1.7.0 CVS file append & NULL overlapped hEvent

[Brian Ford <Brian dot Ford at FlightSafety dot com>]

While trying to test Corinna's fix for my heap error problem here:

http://cygwin.com/ml/cygwin/2007-07/msg00507.html

(which I believe to be fixed, but I can not yet firmly confirm it since I
have not tried the fix isolated from current CVS head as requested, and
CVS head has these other problems) I have come across the following
regressions.

$ uname -a
CYGWIN_NT-5.1 PC1163-8460-XP 1.7.0(0.177/4/2) 2007-07-20 14:35 i686
unknown unknown Cygwin

The first one relates to file appending via bash shell redirection with a
Windows native program providing the output:

$ cat Test.java
public class Test {
    static public void main(String[] argv) {
        System.out.println("message1 to stdout");
        System.err.println("message1 to stderr");
        System.err.println("message2 to stderr");
        System.out.println("message2 to stout");
   }
}

$ /usr/local/java/bin/javac Test.java

$ java Test
message1 to stdout
message1 to stderr
message2 to stderr
message2 to stout

$ date > outfile

$ cat outfile
Mon Jul 23 11:30:14 CDT 2007

$ java Test >> outfile 2>&1

$ cat outfile
message1 to stdout
message1 to stderr
message2 to stderr
message2 to stout

Oops, what happened to the date?

The inspiration for this test case is the java program fetchExc that I use
to get my Outlook mail.  It spawns procmail that I found to be core
dumping on the assertion:

0x6102f482 is in fhandler_base::read_overlapped(void*, unsigned int&)
(../../../../cygwin/winsup/cygwin/fhandler.cc:1739).
1734    void
1735    fhandler_base::read_overlapped (void *ptr, size_t& len)
1736    {
1737    #ifdef DEBUGGING
1738      assert (get_overlapped ());
1739      assert (get_overlapped ()->hEvent);
1740    #endif

I also noticed this, which I presume to be a simple test case for the
procmail problem above:

$ cygcheck -svr > cygcheck.out
   4435 [main] id 5264 fhandler_base::wait_overlapped: hEvent is zero?
   3730 [main] id 6088 fhandler_base::wait_overlapped: hEvent is zero?
   2879 [main] cygrunsrv 5132 fhandler_base::wait_overlapped: hEvent is
zero?
  56005 [main] cygrunsrv 1692 fhandler_base::wait_overlapped: hEvent is
zero?

Thanks.

-- 
Brian Ford
Lead Realtime Software Engineer
VITAL - Visual Simulation Systems
FlightSafety International
the best safety device in any aircraft is a well-trained crew...

Attachment: cygcheck.out
Description: Text document

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/