Hey!

Thanks for the tip.

What made you think that this command does anything useful? 
well I was fumbling around and I thought make files should yield something. Turns out it does something else but I wanted to know why it running "make -f Makefile.devel" gave me an error? 

> ISTR that you also expressed interest in libffcall - you might want to
> start with building and installing it.
I was thinking of sticking to clisp itself, but I will definitely look into it. Thanks ;)

>>    3.   readline:   no (user requested: default)
> what is your platform that it lacks readline?
I run Ubuntu 13.04 on a Intel i5 machine though I don't know why the program said that my system lacks readline. 

"make check" either prints "tests passed" or instructions on where to
> find the failures (see tests/*.erg)

I carefully inspected each step this time while running "make check". And there was just this one single error
in the "src/tests/backquot.erg" file.  I suspect if it is an error I should expect.





On Wed, Mar 12, 2014 at 2:01 AM, Sam Steingold <sds@gnu.org> wrote:
Hi,

> * shashank <funfunaxtebil@tznvy.pbz> [2014-03-12 01:20:56 +0500]:
>
> I am Shashank (a gsoc aspirant) and I have been trying to build clisp
> however I can't seem to progress.
> I have cloned the mercuial repository from here -
> http://sourceforge.net/p/clisp/clisp/ci/default/tree/doc/
>
> In an earlier exchange of email, Sam Steingold asked me to:
>> Clone the clisp mercurial repo and build from it with threads.
>> Observe the crashes in the tests.
>> Investigate how you would replace the CLISP hash table implementation
>> with the lock-free one.
>
> However as soon as I do :
>
>
>    1. $ make -f Makefile.devel CFLAGS=MULTITHREAD
>
>
> I get this error :     (The complete stacktrace is
> here<http://pastebin.com/PpHsT1vU>
> )
>
>
>    1. find: `@srcdir@\\..\\modules\\': No such file or directory

What made you think that this command does anything useful? :-)

> So I tried build using configure
>
>    1. $ ./configure --with-threads=POSIX_THREADS
>
> the command runs successfully and in the end, I get the following
>
>
>    1. Configure findings:
>    2.   FFI:        no (user requested: default)

ISTR that you also expressed interest in libffcall - you might want to
start with building and installing it.

>    3.   readline:   no (user requested: default)

what is your platform that it lacks readline?

>    4. libsigsegv: yes
>    5. ./makemake  --with-threads=POSIX_THREADS  > Makefile
>    6. make: `config.lisp` is up to date.
>    7.
>    8. To continue building CLISP, the following commands are recommended
>    9.   (cf. unix/INSTALL step 4 ff):
>    10.     cd src
>    11.     vi config.lisp
>    12.     make
>    13.     make check

you could also wrap the whole thing in one command, by passing "--cbc"
(config/build/check) to configure.

> However, I can't possibly seem to locate errors while building with
> threads, although I may be wrong in execution or looking at the wrong

"make check" either prints "tests passed" or instructions on where to
find the failures (see tests/*.erg)

> place.  Can you guide me as to how I shall proceed in this
> regard. Also, how and where the CLISP hash table implementation is
> used, which in turn should be replaced by a lock-free implementation.

See clisp/src/hashtbl.d for the hash table implementation.

--
Sam Steingold (http://sds.podval.org/) on darwin Ns 10.3.1265
http://www.childpsy.net/ http://pmw.org.il http://openvotingconsortium.org
http://jihadwatch.org http://islamexposedonline.com
Feynman: 'Philosophy of science is as useful to scientists as ornithology is to birds'