BUG: infinite loop in _IO_flush_all()

[Eric Ding]

Ack!  Wait, my last e-mail was sent before my brain kicked in.  The
reason I can't repro is because I did a workaround involving using
_exit() rather than exit().  But I can't try to reproduce this until
tomorrow, because we're in the midst of a code freeze.

In the meantime, I checked again -- the system's actually got
libc.so.5.4.38 installed.  The reason that it's using libc.so.5.4.33 is
because there's a libc and libdl installed in the JRE directories that
are in the LD_LIBRARY_PATH when the program is called from elsewhere in
our code.  So it's using those libs.

Eric

>>>>> N8TM@aol.com writes:

> In a message dated 7/29/98 11:58:38 PM Pacific Daylight Time,
> ericding@applix.com writes:

>> I believe the executable is actually using libc.5.4.33
> Not unless you have that version installed.  That's a good library version.
> Why don't you just install it via rpm and be sure of it?