linux-ntfs-dev — Development discussions and bug reports.

Re: [Linux-NTFS-Dev] Problem with ParentMftReference

[Anton A. <ai...@ca...>]

Or you may be forgetting about hard links?  i.e. where there are  
multiple parents for the same mft record (it will have multiple file  
name attributes in it each pointing to a different parent).

Other than that you must either be seeing corruption or are doing  
something wrong.  If you show an example we can probably tell you  
which.  (-:

Best regards,

	Anton

On 27 Apr 2007, at 10:05, Anton Altaparmakov wrote:

> Hi,
>
> Are you forgetting about extent mft records?  You said you are  
> reading the MFT sequentially so there will be mft records that  
> point back to their base mft record...
>
> Best regards,
>
> 	Anton
>
> On 27 Apr 2007, at 09:09, pawan rathi wrote:
>
>> Dear Anton,
>>         I have taken care of the fixups in my code both for the  
>> MFT record and the Index record. So, that is not the problem for  
>> sure.
>>        I thought the reason might be the inconsistency caused in  
>> the file-system because of improper shutdown or anything else. So  
>> i did "chkdsk" on the file system. But still couldn't figure out  
>> the problem.
>>
>> Pawan.
>>
>> Anton Altaparmakov <ai...@ca...> wrote:
>> On 27 Apr 2007, at 06:28, pawan rathi wrote:
>>
>> > Hi All,
>> > I am reading the MFT sequentially. In the FileName attribute of
>> > each MFT record there is a field called ParentMftReference which
>> > contains the MFTReference of it's parent.
>> > for eg. if there is a file called 1.txt in Directory A and
>> > MFTReference of A = 0x20
>> > then ParentMftReference of 1.txt = 0x20.
>> >
>> > Ideally it should happen as shown above. Even i am gettin the same
>> > results maximum times.
>> >
>> > But, sometimes the ParentMftReference of a child doesn't match with
>> > the MftReference of the Parent.
>> >
>> > If anyone has faced such issue before or if anyone knows how to
>> > resolve it or if you can put some light into it, please revert  
>> back.
>> >
>> You probably forgot to remove the fixups from your MFT records and/or
>> index records...
>>
>> See http://www.linux-ntfs.org/content/view/109/48#concept_fixup
>>
>> Best regards,
>>
>> Anton

-- 
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer, http://www.linux-ntfs.org/

Re: [Linux-NTFS-Dev] Problem with ParentMftReference

[Anton A. <ai...@ca...>]

Hi,

Are you forgetting about extent mft records?  You said you are  
reading the MFT sequentially so there will be mft records that point  
back to their base mft record...

Best regards,

	Anton

On 27 Apr 2007, at 09:09, pawan rathi wrote:

> Dear Anton,
>         I have taken care of the fixups in my code both for the MFT  
> record and the Index record. So, that is not the problem for sure.
>        I thought the reason might be the inconsistency caused in  
> the file-system because of improper shutdown or anything else. So i  
> did "chkdsk" on the file system. But still couldn't figure out the  
> problem.
>
> Pawan.
>
> Anton Altaparmakov <ai...@ca...> wrote:
> On 27 Apr 2007, at 06:28, pawan rathi wrote:
>
> > Hi All,
> > I am reading the MFT sequentially. In the FileName attribute of
> > each MFT record there is a field called ParentMftReference which
> > contains the MFTReference of it's parent.
> > for eg. if there is a file called 1.txt in Directory A and
> > MFTReference of A = 0x20
> > then ParentMftReference of 1.txt = 0x20.
> >
> > Ideally it should happen as shown above. Even i am gettin the same
> > results maximum times.
> >
> > But, sometimes the ParentMftReference of a child doesn't match with
> > the MftReference of the Parent.
> >
> > If anyone has faced such issue before or if anyone knows how to
> > resolve it or if you can put some light into it, please revert back.
> >
> You probably forgot to remove the fixups from your MFT records and/or
> index records...
>
> See http://www.linux-ntfs.org/content/view/109/48#concept_fixup
>
> Best regards,
>
> Anton

