote, 2021-11-26 15:35: > Thanks to those who have sent early reports. I have fixed the JVM and > Bignum problems The bignum fix works, but the build process stops later on (tested only for x32): gccx32 -I/var/tmp/bigloo/lib/bigloo/4.4c -O2 -fcf-protection=none -fno-stack-protector -fno

[bigloo] xml-parse: file position of elements

Hi. I am using bigloo's xml-parse. It is really nice (but undocumented?). Does anybody know a way to determine the file offset of an xml element in the result? An option would be really powerful. Greetings Sven

Re: [bigloo] xml-parse: file position of elements

Hello Joe. Joseph Donaldson wrote 2022-01-30 02:37: > The current implementation of xml-parse does not track the file offset of > the xml elements it parses. It may be possible to modify it do so. Can you > elaborate a bit further on your exact use case? I need to propose patches for quite comple

Re: [bigloo] xml-parse: file position of elements

Hello Joe. Joseph Donaldson, 2022-02-01 18:21: > Hello, Sven, > Here is an updated version that moves the start-position parameter to > position 4 and has the value of the element's starting < character. > I also removed some debugging print statements that were erroneously left in > place. Let

Re: [bigloo] xml-parse: file position of elements

Hello Joe. donaldsonjw, 2022-02-06 01:00: > Here is another version of xml.scm that I believe addresses the problem you > pointed out. Let me know how it works for you. Thanks again. I had to fix two calls of 'make' by moving the position to the end, see attached file, then everyting worked lik

Re: [bigloo] xml-parse: file position of elements

t Regards, > Joe Donaldson > > On Thursday, February 10, 2022, 09:42:43 AM PST, Sven Hartrumpf > wrote: > > Hello Joe. > > donaldsonjw, 2022-02-06 01:00: >> Here is another version of xml.scm that I believe addresses the problem you >> pointed out. Let me kno

[bigloo] bigloo-4.5a, release time

Hi Manuel. Thanks for preparing a new release. srfi27 tests report 4 errors: --- make fulltest ... make[3]: Leaving directory '/var/tmp/bigloo/api/srfi27/recette' cond-expand...ok. srfi27.large-numbers...*** ERROR:srfi27: (random-integer n) returned illegal value -- -2277703446941 ok. srfi27.rea

[bigloo] new warnings in 4.5a: bint expected, bignum provided

Hi. I am seeing new (spurious?) warnings for the current unstable version (x86_64 and x32-ABI): "bint" expected, "bignum" provided A minimal example: bugz11.scm: --- (module bugz11 (export (list-ref-check::obj l pos::bint)) (main main)) (define (list-ref-check l pos) (let iter ()

Re: [bigloo] new warnings in 4.5a: bint expected, bignum provided

Hi Manuel. Manuel wrote, 2022-11-03 11:50: > Hi Sven, > >> Hi. >> >> I am seeing new (spurious?) warnings for the current unstable version >> (x86_64 and x32-ABI): >> >> "bint" expected, "bignum" provided >> >> A minimal example: >> >> bugz11.scm: >> --- >> (module bugz11 >> (export >>(list

[bigloo] -freturn -freturn-goto

Dear bigloo users. If I compile programs (normal C backend, gcc 10.5.0) with -O6 and with -O6 -freturn -freturn-goto, the resulting binaries are byte-identical. Do these options require other options to have any effect? Greetings Sven

[bigloo] Re: warnings from gcc for crc.c

Hi Manuel. Tue, 23 Jul 2013 16:01:34 +0200, Manuel.Serrano wrote: >> gcc -I/var/tmp/bigloo4.0b/lib/4.0b -Ofast -ffast-math -fno-stack-protector >> -fno-strict-aliasing -fomit-frame-pointer -march=corei7 >> -mpreferred-stack-boundary=2 -pipe -fPIC >> -I/var/tmp/bigloo4.0b/gc/gc-7.3alpha3-201303

[bigloo] INTERNAL-ERROR from bigloo when compiling

Hi Manuel. I have seen several source files that cause the bigloo compiler to crash like this: *** INTERNAL-ERROR: Illegal signal caught --- aborting...11 (C and java backend; seems to occur only if one uses bigloo optimization -O2 or higher.) As it is not easy to isolate, I want to ask the foll

Re: [bigloo] bigloo compile error

MM: >> Is there any minimal version of clang that is required to >> compile bigloo? >> >> I'm asking because the following lines exist in the package for >> MacPorts and I'm curious if clang < 3.3 would still fail to work with >> the latest version and if -fbracket-depth=4096 still needs to be add

[bigloo] SIGPIPE crash when writing to closed socket port

Hi all. If a bigloo program writes to a socket port that has already been closed, the program crashes with a SIGPIPE signal and a backtrace in gdb like this one: (gdb) bt #0 0x5430 in __kernel_vsyscall () #1 0x086a4a12 in __write_nocancel () at ../sysdeps/unix/syscall-template.S:81 #2 0x08

Re: [bigloo] SIGPIPE crash when writing to closed socket port

Hi. MS: > SIGPIPE is a signal, not an error code. Hence, you should rather install > a signal handler for catching it. Try something as: > > (signal sigpipe (lambda (n) (print "GOT IT"))) This works perfectly. Thanks for your help, Manuel! Ciao Sven

Re: [bigloo] typo in new configure option

SH: >> How about switching to gc 7.4.0? MS: > Done. Great. > I will appreciate if you could give it a try... Works fine. (Manuel, I send you some unrelated GCC warnings in a separate mail.) Ciao Sven

