This is the mail archive of the java-patches@gcc.gnu.org mailing list for the Java project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Patch: core protocol take 2


"Anthony Green" <green@redhat.com> writes:

> I just checked in my changes without fixing this.  I intend to rewrite this
> code very soon anyways so that we have per classloader resource files (once
> Per's changes are in, if they aren't already).

The patch http://gcc.gnu.org/ml/java-patches/2001-q3/msg00366.html
("SharedLibLoader patch") is not checked in.  I'm unsure how best to
conditionally compile the dependency on dlopen. Some feedback would
be welcome.  Using libltdl has the problem that it always sets
RTLD_GLOBAL.  One queston is whether the libtool people would be
interested in extending the API to handle a variant of lt_dlopen
that does *not* pass RTLD_GLOBAL.  If so there are two basic options:
A global flag (perhaps set with a function); or new functions.
In the latter case, we can create sibling functions that are
like ld_dlopen/lt_dlopenext not not passing RTLD_GLOBAL, or we can
create a generalized function that takes an extra 'flags' argument.
The latter is probably the best solution, but it's a fair amount
of work.

Alternatively, I can use plain dlopen, which I think requires extra
autoconf testing, but should otherwise be straightforward,  If someone
can suggest the approved autconf patch, that would help a lot.
-- 
	--Per Bothner
per@bothner.com   http://www.bothner.com/per/


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]