in reply to Re^9: baton passing threads and cond_signal
in thread baton passing threads and cond_signal

The problem might be that your system's implementation of cond_wait isn't atomic

Which would be a bug worth reporting. An easy way to verify this would be to run the code above without the yields on single cpu Linux system and see if the problem occurs there?

  • Comment on Re^10: baton passing threads and cond_signal

Replies are listed 'Best First'.
Re^11: baton passing threads and cond_signal
by BrowserUk (Patriarch) on Aug 22, 2007 at 19:00 UTC

    Can we take that the lack of a refutation that this problem also occurs on a single cpu Linux system to mean that the problem lies in the semantics of the whole cond_* mechanism and not just in the Win32 emulation of it?

    Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
    "Science is about questioning the status quo. Questioning authority".
    In the absence of evidence, opinion is indistinguishable from prejudice.
[reply]
[d/l]

      The people discussing perlbug 41574 think it's specific to (some versions of) Windows.
      The only person in this thread reporting the problem uses WinXP.
      I only have access to Windows machines.

      I'd take the lack of refutation as nothing more than people not reading the thread this deep.

[reply]

        Fair do's ike {.oO(I hope that abbrev is endearing and not insulting)}. I have an axe to grind here in that I think that the pthreads APIs are broken, hence my following along on this thread.

        If the OP doesn't raise another root level post to clarify the problem, I will.

        Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
        "Science is about questioning the status quo. Questioning authority".
        In the absence of evidence, opinion is indistinguishable from prejudice.
[reply]

In Section Seekers of Perl Wisdom