[Top] [All Lists]

Re: The mmap() problem is back

To: Ethan Benson <erbenson@xxxxxxxxxx>
Subject: Re: The mmap() problem is back
From: Keith Owens <kaos@xxxxxxx>
Date: Tue, 06 Aug 2002 00:21:25 +1000
Cc: linux-xfs@xxxxxxxxxxx
In-reply-to: Your message of "Mon, 05 Aug 2002 01:07:31 PST." <20020805010731.F24705@plato.local.lan>
Sender: owner-linux-xfs@xxxxxxxxxxx
On Mon, 5 Aug 2002 01:07:31 -0800, 
Ethan Benson <erbenson@xxxxxxxxxx> wrote:
>On Mon, Aug 05, 2002 at 10:58:06AM +0200, Wessel Dankers wrote:
>> On 2002-08-03 18:59:41+0200, Wessel Dankers wrote:
>> > Also of note may be that I'm running it with the ngpt futex patch and t=
>> > O(1) scheduler patch.
>> I just tried a kernel without any patches whatsoever: it exhibits the same
>> problem. I did notice that the file needs to be written in small chunks at
>> a time. For example:
>>      dd if=3D/dev/zero of=3Dfoo bs=3D1024 count=3D9
>i have reproduced this problem quite easily as well with a plain
>2.4.19 + split patches on powerpc, its not difficult at all to reproduce.

Could people reporting this problem please include their compiler and
binutils versions (gcc -v ; ld -v).  I am running 2.4.19-xfs from
immediately after it was released, and I cannot reproduce the problem.
Tested on UP and SMP using
  Reading specs from /usr/lib/gcc-lib/i386-redhat-linux/2.96/specs
  gcc version 2.96 20000731 (Red Hat Linux 7.1 2.96-81)
  GNU ld version 2.10.91 (with BFD

<Prev in Thread] Current Thread [Next in Thread>