-- 
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer, http://www.linux-ntfs.org/

Re: [Linux-NTFS-Dev] Problem with ParentMftReference

[Anton A. <ai...@ca...>]

On 27 Apr 2007, at 06:28, pawan rathi wrote:

> Hi All,
>     I am reading the MFT sequentially. In the FileName attribute of  
> each MFT record there is a field called ParentMftReference which  
> contains the MFTReference of it's parent.
>  for eg. if there is a file called 1.txt in Directory A and  
> MFTReference of A = 0x20
>            then ParentMftReference of 1.txt = 0x20.
>
> Ideally it should happen as shown above. Even i am gettin the same  
> results maximum times.
>
> But, sometimes the ParentMftReference of a child doesn't match with  
> the MftReference of the Parent.
>
> If anyone has faced such issue before or if anyone knows how to  
> resolve it or if you can put some light into it, please revert back.
>
You probably forgot to remove the fixups from your MFT records and/or  
index records...

See http://www.linux-ntfs.org/content/view/109/48#concept_fixup

Best regards,

	Anton
-- 
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer, http://www.linux-ntfs.org/

[Linux-NTFS-Dev] Problem with ParentMftReference

[pawan r. <paw...@ya...>]

Hi All,
    I am reading the MFT sequentially. In the FileName attribute of each MFT record there is a field called ParentMftReference which contains the MFTReference of it's parent.
 for eg. if there is a file called 1.txt in Directory A and MFTReference of A = 0x20
           then ParentMftReference of 1.txt = 0x20.

Ideally it should happen as shown above. Even i am gettin the same results maximum times.

But, sometimes the ParentMftReference of a child doesn't match with the MftReference of the Parent.

If anyone has faced such issue before or if anyone knows how to resolve it or if you can put some light into it, please revert back.


Thanks and Regards,
Pawan Rathi.

       
---------------------------------
 SHOUT IT OUT! Tell everyone, from anywhere, that you're online on Yahoo! Messenger 

[Linux-NTFS-Dev] Unable to get INDEX ROOT attribute

[AD K. <rao...@ms...>]

Hi,

I would like to find out if for any reason I am unable to get the ROOT
ATTRIBUTE. I seem to get both ROOT and ALLOCATION attributes for all folders
except i386 folder. Can anyone help me to suggest what I am missing here. 

I am doing test on Windows2003 and Windows200 systems, both have similar
issues.

Thanks a lot in advance.
-- 
View this message in context: http://www.nabble.com/Unable-to-get-INDEX-ROOT-attribute-tf3649191.html#a10193105
Sent from the linux-ntfs-dev mailing list archive at Nabble.com.

Re: [Linux-NTFS-Dev] Writing mft records without existing buffers is not implemented yet.

[Michael B. <mic...@gm...>]

This is unfortunately the latest kernel available on CentOS 4 (RHEL4
clone) unless I want to roll my own, which isn't that fun from a
support POV.

Best regards
 Michael Boman

On 4/17/07, Anton Altaparmakov <ai...@ca...> wrote:
> Hi,
>
> Thanks for the report!
>
> I am afraid you are using an ancient NTFS driver (and kernel for that
> matter!).
>
> This was implemented/fixed several years ago...
>
> If you upgrade to a more recent kernel this error will go away.
>
> Best regards,
>
>         Anton
> --
> Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
> Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
> Linux NTFS maintainer, http://www.linux-ntfs.org/
>
>
>


-- 
IT Security Researcher & Developer
http://proxy.11a.nu | http://securitytinker.com
"It is well known that a vital ingredient of success is not knowing
that what you're attempting can't be done." -- Terry Pratchett

Re: [Linux-NTFS-Dev] Writing mft records without existing buffers is not implemented yet.

[Anton A. <ai...@ca...>]

Hi,

Thanks for the report!

I am afraid you are using an ancient NTFS driver (and kernel for that  
matter!).

This was implemented/fixed several years ago...

