Guys,
I'm building the DBI on AIX and I'm seeing something I'm not sure I like in the test results. I'm using the pre-built perl that comes with AIX 5.3 and IBM's C compiler that was used to build it. To cut down on most of the miscellaneous messages about skipping SQL-Statement, I went and installed SQL-Statement-1.33 before I attempted to build DBI again. But even with cleaner log files, it did nothing for the log entries I was worried about. I saved & reviewed the output of "perl Makefile.PL" & "make", but didn't see anything in there that looked suspicious. (logs available on request) Here's a sample of the messages I'm talking about. t/35thrclone....................Attempt to free unreferenced scalar: SV 0x3081c914 during global destruction. Attempt to free unreferenced scalar: SV 0x306b9244 during global destruction. Attempt to free unreferenced scalar: SV 0x305501c4 during global destruction. Attempt to free unreferenced scalar: SV 0x3037bcf4 during global destruction. Ok These messages appear for several test cases. But the log ends with a message that implies all is OK: t/zvxnp_85gofer.................ok 18/234 skipped: caching of metadata methods returning sth not yet implemented All tests successful, 38 tests and 553 subtests skipped. Files=178, Tests=7937, 147 wallclock secs (111.78 cusr + 25.73 csys = 137.51 CPU) PERL_DL_NONLAZY=1 /usr/bin/perl "-Iblib/lib" "-Iblib/arch" test.pl test.pl DBI test application $Revision: 12537 $ Switch: DBI 1.616 by Tim Bunce, 1.616 Available Drivers: DBM, ExampleP, File, Gofer, Proxy, Sponge dbi:ExampleP:: testing 3 sets of 20 connections: Connecting... 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Disconnecting... Connecting... 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Disconnecting... Connecting... 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Disconnecting... connect 20 and disconnect them, 3 times: 0.0050s / 60 = 0.0001s Testing handle creation speed... 27777 NullP sth/s perl 5.008002 aix-thread-multi (cc_r 5.0.2.0 -O) 0.000036s test.pl done Target "test" is up to date. Are these messages about attempts to free unreferenced scalars important? It sounds like I'm hitting some type of problem that's being ignored in the pass/fail tests. If I can ignore it, I'm good to go, but I'm a bit nervous about this warning. I can provide the output of "perl Makefile.PL", "make", and "make test" on request. Curtis