GNU bug report logs - #64588
BUG: stat -f incorrectly reports a BTRFS device as a tmpfs

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 64588 in the body.
You can then email your comments to 64588 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: support <support <at> roosoft.ltd.uk>
To: bug-coreutils <bug-coreutils <at> gnu.org>
Subject: BUG: stat -f incorrectly reports a BTRFS device as a tmpfs
Date: Wed, 12 Jul 2023 23:46:38 +0100

coreutils version 9.1-1 on Debian testing, but also exists in previous 
versions.


Example:

stat -f  /dev/sdb
  File: "/dev/sdb"
    ID: eb91af7d7bda02dd Namelen: 255     Type: tmpfs
Block size: 4096       Fundamental block size: 4096
Blocks: Total: 2012292    Free: 2012292    Available: 2012292
Inodes: Total: 2012292    Free: 2011839


stat -f  /dev/disk/by-uuid/6abaa68a-2670-4d8b-8d2a-fd7321df9242
  File: "/dev/disk/by-uuid/6abaa68a-2670-4d8b-8d2a-fd7321df9242"
    ID: eb91af7d7bda02dd Namelen: 255     Type: tmpfs
Block size: 4096       Fundamental block size: 4096
Blocks: Total: 2012292    Free: 2012292    Available: 2012292
Inodes: Total: 2012292    Free: 2011839


blkid /dev/sdb
/dev/sdb: UUID="6abaa68a-2670-4d8b-8d2a-fd7321df9242" 
UUID_SUB="d15f1846-8f18-4ba0-9e2a-a6aaa7bbf83b" BLOCK_SIZE="4096" 
TYPE="btrfs"


Not sure why the reported fstype is wrong but obviously it is a bug.


TIA

Message #8 received at 64588 <at> debbugs.gnu.org (full text, mbox):

From: Pádraig Brady <P <at> draigBrady.com>
To: "debug <at> roosoft.ltd.uk" <debug <at> roosoft.ltd.uk>, 64588 <at> debbugs.gnu.org
Subject: Re: bug#64588: BUG: stat -f incorrectly reports a BTRFS device as a
 tmpfs
Date: Thu, 13 Jul 2023 10:59:48 +0100

tag 64588 notabug
close 64588
stop

On 12/07/2023 23:46, support wrote:
> coreutils version 9.1-1 on Debian testing, but also exists in previous
> versions.
> 
> 
> Example:
> 
> stat -f  /dev/sdb
>     File: "/dev/sdb"
>       ID: eb91af7d7bda02dd Namelen: 255     Type: tmpfs
> Block size: 4096       Fundamental block size: 4096
> Blocks: Total: 2012292    Free: 2012292    Available: 2012292
> Inodes: Total: 2012292    Free: 2011839
> 
> 
> stat -f  /dev/disk/by-uuid/6abaa68a-2670-4d8b-8d2a-fd7321df9242
>     File: "/dev/disk/by-uuid/6abaa68a-2670-4d8b-8d2a-fd7321df9242"
>       ID: eb91af7d7bda02dd Namelen: 255     Type: tmpfs
> Block size: 4096       Fundamental block size: 4096
> Blocks: Total: 2012292    Free: 2012292    Available: 2012292
> Inodes: Total: 2012292    Free: 2011839
> 
> 
> blkid /dev/sdb
> /dev/sdb: UUID="6abaa68a-2670-4d8b-8d2a-fd7321df9242"
> UUID_SUB="d15f1846-8f18-4ba0-9e2a-a6aaa7bbf83b" BLOCK_SIZE="4096"
> TYPE="btrfs"
> 
> 
> Not sure why the reported fstype is wrong but obviously it is a bug.

This is expected. From the info docs:

"stat does not search for specified device nodes in the file system list,
instead operating on them directly"

I.e. /dev/sdb is a device node in the /dev file system, which is tmpfs.
You can confirm this with:

  strace -e statfs stat -f -c %T /dev/sdb
  strace -e statfs df --output=fstype /dev/sdb

If you wanted to operate on a file system mounted on the device node you could:

  stat -f $(df --output=target /dev/sdb | tail -n1)

cheers,
Pádraig

Message #15 received at 64588 <at> debbugs.gnu.org (full text, mbox):

From: support <support <at> roosoft.ltd.uk>
To: Pádraig Brady <P <at> draigBrady.com>,
 "debug <at> roosoft.ltd.uk" <debug <at> roosoft.ltd.uk>, 64588 <at> debbugs.gnu.org
Subject: Re: bug#64588: BUG: stat -f incorrectly reports a BTRFS device as a
 tmpfs
Date: Thu, 13 Jul 2023 11:33:57 +0100

Hmmm ok,

Well the reason I flagged it as a bug, is because the btrfsmaintenance 
package uses a script to test to see if a filesystem is btrfs and 
incorrectly uses stat -f to do it. This must have worked in the past 
else they would not use it.