If you upgrade to a more recent kernel this error will go away.

Best regards,

	Anton
-- 
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer, http://www.linux-ntfs.org/

[Linux-NTFS-Dev] Writing mft records without existing buffers is not implemented yet.

[Michael B. <mic...@gm...>]

NTFS-fs error (device sdb2): write_mft_record_nolock(): Writing mft
records without existing buffers is not implemented yet.  Please email
lin...@li... and say that you saw this
message.  Thank you.


# uname -a
Linux vm2 2.6.9-42.0.10.ELsmp #1 SMP Tue Feb 27 10:11:19 EST 2007 i686
i686 i386 GNU/Linux


# rpm -qa | grep ntfs
kernel-module-ntfs-2.6.9-42.0.10.ELsmp-2.1.20-0.rr.10.0


Best regards
 Michael Boman

-- 
IT Security Researcher & Developer
http://proxy.11a.nu | http://securitytinker.com
"It is well known that a vital ingredient of success is not knowing
that what you're attempting can't be done." -- Terry Pratchett

Re: [Linux-NTFS-Dev] [KJ] Trivial patch2:changing const char foo* to const char foo[] in fssubsystem

[<mar...@gm...>]

2007/4/16, sur...@wi... <sur...@wi...>:
your patch is corrupted - try to apply it

>diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c
>index e6f5799..b9ef210 100644
>--- a/fs/binfmt_misc.c
>+++ b/fs/binfmt_misc.c
>@@ -445,7 +445,7 @@ static void entry_status(Node *e, char *page)
> {
> 	char *dp;
> 	char *status = "disabled";
>-	const char * flags = "flags: ";
>+	const char  flags[] = "flags: ";
>
> 	if (test_bit(Enabled, &e->flags))
> 		status = "enabled";
it creates more code (on every call, this string will be copied)
mark this variable as static

>@@ -2369,7 +2369,7 @@ int ntfs_truncate(struct inode *vi)
> 	ntfs_attr_search_ctx *ctx;
> 	MFT_RECORD *m;
> 	ATTR_RECORD *a;
>-	const char *te = "  Leaving file length out of sync with i_size.";
>+	const char te[] = "  Leaving file length out of sync with i_size.";
> 	int err, mp_size, size_change, alloc_change;
> 	u32 attr_len;
>
again

> diff --git a/fs/ntfs/super.c b/fs/ntfs/super.c
> index 1594c90..1fe6ddd 100644
> --- a/fs/ntfs/super.c
> +++ b/fs/ntfs/super.c
> @@ -647,7 +647,7 @@ not_ntfs:
>  static struct buffer_head *read_ntfs_boot_sector(struct super_block
> *sb,
>                 const int silent)
>  {
> -       const char *read_err_str = "Unable to read %s boot sector.";
> +       const char read_err_str[] = "Unable to read %s boot sector.";
>         struct buffer_head *bh_primary, *bh_backup;
>         sector_t nr_blocks = NTFS_SB(sb)->nr_blocks;
>
again

> @@ -2156,9 +2156,9 @@ get_ctx_vol_failed:
>         }
>         /* If (still) a read-write mount, stamp the transaction log. */
>         if (!(sb->s_flags & MS_RDONLY) && !ntfs_stamp_usnjrnl(vol)) {
> -               static const char *es1 = "Failed to stamp transaction log "
> +               static const char []es1 = "Failed to stamp transaction log "
>                                 "($UsnJrnl)";
> -               static const char *es2 = ".  Run chkdsk.";
> +               static const char []es2 = ".  Run chkdsk.";
>
>                 /* Convert to a read-only mount. */
>                 if (!(vol->on_errors & (ON_ERRORS_REMOUNT_RO |
doesn't compile

[Linux-NTFS-Dev] [Fwd: Trivial patch2:changing const char foo* to const char foo[]in fs subsystem]

[<sur...@wi...>]

Files modified=20
fs/binfmt_misc.c
fs/ntfs/attrib.c
fs/ntfs/inode.c
fs/ntfs/mft.c
fs/ntfs/super.c