[bigloo] Allowing gzip encoding in http requests via bigloo's http

Hi all. Is there a way to allow a compressing encoding in bigloo's 'http' function? I.e. a keyword option like :encoding 'gzip ? Given that bigloo has built-in functions for decompressing gzip output this should be easy to implement. Ciao Sven

Re: [bigloo] Closing Sockets

Hi. Thu, 3 Apr 2014 00:21:30 -0300, fernando.ribeiro wrote: > From the docs, socket-shutdown should call socket-close when the close > parameter is #t, but I am having issues with the following code: > > (define socket (make-client-socket "localhost" 61613)) > > (socket-shutdown socket) ; doesn'

Re: [bigloo] problem to build bigloo-lib on a Mac

Hi Guilhem. Sat, 21 Jun 2014 10:53:21 +0200, gdw wrote: > I have built and installed the latest bigloo 4.1a using the default settings > (/usr/local/) and I am trying to install bigloo-lib 0.21b. This version of bigloo-lib is 7 years older than bigloo - so expect trouble :-) ... > checking the

Re: [bigloo] bdb help

Fri, 18 Jul 2014 09:55:24 +0200, gdw wrote: > I am still unable to use bdb: when I compile a scheme file, the compiler says ... > *** WARNING:bigloo:-gbdb > Bdb not available (see Bigloo configuration option) ignoring option bdb is part of bee, hence the following line from configure --help is re

[bigloo] Reset counters when profiling

Hi all. Is there a bigloo function to reset all counters (to 0) when a program is run in profile mode? Background: I need profiling results, but only after my program reached a specific point and until it terminates. Ciao Sven

[bigloo] latest alpha needs new bglstone?

Hi Manuel. Thanks for the new alpha version 2014-10-13. Now I can use the 32bit build as needed (using ./configure --disable-libuv). As a side note: I think you must update bglstone now due to the rgc changes: ... make[4]: Entering directory '/var/tmp/bglstone/src/cgc/bigloo' bglafile -o .afile

Re: [bigloo] bee install problem

Hi Raoul. You wrote on 2014-08-11: > ubuntu 14.04 lts > Linux superlap2 3.13.0-30-generic #55-Ubuntu SMP Fri Jul 4 21:43:42 > UTC 2014 i686 i686 i686 GNU/Linux > GNU Emacs 24.3.1 ... > ude/ude-compile.el:18:1:Error: Symbol's value as variable is void: > image-load-path That's really strange. I ha

[bigloo] BWD GC 7.4.2

Hi Manuel. I just browsed the ChangeLog file for the BWD GC 7.4.2. Looks like a definite improvement. What do you think? Merry Christmas! Sven

Re: [bigloo] Bigloo Regexp: m/ /ms (now: Regular Parsing)

Hi Laurent. Laurent Bloch wrote: > Here is my try on it. Besides the ugly copying operations, it seems to me > that the multiline option doesn't make what I was waiting for: finding a > pattern across a newline. Your raw-sequence-file.txt does not contain any newlines, just some ASCII codes 0. Pl

[bigloo] Unicode escape with \u in json-read

Hi Manuel. I stumbled across this comment in json-read from file ./api/packrat/src/Llib/json.scm : ;; we don't support the "u" escape for unicode Do you plan to fix this? Or should one preprocess json data (converting the \u0123 notation to UTF-8, or similar) before calling json-read? Ciao Sve

Re: [bigloo] Unicode escape with \u in json-read

SH: >> I stumbled across this comment >> in json-read from file ./api/packrat/src/Llib/json.scm : >> >> ;; we don't support the "u" escape for unicode >> >> Do you plan to fix this? MS: > No, I don't as this is already supported by the JSON-PARSE function of the > Bigloo web library (see api/we

[bigloo] utf8-substring vs. subucs2-string

Hi Manuel. The ucs2 version seems to be named inconsistently. Should not this be ucs2-substring ? Ciao Sven

[bigloo] sha256sum broken on 64-bit systems

Hi all. I would like to warn other users that sha256sum seems to be broken on 64-bit (ok on 32-bit). Here are some examples (roughly 0.1% of random strings affected): s: input string next line: bigloo output (sha256sum-string) following line: correct output (output by bigloo32 or md5sum binary):

Re: [bigloo] sha256sum broken on 64-bit systems

Hi Manuel. >> > I'm currently preparing the new tarball. >> >> Please let me know when it is ready. > Oops. I have forgotten to upload it yesterday. It is now available (4.3a). Thanks! I have run some tests with millions of random strings and the 64bit version of sha256sum... now works as correc

Re: [bigloo] Do function fail in interpreted mode

Hi Pierre-Francois. You wrote, 2016-04-25 16:49: > Hello Manuel, > > I am using the latest 4.3c There is no such beast. 4.2c or 4.3a (alpha)? > (I am switching from an old 3.1b ...) Welcome back - this switching is a worthwhile project, but expect some trouble ahead :-) > and I face some issu

Re: [bigloo] Fail to compile bigloo with

Hi Pierre-Francois. You wrote: > I have some files failing when using these optimisations: > cfa-tracking (at execution) > dataflow-types (at execution) > initflow (I think at execution) > > Is it "normal" ? I hope not :-) All three optimizations are already turned on at a low optimization leve

Re: [bigloo] Performance

Hi Bob. bobeob wrote on 2016-08-12 05:11: > Hi, I'm just wondering how I might speed up my code with little or no > effort! > Do class fields marked as read-only improve performance? I have never used classes, so I have no idea:-) > Might I get faster code by merging all my modules into one big