t/pmc/io.t: failure in I/O buffering test

[jkeenan]

We began to observe this test failure last night (Jan 20 2008) before the release. A number of people observed this on Linux boxes; the NetBSD Smolder box was reporting it as well. Curiously, I did not get the error on Darwin, the the FreeBSD Smolder box was not reporting it either.

The attached failure report was run on the same Linux box as the one on which I got the failure last night -- the same box I've been building/testing Parrot for years. This report was run at r35869 -- i.e., post-release. To make things even weirder, I did not get the error at the same rev number on a different Linux box.

[jkeenan]

t/pmc/io.t fails test 23, plus backtrace

[jkeenan]

At r35900 tonight, I observed that this test had once again begun to pass for me. However, it still failed at the same rev on our NetBSD Smolder reporter.

I can't identify what might have caused the test to resume passing.

[jkeenan]

This test is passing again on our NetBSD Smolder reporter (see, e.g.,  report 16966), so, keeping fingers crossed, I'm closing this ticket.

Thank you very much,
kid51