Can you pass a dump of what C compiler& linker invocations are being made?
You do this by ensuring that in Gambit's OS environment there's the var GAMBC_CC_VERBOSE set to true, so in bash,
GAMBC_CC_VERBOSE=TRUE gsc
(The verbosity stuff is effectuated by the gambc-cc shellscript.)
Den 13 februari 2012 22:00 skrev Álvaro Castro-Castilla < alvaro.castro.castilla@gmail.com>:
hum... I still get the same *** WARNING -- Could not find C function: "____20_xml_2e_o1" *** ERROR IN (console)@1.1 -- /home/alvatar/gambit-xml/xml.o1: undefined symbol: ___release_scmobj
I've surrounded all those calls with ___EXT()
2012/2/13 Marc Feeley feeley@iro.umontreal.ca
Can you try the change I suggest to see if the problem goes away?
Marc
On 2012-02-13, at 5:59 AM, Álvaro Castro-Castilla wrote:
Marc, if you mean my error, it was on Linux 3.1 (Gentoo), and happens
also on latest Ubuntu on VirtualBox.
2012/2/13 Marc Feeley feeley@iro.umontreal.ca
On 2012-02-12, at 8:43 AM, Álvaro Castro-Castilla wrote:
Hi Taylor,
Do you plan on making full bindings for libxml2? I'd be interested in
packaging it as a Blackhole module when you are ready.
Also, when following your steps I got this error:
(load "xml")
*** WARNING -- Could not find C function: "____20_xml_2e_o1" *** ERROR IN (console)@1.1 -- /home/alvatar/gambit-xml/xml.o1:
undefined symbol: ___release_scmobj
Using gambit 4.6.3
Best regards,
Álvaro
I'm not sure this is the cause of the bug, but I have noticed that
Taylor's code is not completely portable. It contains references to Gambit runtime functions that are not wrapped in a call to ___EXT. In other words, the calls
vect = ___make_vector(4, ___FAL, ___STILL); ___release_scmobj(x); tmp = ___make_pair(x, lst, ___STILL);
should be written
vect = ___EXT(___make_vector)(4, ___FAL, ___STILL); ___EXT(___release_scmobj)(x); tmp = ___EXT(___make_pair)(x, lst, ___STILL);
This is needed on some operating systems to link with the Gambit
runtime system functions. On the operating systems that don't need anything special, the call ___EXT(f) is replaced with f, so there is no overhead.
By the way, on which OS did this problem occur?
Marc
Gambit-list mailing list Gambit-list@iro.umontreal.ca https://webmail.iro.umontreal.ca/mailman/listinfo/gambit-list