-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi there,
I apologize if this hits the list twice. I sent it before I was
subscribed and it appears to me that it was silently discarded. This
is a verbatim repost except for the information about the kernel
version:
I am trying to use the jdk117_v3 on a glibc 2.1.2 system under a 2.2.9
single processor kernel. I am using native threads and the Invocation
API embedded in a large application. My problem is that everytime the
garbage collector runs, most of the threads seem to get SIGSTOP but
never get started again. The program even detaches from the shell as
if I had hit Ctrl-Z.
I can repeat this behavior at will by pressing a button in my
application that forces a call to Runtime.getRuntime().gc(), but it
also happens whenever the gc gets called upon allocation failure.
I've tried to repeat this behavior in a smaller test program, but I
cannot. The program it happens in is very large, but it does no
threading or signal handling of its own.
Is it possible that the code that sends SIGSTOP upon garbage
collection is sending it to itself? I've tried using strace to
determine this, but the behavior doesn't happen during strace because
SIGSTOP is blocked.
Are there any ideas on how I can track this down? Thanks.
- --
Gene McCulley Voice: (407) 265-0772
[EMAIL PROTECTED] Fax: (407) 265-0773
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v0.9.10 (GNU/Linux)
Comment: For info see http://www.gnupg.org
iD8DBQE31m/WK+LjkZ54DewRArc6AKCLUJGk6rZkYNYxWhOJLnuHz6Y3aACgh7xl
lm8WYTKo74PHWyqcLCQQCZ8=
=whdI
-----END PGP SIGNATURE-----
----------------------------------------------------------------------
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