As suggested by Jeff Garzik,=20
----
>The string form [const] char *foo =3D "blah";
>creates two variables in the final assembly output, a static string,
>and
>a char pointer to the static string. The alternate string form
>[const] char foo[] =3D "blah"; is better because it declares a single
>variable.
------

Cleaned up for the new modifications.=20


Signed-off-by: Surya Prabhakar <sur...@wi...>
---

diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c
index e6f5799..b9ef210 100644
--- a/fs/binfmt_misc.c
+++ b/fs/binfmt_misc.c
@@ -445,7 +445,7 @@ static void entry_status(Node *e, char *page)
 {
 	char *dp;
 	char *status =3D "disabled";
-	const char * flags =3D "flags: ";
+	const char  flags[] =3D "flags: ";
=20
 	if (test_bit(Enabled, &e->flags))
 		status =3D "enabled";
diff --git a/fs/ntfs/attrib.c b/fs/ntfs/attrib.c
index 7659cc1..6f9345d 100644
--- a/fs/ntfs/attrib.c
+++ b/fs/ntfs/attrib.c
@@ -869,7 +869,7 @@ static int ntfs_external_attr_find(const ATTR_TYPE
type,
 	ntfschar *al_name;
 	u32 al_name_len;
 	int err =3D 0;
-	static const char *es =3D " Unmount and run chkdsk.";
+	static const char es[] =3D " Unmount and run chkdsk.";
=20
 	ni =3D ctx->ntfs_ino;
 	base_ni =3D ctx->base_ntfs_ino;
diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c
index f8bf8da..f580025 100644
--- a/fs/ntfs/inode.c
+++ b/fs/ntfs/inode.c
@@ -1871,7 +1871,7 @@ int ntfs_read_inode_mount(struct inode *vi)
 	} else /* if (!err) */ {
 		ATTR_LIST_ENTRY *al_entry, *next_al_entry;
 		u8 *al_end;
-		static const char *es =3D "  Not allowed.  $MFT is corrupt.  "
+		static const char es[] =3D "  Not allowed.  $MFT is corrupt.  "
 				"You should run chkdsk.";
=20
 		ntfs_debug("Attribute list attribute found in $MFT.");
@@ -2333,7 +2333,7 @@ int ntfs_show_options(struct seq_file *sf, struct
vfsmount *mnt)
=20
 #ifdef NTFS_RW
=20
-static const char *es =3D "  Leaving inconsistent metadata.  Unmount and
run "
+static const char es[] =3D "  Leaving inconsistent metadata.  Unmount and
run "
 		"chkdsk.";
=20
 /**
@@ -2369,7 +2369,7 @@ int ntfs_truncate(struct inode *vi)
 	ntfs_attr_search_ctx *ctx;
 	MFT_RECORD *m;
 	ATTR_RECORD *a;
-	const char *te =3D "  Leaving file length out of sync with i_size.";
+	const char te[] =3D "  Leaving file length out of sync with i_size.";
 	int err, mp_size, size_change, alloc_change;
 	u32 attr_len;
=20
diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c
index 2ad5c8b..a560bb0 100644
--- a/fs/ntfs/mft.c
+++ b/fs/ntfs/mft.c
@@ -409,7 +409,7 @@ void __mark_mft_record_dirty(ntfs_inode *ni)
 	__mark_inode_dirty(VFS_I(base_ni), I_DIRTY_SYNC | I_DIRTY_DATASYNC);
 }
=20
-static const char *ntfs_please_email =3D "Please email "
+static const char ntfs_please_email[] =3D "Please email "
 		"lin...@li... and say that you saw "
 		"this message.  Thank you.";
=20
@@ -1106,7 +1106,7 @@ bool ntfs_may_write_mft_record(ntfs_volume *vol,
const unsigned long mft_no,
 	return true;
 }
=20
-static const char *es =3D "  Leaving inconsistent metadata.  Unmount and
run "
+static const char es[] =3D "  Leaving inconsistent metadata.  Unmount and
run "
