[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <1177115236.3778.23.camel@dyn9047017103.beaverton.ibm.com>
Date: Fri, 20 Apr 2007 17:27:15 -0700
From: Mingming Cao <cmm@...ibm.com>
To: Eric Sandeen <sandeen@...hat.com>
Cc: Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
ext4 development <linux-ext4@...r.kernel.org>
Subject: Re: [PATCH] fix ext2 allocator overflows above 31 bit blocks
On Fri, 2007-04-20 at 18:14 -0500, Eric Sandeen wrote:
> Andreas Dilger wrote:
> > On Apr 20, 2007 12:10 -0500, Eric Sandeen wrote:
> >> If ext3 can do 16T, ext2 probably should be able to as well.
> >> There are still "int" block containers in the block allocation path
> >> that need to be fixed up.
> >
> > Yeah, but who wants to do 16TB e2fsck on every boot? I think there
> > needs to be some limits imposed for the sake of usability.
>
> I figure this is in the fine tradition of "enough rope to hang oneself"
>
> If you have 16T of filesystem you probably know enough to not hang
> yourself this way.
>
> *shrug*
>
> It's a bug, today.
They are fixed in mm tree, as part of the patches which backports ext3
block reservation code to ext2. filesystem block numbers are all
ext2_fsblk_t type(i.e. unsigned long)(see ext2_new_blocks()). Maybe need
a round of thorough review to see if anything left, but I think what in
mm tree looks good.
And those patches in mm tree also backports the ext3 best-effort
allocates multiple blocks code (allocate multiple blocks within the
block reservation window as much as possible), FYI.
Mingming
> If we need another change to limit ext2 to 500G or
> something, fine by me. :)
>
> -Eric
> -
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
-
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists