[PARI] about the behavior of install()

[Philippe Elbaz-Vincent <pev@pev.math.univ-montp2.fr>]

Dear PARI Experts,

I am puzzled by the behavior of the function install(), or should I say
its internal mechanism, and I will appreciate any help on this matter.
I will illustrate my question with the matexp function given in the
examples/ directory.
all the following was done on a i686 running linux (ix86 kernel) 32-bit
version, with 2.0.21 (beta) [glibc-2.1.3, gcc-2.95.2], but this behavior was observed
 with  previous CVS 
versions of PARI (also tested against egcs-1.1.2):

I open a gp session, and install the function 'mexp' as described
(00:40) gp > 
install(matexp,Gp,mexp,"~pev/PARI-Dev/pari/examples/libmatexp.so")
time = 0 ms.

I test it
(00:40) gp > mexp([0,1;1,0])
time = 0 ms.
%2 = 
[1.543080634815243778477905620 1.175201193643801456882381850]

[1.175201193643801456882381850 1.543080634815243778477905620]

so far so good. Now, say I'm unhappy with the behavior of the function, I
decide to "change it" by adding a kind message like a printf("hello !\n");
Then, I do a 'make clean' and a 'make all' in the dir examples/
In my gp session, I kill the function 'mexp', and install it again,
expecting my "new" function

(00:40) gp > kill(mexp)
time = 0 ms.
(00:40) gp >
install(matexp,Gp,mexp,"~pev/PARI-Dev/pari/examples/libmatexp.so")
time = 0 ms.

and I get exactly the same output than above, nothing has changed !
But, if I quit my gp session,  re-open it, and install again 'mexp', I
get;
(00:41) gp >
install(matexp,Gp,mexp,"~pev/PARI-Dev/pari/examples/libmatexp.so")
time = 0 ms.
(00:41) gp > mexp([0,1;1,0])
hello !
time = 0 ms.
%1 = 
[1.543080634815243778477905620 1.175201193643801456882381850]

[1.175201193643801456882381850 1.543080634815243778477905620]

which is the new 'mexp'...
So my question is; does this illustrate the expected behavior of install()
? If yes, what is the correct way to kill and re-install a loadable
function ? (because it's quite cumbersome when you're testing/debugging a
function)

Thanking you in advance.

best regards,Ph.

detail of the config (if useful); Linux-2.2.17+ReiserFS patch+I2C-2.5.2,
p2/450,glibc-2.1.3, readline v4.0 enabled, gcc-2.95.2, 
compile flag: -O3 -DGCC_INLINE -Wall -Wno-implicit 
-fexpensive-optimizations   -fomit-frame-pointer

Laboratoire Geometrie-Topologie et Algebre. UMR CNRS 5030.
E-mail: pev@pev.math.univ-montp2.fr | Phone: +33 (0)467143958 
http://www.math.univ-montp2.fr/~pev | Fax: +33 (0)467143558