[PATCH] btrfs-progs: force overwrite should wipe stale SB

[PATCH] btrfs-progs: force overwrite should wipe stale SB

[Anand Jain]

Here is a test case which says it all..

mkfs.xfs -f $DEV
mkfs.btrfs -f $DEV
mount $DEV $MNT
mount: /dev/vdiskc: more filesystems detected. This should not happen,
       use -t <type> to explicitly specify the filesystem type or
       use wipefs(8) to clean up the device.

mount: you must specify the filesystem type

with this patch btrfs_prepare_device() also wipes old FS if any,
btrfs_prepare_device() is called after we have verified that
user has provided -f option.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
 utils.c |   31 +++++++++++++++++++++++++++++++
 1 files changed, 31 insertions(+), 0 deletions(-)

diff --git a/utils.c b/utils.c
index 0064587..8b0c751 100644
--- a/utils.c
+++ b/utils.c
@@ -678,6 +678,35 @@ int btrfs_add_to_fsid(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
+static void wipe_existing_fs(int fd)
+{
+	blkid_probe     pr = NULL;
+
+	pr = blkid_new_probe();
+	if (!pr)
+		return;
+
+	if (blkid_probe_set_device(pr, fd, 0, 0))
+		goto out;
+
+	blkid_probe_enable_superblocks(pr, 1);
+	blkid_probe_set_superblocks_flags(pr,
+				BLKID_SUBLKS_MAGIC |
+				BLKID_SUBLKS_TYPE |
+				BLKID_SUBLKS_USAGE |
+				BLKID_SUBLKS_LABEL |
+				BLKID_SUBLKS_UUID);
+
+	blkid_probe_enable_partitions(pr, 1);
+
+	while (blkid_do_probe(pr) == 0)
+		blkid_do_wipe(pr, 0);
+
+	fsync(fd);
+out:
+	blkid_free_probe(pr);
+}
+
 int btrfs_prepare_device(int fd, char *file, int zero_end, u64 *block_count_ret,
 			   u64 max_block_count, int *mixed, int discard)
 {
@@ -729,6 +758,8 @@ int btrfs_prepare_device(int fd, char *file, int zero_end, u64 *block_count_ret,
 		return 1;
 	}
 
+	wipe_existing_fs(fd);
+
 	*block_count_ret = block_count;
 	return 0;
 }
-- 
1.7.1

Re: [PATCH] btrfs-progs: force overwrite should wipe stale SB

[David Sterba]

On Fri, Sep 05, 2014 at 11:51:24PM +0800, Anand Jain wrote:
> Here is a test case which says it all..
> 
> mkfs.xfs -f $DEV
> mkfs.btrfs -f $DEV
> mount $DEV $MNT
> mount: /dev/vdiskc: more filesystems detected. This should not happen,
>        use -t <type> to explicitly specify the filesystem type or
>        use wipefs(8) to clean up the device.
> 
> mount: you must specify the filesystem type
> 
> with this patch btrfs_prepare_device() also wipes old FS if any,
> btrfs_prepare_device() is called after we have verified that
> user has provided -f option.

Good catch.

Re: [PATCH] btrfs-progs: force overwrite should wipe stale SB

[David Sterba]

On Fri, Sep 05, 2014 at 11:51:24PM +0800, Anand Jain wrote:
> +static void wipe_existing_fs(int fd)
> +{
...
> +	blkid_probe_enable_partitions(pr, 1);
> +
> +	while (blkid_do_probe(pr) == 0)
> +		blkid_do_wipe(pr, 0);

Reported by a user on IRC, I verified it on one of my old test boxes.
The function is missing in older versions of blkid and causes comilation
errors:

    [CC]     utils.o
utils.c: In function ‘wipe_existing_fs’:
utils.c:705:3: warning: implicit declaration of function ‘blkid_do_wipe

utils.o: In function `wipe_existing_fs':
/labs/dsterba/btrfs-progs/utils.c:705: undefined reference to `blkid_do_wipe'
collect2: ld returned 1 exit status

blkid_do_wipe is from Novembe 2011, so nothing really new, but I'd still
like to keep the backward compatibility of btrfs-progs.  The
implementation of blkid_do_wipe uses some internal data so it's not
possible to simply copy the implementation. Maybe wiping the first
megabyte is enough.

Re: [PATCH] btrfs-progs: force overwrite should wipe stale SB

[Anand Jain]

On 10/01/2014 07:26 PM, David Sterba wrote:
> On Fri, Sep 05, 2014 at 11:51:24PM +0800, Anand Jain wrote:
>> +static void wipe_existing_fs(int fd)
>> +{
> ...
>> +	blkid_probe_enable_partitions(pr, 1);
>> +
>> +	while (blkid_do_probe(pr) == 0)
>> +		blkid_do_wipe(pr, 0);
>
> Reported by a user on IRC, I verified it on one of my old test boxes.
> The function is missing in older versions of blkid and causes comilation
> errors:
>
>      [CC]     utils.o
> utils.c: In function ‘wipe_existing_fs’:
> utils.c:705:3: warning: implicit declaration of function ‘blkid_do_wipe
>
> utils.o: In function `wipe_existing_fs':
> /labs/dsterba/btrfs-progs/utils.c:705: undefined reference to `blkid_do_wipe'
> collect2: ld returned 1 exit status
 >
> blkid_do_wipe is from Novembe 2011, so nothing really new, but I'd still
> like to keep the backward compatibility of btrfs-progs.  The

> implementation of blkid_do_wipe uses some internal data so it's not
> possible to simply copy the implementation. Maybe wiping the first
> megabyte is enough.

  for the sake of backward compatible.. the only choice I see as of now
  yes to write our own wipe. will send one.

