[Updated on 2006-9-12] Linux AIO module, Re: PERFORCE change 105930 for review

From: Intron is my alias on the Internet <mag_at_intron.ac>
Date: Tue, 12 Sep 2006 19:25:52 +0800
Download:  http://ftp.intron.ac/tmp/linux_aio-20060912.tar.bz2

This update includes:

1. Some minor modifications in comments and debugging messages.
2. Calling uma_zcreate() without the flag UMA_ZONE_NOFREE.

Alexander, please commit the only modified file "linux_aio.c" to the
Perforce repository. When you commit my module to the CVS repository,
please remove the two problems listed in my previous "readme.txt".

------------------------------------------------------------------------
                                                From Beijing, China


Attilio Rao wrote:

> 2006/9/12, Attilio Rao <attilio_at_freebsd.org>:
>> 2006/9/12, Alexander Leidinger <netchild_at_freebsd.org>:
>> > Quoting John Baldwin <jhb_at_freebsd.org> (from Mon, 11 Sep 2006 15:28:48 
>> -0400):
>> >
>> > > I've only gotten this message when I've leaked memory in a zone.  Do 
>> you have
>> > > an example self-contained kernel module that produces this?
>> >
>> > Download
>> > http://www.Leidinger.net/FreeBSD/linuxolator/linuxolator-p4.diff (or
>> > check out the p4-branch) and recompile the aio and linux modules.
>> 
>> You might put the code under "Initialize the new context"
>> (linux_io_setup) into the init function for your UMA object, and so
>> the destructive pair.
>> 
>> It might not solve your problem, but it is more elegant.
> 
> BTW, the error that you receive effectively happens only on memory leak.
> This is called from the keg destructor (uma_core.c).
> 
> Attilio
> 
> 
> -- 
> Peace can only be achieved by understanding - A. Einstein
Received on Tue Sep 12 2006 - 09:25:54 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:00 UTC