 		"chkdsk.";
=20
 /**
diff --git a/fs/ntfs/super.c b/fs/ntfs/super.c
index 1594c90..1fe6ddd 100644
--- a/fs/ntfs/super.c
+++ b/fs/ntfs/super.c
@@ -647,7 +647,7 @@ not_ntfs:
 static struct buffer_head *read_ntfs_boot_sector(struct super_block
*sb,
 		const int silent)
 {
-	const char *read_err_str =3D "Unable to read %s boot sector.";
+	const char read_err_str[] =3D "Unable to read %s boot sector.";
 	struct buffer_head *bh_primary, *bh_backup;
 	sector_t nr_blocks =3D NTFS_SB(sb)->nr_blocks;
=20
@@ -1756,9 +1756,9 @@ static bool load_system_files(ntfs_volume *vol)
 #ifdef NTFS_RW
 	/* Get mft mirror inode compare the contents of $MFT and $MFTMirr. */
 	if (!load_and_init_mft_mirror(vol) || !check_mft_mirror(vol)) {
-		static const char *es1 =3D "Failed to load $MFTMirr";
-		static const char *es2 =3D "$MFTMirr does not match $MFT";
-		static const char *es3 =3D ".  Run ntfsfix and/or chkdsk.";
+		static const char es1[] =3D "Failed to load $MFTMirr";
+		static const char es2[] =3D "$MFTMirr does not match $MFT";
+		static const char es3[] =3D ".  Run ntfsfix and/or chkdsk.";
=20
 		/* If a read-write mount, convert it to a read-only mount. */
 		if (!(sb->s_flags & MS_RDONLY)) {
@@ -1880,11 +1880,11 @@ get_ctx_vol_failed:
 #ifdef NTFS_RW
 	/* Make sure that no unsupported volume flags are set. */
 	if (vol->vol_flags & VOLUME_MUST_MOUNT_RO_MASK) {
-		static const char *es1a =3D "Volume is dirty";
-		static const char *es1b =3D "Volume has been modified by chkdsk";
-		static const char *es1c =3D "Volume has unsupported flags set";
-		static const char *es2a =3D ".  Run chkdsk and mount in Windows.";
-		static const char *es2b =3D ".  Mount in Windows.";
+		static const char es1a[] =3D "Volume is dirty";
+		static const char es1b[] =3D "Volume has been modified by chkdsk";
+		static const char es1c[] =3D "Volume has unsupported flags set";
+		static const char es2a[] =3D ".  Run chkdsk and mount in Windows.";
+		static const char es2b[] =3D ".  Mount in Windows.";
 		const char *es1, *es2;
=20
 		es2 =3D es2a;
@@ -1926,9 +1926,9 @@ get_ctx_vol_failed:
 	rp =3D NULL;
 	if (!load_and_check_logfile(vol, &rp) ||
 			!ntfs_is_logfile_clean(vol->logfile_ino, rp)) {
-		static const char *es1a =3D "Failed to load $LogFile";
-		static const char *es1b =3D "$LogFile is not clean";
-		static const char *es2 =3D ".  Mount in Windows.";
+		static const char es1a[] =3D "Failed to load $LogFile";
+		static const char es1b[] =3D "$LogFile is not clean";
+		static const char es2[] =3D ".  Mount in Windows.";
 		const char *es1;
=20
 		es1 =3D !vol->logfile_ino ? es1a : es1b;
@@ -1974,10 +1974,10 @@ get_ctx_vol_failed:
 	 */
 	err =3D check_windows_hibernation_status(vol);
 	if (unlikely(err)) {
-		static const char *es1a =3D "Failed to determine if Windows is "
+		static const char es1a[] =3D "Failed to determine if Windows is "
 				"hibernated";
-		static const char *es1b =3D "Windows is hibernated";
-		static const char *es2 =3D ".  Run chkdsk.";
+		static const char es1b[] =3D "Windows is hibernated";
+		static const char es2[] =3D ".  Run chkdsk.";
 		const char *es1;