Anand


> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

[PATCH v2] btrfs-progs: force overwrite should wipe stale SB

[Anand Jain]

(I am unable to reproduce the issue, tried to go back with progs versions
but still the same. So as of now this code remains untested, suggest to
wait till we have a reproducible test case).

Here is a test case which says it all..

mkfs.xfs -f $DEV
mkfs.btrfs -f $DEV
mount $DEV $MNT
mount: /dev/vdiskc: more filesystems detected. This should not happen,
       use -t <type> to explicitly specify the filesystem type or
       use wipefs(8) to clean up the device.

mount: you must specify the filesystem type

with this patch btrfs_prepare_device() also wipes old FS if any,
btrfs_prepare_device() is called after we have verified that
user has provided -f option.

v2: to satisfy the backward compatibility issue, replace
blkid_do_wipe() with local wipe function.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
---
 utils.c | 38 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 38 insertions(+)

diff --git a/utils.c b/utils.c
index fb78dd6..5f21737 100644
--- a/utils.c
+++ b/utils.c
@@ -680,6 +680,42 @@ int btrfs_add_to_fsid(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
+static void btrfs_wipe_existing_sb(int fd)
+{
+	const char *off = NULL;
+	size_t len = 0;
+	loff_t offset;
+	char buf[BUFSIZ];
+	int rc = 0;
+	blkid_probe pr = NULL;
+
+	pr = blkid_new_probe();
+	if (!pr)
+		return;
+
+	if (blkid_probe_set_device(pr, fd, 0, 0))
+		goto out;
+
+	rc = blkid_probe_lookup_value(pr, "SBMAGIC_OFFSET", &off, NULL);
+	if (!rc)
+		rc = blkid_probe_lookup_value(pr, "SBMAGIC", NULL, &len);
+
+	if (rc || len == 0 || off == NULL)
+		goto out;
+
+	offset = strtoll(off, NULL, 10);
+	if (len > sizeof(buf))
+		len = sizeof(buf);
+
+	memset(buf, 0, len);
+	rc = pwrite(fd, buf, len, offset);
+	fsync(fd);
+
+out:
+	blkid_free_probe(pr);
+	return;
+}
+
 int btrfs_prepare_device(int fd, char *file, int zero_end, u64 *block_count_ret,
 			   u64 max_block_count, int *mixed, int discard)
 {
@@ -731,6 +767,8 @@ int btrfs_prepare_device(int fd, char *file, int zero_end, u64 *block_count_ret,
 		return 1;
 	}
 
+	btrfs_wipe_existing_sb(fd);
+
 	*block_count_ret = block_count;
 	return 0;
 }
-- 
2.0.0.153.g79dcccc

Re: [PATCH v2] btrfs-progs: force overwrite should wipe stale SB

[David Sterba]

On Thu, Oct 02, 2014 at 07:22:09AM +0800, Anand Jain wrote:
> (I am unable to reproduce the issue, tried to go back with progs versions
> but still the same. So as of now this code remains untested, suggest to
> wait till we have a reproducible test case).
> 
> Here is a test case which says it all..
> 
> mkfs.xfs -f $DEV
> mkfs.btrfs -f $DEV
> mount $DEV $MNT
> mount: /dev/vdiskc: more filesystems detected. This should not happen,
>        use -t <type> to explicitly specify the filesystem type or
>        use wipefs(8) to clean up the device.
> 
> mount: you must specify the filesystem type
> 
> with this patch btrfs_prepare_device() also wipes old FS if any,
> btrfs_prepare_device() is called after we have verified that
> user has provided -f option.
> 
> v2: to satisfy the backward compatibility issue, replace
> blkid_do_wipe() with local wipe function.

Thank you, works for me, added to 3.17 queue.

Re: [PATCH v2] btrfs-progs: force overwrite should wipe stale SB

[Anand Jain]

  Thanks.
Anand

On 10/10/2014 03:39 PM, David Sterba wrote:
> On Thu, Oct 02, 2014 at 07:22:09AM +0800, Anand Jain wrote:
>> (I am unable to reproduce the issue, tried to go back with progs versions
>> but still the same. So as of now this code remains untested, suggest to
>> wait till we have a reproducible test case).
>>
>> Here is a test case which says it all..
>>
>> mkfs.xfs -f $DEV
>> mkfs.btrfs -f $DEV
>> mount $DEV $MNT
>> mount: /dev/vdiskc: more filesystems detected. This should not happen,
>>         use -t <type> to explicitly specify the filesystem type or
>>         use wipefs(8) to clean up the device.
>>
>> mount: you must specify the filesystem type
>>
>> with this patch btrfs_prepare_device() also wipes old FS if any,
>> btrfs_prepare_device() is called after we have verified that
>> user has provided -f option.
>>
>> v2: to satisfy the backward compatibility issue, replace
>> blkid_do_wipe() with local wipe function.
>
> Thank you, works for me, added to 3.17 queue.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>