linux-ext4.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "George Spelvin" <linux@horizon.com>
To: linux-ext4@vger.kernel.org
Cc: linux@horizon.com
Subject: mke2fs -O 64bit -E resize=<anything> divides by 0
Date: 11 Nov 2012 17:27:16 -0500	[thread overview]
Message-ID: <20121111222716.11311.qmail@science.horizon.com> (raw)

I'm using v1.43-WIP-2012-09-22-10-g41bf599, last commit Oct. 14.

I'm trying to create a file system with 64bit support and specify a
maximum resize limit of 64 TiB = 2^34 blocks = 17179869184.

(gdb) run -n -t ext4 -O 64bit -E resize=4294967296 /dev/md1
Starting program: /root/e2fsprogs/misc/mke2fs -n -t ext4 -O 64bit -E resize=4294967295 /dev/md1
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
mke2fs 1.43-WIP (22-Sep-2012)

Program received signal SIGFPE, Arithmetic exception.
0x0000000000405f5a in parse_extended_opts (opts=<optimized out>, 
    param=0x64e200) at mke2fs.c:800
800                             gdpb = EXT2_DESC_PER_BLOCK(param);

The issue is that 

#define EXT2_DESC_PER_BLOCK(s)          (EXT2_BLOCK_SIZE(s) / EXT2_DESC_SIZE(s))
#define EXT2_DESC_SIZE(s)                                                \
       ((EXT2_SB(s)->s_feature_incompat & EXT4_FEATURE_INCOMPAT_64BIT) ? \
        (s)->s_desc_size : EXT2_MIN_DESC_SIZE)

and s_desc_size is 0 because parse_extended_opts is called from PRS which
is called very early in main() at line 2320, while s_desc_size is set up
in ext2fs_initialize, which is not called from main() until mke2fs.c:2353.

As a temporary workaround, I notice that ext2fs_initialize sets s_desc_size to
the fixed value EXT2_MIN_DESC_SIZE_64BIT, so I changed the #define as follows:

#define EXT2_DESC_SIZE(s)                                                \
       ((EXT2_SB(s)->s_feature_incompat & EXT4_FEATURE_INCOMPAT_64BIT) ? \
        (s)->s_desc_size ?: EXT2_MIN_DESC_SIZE_64BIT : EXT2_MIN_DESC_SIZE)

... which seems to work.


(One point that occurred to me while wrestling with this is that the
default resize limit of initial size * 1000 should perhaps be clamped
to 2^32 if 64bit is not enabled.)

             reply	other threads:[~2012-11-11 22:27 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-11-11 22:27 George Spelvin [this message]
2012-11-12  1:54 ` mke2fs -O 64bit -E resize=<anything> divides by 0 Andreas Dilger
2012-11-12  4:32   ` George Spelvin
2012-11-12  4:37   ` Eric Sandeen
2012-11-12  6:32     ` George Spelvin
2012-11-15 15:38       ` George Spelvin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20121111222716.11311.qmail@science.horizon.com \
    --to=linux@horizon.com \
    --cc=linux-ext4@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).