=20
 		es1 =3D err < 0 ? es1a : es1b;
@@ -2049,8 +2049,8 @@ get_ctx_vol_failed:
 	/* If (still) a read-write mount, empty the logfile. */
 	if (!(sb->s_flags & MS_RDONLY) &&
 			!ntfs_empty_logfile(vol->logfile_ino)) {
-		static const char *es1 =3D "Failed to empty $LogFile";
-		static const char *es2 =3D ".  Mount in Windows.";
+		static const char es1[] =3D "Failed to empty $LogFile";
+		static const char es2[] =3D ".  Mount in Windows.";
=20
 		/* Convert to a read-only mount. */
 		if (!(vol->on_errors & (ON_ERRORS_REMOUNT_RO |
@@ -2089,8 +2089,8 @@ get_ctx_vol_failed:
 #ifdef NTFS_RW
 	/* Find the quota file, load it if present, and set it up. */
 	if (!load_and_init_quota(vol)) {
-		static const char *es1 =3D "Failed to load $Quota";
-		static const char *es2 =3D ".  Run chkdsk.";
+		static const char es1[] =3D "Failed to load $Quota";
+		static const char es2[] =3D ".  Run chkdsk.";
=20
 		/* If a read-write mount, convert it to a read-only mount. */
 		if (!(sb->s_flags & MS_RDONLY)) {
@@ -2113,8 +2113,8 @@ get_ctx_vol_failed:
 	/* If (still) a read-write mount, mark the quotas out of date. */
 	if (!(sb->s_flags & MS_RDONLY) &&
 			!ntfs_mark_quotas_out_of_date(vol)) {
-		static const char *es1 =3D "Failed to mark quotas out of date";
-		static const char *es2 =3D ".  Run chkdsk.";
+		static const char es1[] =3D "Failed to mark quotas out of date";
+		static const char es2[] =3D ".  Run chkdsk.";
=20
 		/* Convert to a read-only mount. */
 		if (!(vol->on_errors & (ON_ERRORS_REMOUNT_RO |
@@ -2133,8 +2133,8 @@ get_ctx_vol_failed:
 	 * it, and set it up.
 	 */
 	if (!load_and_init_usnjrnl(vol)) {
-		static const char *es1 =3D "Failed to load $UsnJrnl";
-		static const char *es2 =3D ".  Run chkdsk.";
+		static const char es1[] =3D "Failed to load $UsnJrnl";
+		static const char es2[] =3D ".  Run chkdsk.";
=20
 		/* If a read-write mount, convert it to a read-only mount. */
 		if (!(sb->s_flags & MS_RDONLY)) {
@@ -2156,9 +2156,9 @@ get_ctx_vol_failed:
 	}
 	/* If (still) a read-write mount, stamp the transaction log. */
 	if (!(sb->s_flags & MS_RDONLY) && !ntfs_stamp_usnjrnl(vol)) {
-		static const char *es1 =3D "Failed to stamp transaction log "
+		static const char []es1 =3D "Failed to stamp transaction log "
 				"($UsnJrnl)";
-		static const char *es2 =3D ".  Run chkdsk.";
+		static const char []es2 =3D ".  Run chkdsk.";
=20
 		/* Convert to a read-only mount. */
 		if (!(vol->on_errors & (ON_ERRORS_REMOUNT_RO |
@@ -2389,13 +2389,13 @@ static void ntfs_put_super(struct super_block
*sb)
 		mutex_unlock(&vol->mft_ino->i_mutex);
 		write_inode_now(vol->mft_ino, 1);
 		if (!list_empty(&sb->s_dirty)) {
-			static const char *_s1 =3D "inodes";
-			static const char *_s2 =3D "";
+			static const char _s1[] =3D "inodes";
+			static const char _s2[] =3D "";
 			s1 =3D _s1;
 			s2 =3D _s2;
 		} else {
-			static const char *_s1 =3D "mft pages";
-			static const char *_s2 =3D "They have been thrown "
+			static const char _s1[] =3D "mft pages";
+			static const char _s2[] =3D "They have been thrown "
 					"away.  ";
 			s1 =3D _s1;
 			s2 =3D _s2;