The code snippet is:


# function: is_btrfs
# parameter: path to a mounted filesystem
#
# check if filesystem is a btrfs
is_btrfs() {
    local FS=$(stat -f --format=%T "$1")
    [ "$FS" = "btrfs" ] && return 0
    return 1
}


Obviously that is targeted at mounted filesystems, but the documentation 
and usage always shows "<path>|<device>" as a parameter.

I guess I should file a bug against that package but I am curious as to 
when the default behaviour changed.

Thanks for your response.

On 13/07/2023 10:59, Pádraig Brady wrote:
> tag 64588 notabug
> close 64588
> stop
>
> On 12/07/2023 23:46, support wrote:
>> coreutils version 9.1-1 on Debian testing, but also exists in previous
>> versions.
>>
>>
>> Example:
>>
>> stat -f  /dev/sdb
>>     File: "/dev/sdb"
>>       ID: eb91af7d7bda02dd Namelen: 255     Type: tmpfs
>> Block size: 4096       Fundamental block size: 4096
>> Blocks: Total: 2012292    Free: 2012292    Available: 2012292
>> Inodes: Total: 2012292    Free: 2011839
>>
>>
>> stat -f  /dev/disk/by-uuid/6abaa68a-2670-4d8b-8d2a-fd7321df9242
>>     File: "/dev/disk/by-uuid/6abaa68a-2670-4d8b-8d2a-fd7321df9242"
>>       ID: eb91af7d7bda02dd Namelen: 255     Type: tmpfs
>> Block size: 4096       Fundamental block size: 4096
>> Blocks: Total: 2012292    Free: 2012292    Available: 2012292
>> Inodes: Total: 2012292    Free: 2011839
>>
>>
>> blkid /dev/sdb
>> /dev/sdb: UUID="6abaa68a-2670-4d8b-8d2a-fd7321df9242"
>> UUID_SUB="d15f1846-8f18-4ba0-9e2a-a6aaa7bbf83b" BLOCK_SIZE="4096"
>> TYPE="btrfs"
>>
>>
>> Not sure why the reported fstype is wrong but obviously it is a bug.
>
> This is expected. From the info docs:
>
> "stat does not search for specified device nodes in the file system list,
> instead operating on them directly"
>
> I.e. /dev/sdb is a device node in the /dev file system, which is tmpfs.
> You can confirm this with:
>
>   strace -e statfs stat -f -c %T /dev/sdb
>   strace -e statfs df --output=fstype /dev/sdb
>
> If you wanted to operate on a file system mounted on the device node 
> you could:
>
>   stat -f $(df --output=target /dev/sdb | tail -n1)
>
> cheers,
> Pádraig

Message #18 received at 64588 <at> debbugs.gnu.org (full text, mbox):

From: Pádraig Brady <P <at> draigBrady.com>
To: "debug <at> roosoft.ltd.uk" <debug <at> roosoft.ltd.uk>, 64588 <at> debbugs.gnu.org
Subject: Re: bug#64588: BUG: stat -f incorrectly reports a BTRFS device as a
 tmpfs
Date: Thu, 13 Jul 2023 13:35:28 +0100

On 13/07/2023 11:33, support wrote:
> Hmmm ok,
> 
> Well the reason I flagged it as a bug, is because the btrfsmaintenance
> package uses a script to test to see if a filesystem is btrfs and
> incorrectly uses stat -f to do it. This must have worked in the past
> else they would not use it.
> 
> The code snippet is:
> 
> 
> # function: is_btrfs
> # parameter: path to a mounted filesystem
> #
> # check if filesystem is a btrfs
> is_btrfs() {
>       local FS=$(stat -f --format=%T "$1")
>       [ "$FS" = "btrfs" ] && return 0
>       return 1
> }
> 
> 
> Obviously that is targeted at mounted filesystems, but the documentation
> and usage always shows "<path>|<device>" as a parameter.
> 
> I guess I should file a bug against that package but I am curious as to
> when the default behaviour changed.

It has never changed AFAIK

thanks,
Pádraig

Message #21 received at 64588 <at> debbugs.gnu.org (full text, mbox):

From: Andreas Schwab <schwab <at> linux-m68k.org>
To: support <support <at> roosoft.ltd.uk>
Cc: 64588 <at> debbugs.gnu.org, "debug <at> roosoft.ltd.uk" <debug <at> roosoft.ltd.uk>,
 Pádraig Brady <P <at> draigBrady.com>
Subject: Re: bug#64588: BUG: stat -f incorrectly reports a BTRFS device as a
 tmpfs
Date: Thu, 13 Jul 2023 18:02:12 +0200

On Jul 13 2023, support wrote:

> # function: is_btrfs
> # parameter: path to a mounted filesystem

A device node is not a path to a mounted filesystem

-- 
Andreas Schwab, schwab <at> linux-m68k.org
GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510  2552 DF73 E780 A9DA AEC1
"And now for something completely different."

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.