|
Re: linux-kernel say bug
2003-12-04 21:33:37
On Thu, Dec 04, 2003 at 04:07:49PM -0800, Greg Harper wrote:
> <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
> <html>
Wish you hadn't posted in html.
I had to go to another system to read this.
> <head>
> <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
> <title></title>
> </head>
> <body bgcolor="#ffffff" text="#000000">
> I get the same error, but if you look at the next line:
> <br>
> <br>
> Dec 4 13:15:25 mail kernel: application bug: dumper(12282) has SIGCHLD
> set to SIG_IGN but calls wait().
> <br>
> Dec 4 13:15:25 mail kernel: (see the NOTES section of 'man 2 wait').
> Workaround activated.
> <br>
> <br>
> Then look at the man page on wait:
> <br>
> <br>
> NOTES
> <br>
> The Single Unix Specification describes a flag
> SA_NOCLDWAIT
> (not sup-
> <br>
> ported under Linux) such that if either this
> flag is set, or the
> action
> <br>
> for SIGCHLD is set to SIG_IGN then children
> that exit do not
> become
> <br>
> zombies and a call to wait() or waitpid() will
> block until all
> children
> <br>
> have exited, and then fail with errno set to
> ECHILD.
> <br>
> <br>
> The original POSIX standard left the behaviour
> of setting
> SIGCHLD to
> <br>
> SIG_IGN unspecified.
> Later standards, including SUSv2 and
> POSIX
> <br>
> 1003.1-2001 specify the behaviour just
> described as an
> XSI-compliance
> <br>
> option. Linux does not
> conform to the second of the two
> points just
> <br>
> described: if a wait() or waitpid() call is
> made while SIGCHLD is
> being
> <br>
> ignored, the call behaves just as though
> SIGCHLD were not being
> ignored,
> <br>
> that is, the call blocks until the
> next child terminates
> and then
> <br>
> returns the PID and status of that child.
> <br>
> <br>
> So its really nothing serious, just something that should be fixed in
> future versions.
> <br>
> <br>
> -Greg
> <br>
> <br>
> <br>
> <br>
> Mats Blomstrand wrote:<br>
> <blockquote cite="mid200312042354.hB4Nswv12985 AT bergelmer.thn.htu DOT se"
> type="cite">
> <blockquote type="cite">
> <pre wrap="">modified I don't recall. Is your's an older amanda release?
> </pre>
> </blockquote>
> <pre wrap=""><!---->
> On that machine its amanda-2.4.4.
> //Mats
>
>
>
> </pre>
> </blockquote>
> </body>
> </html>
>
>>> End of included message <<<
You say "should be fixed in a future release". Did you mean will be fixed
in linux or the application encountering the problem should be fixed.
--
Jon H. LaBadie jon AT jgcomp DOT com
JG Computing
4455 Province Line Road (609) 252-0159
Princeton, NJ 08540-4322 (609) 683-7220 (fax)
|
|
|