--=20
surya.
16/04/2007

[Linux-NTFS-Dev] Trivial patch2:changing const char foo* to const char foo[] in fssubsystem

[<sur...@wi...>]

Files modified=20
fs/binfmt_misc.c
fs/ntfs/attrib.c
fs/ntfs/inode.c
fs/ntfs/mft.c
fs/ntfs/super.c


As suggested by Jeff Garzik,=20
----
>The string form [const] char *foo =3D "blah";
>creates two variables in the final assembly output, a static string,
>and
>a char pointer to the static string. The alternate string form
>[const] char foo[] =3D "blah"; is better because it declares a single
>variable.
------

Cleaned up for the new modifications.=20


Signed-off-by: Surya Prabhakar <sur...@wi...>
---

diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c
index e6f5799..b9ef210 100644
--- a/fs/binfmt_misc.c
+++ b/fs/binfmt_misc.c
@@ -445,7 +445,7 @@ static void entry_status(Node *e, char *page)
 {
 	char *dp;
 	char *status =3D "disabled";
-	const char * flags =3D "flags: ";
+	const char  flags[] =3D "flags: ";
=20
 	if (test_bit(Enabled, &e->flags))
 		status =3D "enabled";
diff --git a/fs/ntfs/attrib.c b/fs/ntfs/attrib.c
index 7659cc1..6f9345d 100644
--- a/fs/ntfs/attrib.c
+++ b/fs/ntfs/attrib.c
@@ -869,7 +869,7 @@ static int ntfs_external_attr_find(const ATTR_TYPE
type,
 	ntfschar *al_name;
 	u32 al_name_len;
 	int err =3D 0;
-	static const char *es =3D " Unmount and run chkdsk.";
+	static const char es[] =3D " Unmount and run chkdsk.";
=20
 	ni =3D ctx->ntfs_ino;
 	base_ni =3D ctx->base_ntfs_ino;
diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c
index f8bf8da..f580025 100644
--- a/fs/ntfs/inode.c
+++ b/fs/ntfs/inode.c
@@ -1871,7 +1871,7 @@ int ntfs_read_inode_mount(struct inode *vi)
 	} else /* if (!err) */ {
 		ATTR_LIST_ENTRY *al_entry, *next_al_entry;
 		u8 *al_end;
-		static const char *es =3D "  Not allowed.  $MFT is corrupt.  "
+		static const char es[] =3D "  Not allowed.  $MFT is corrupt.  "
 				"You should run chkdsk.";
=20
 		ntfs_debug("Attribute list attribute found in $MFT.");
@@ -2333,7 +2333,7 @@ int ntfs_show_options(struct seq_file *sf, struct
vfsmount *mnt)
=20
 #ifdef NTFS_RW
=20
-static const char *es =3D "  Leaving inconsistent metadata.  Unmount and
run "
+static const char es[] =3D "  Leaving inconsistent metadata.  Unmount and
run "
 		"chkdsk.";
=20
 /**
@@ -2369,7 +2369,7 @@ int ntfs_truncate(struct inode *vi)
 	ntfs_attr_search_ctx *ctx;
 	MFT_RECORD *m;
 	ATTR_RECORD *a;
-	const char *te =3D "  Leaving file length out of sync with i_size.";
+	const char te[] =3D "  Leaving file length out of sync with i_size.";
 	int err, mp_size, size_change, alloc_change;
 	u32 attr_len;
=20
diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c
index 2ad5c8b..a560bb0 100644
--- a/fs/ntfs/mft.c
+++ b/fs/ntfs/mft.c
@@ -409,7 +409,7 @@ void __mark_mft_record_dirty(ntfs_inode *ni)
 	__mark_inode_dirty(VFS_I(base_ni), I_DIRTY_SYNC | I_DIRTY_DATASYNC);
 }
=20
-static const char *ntfs_please_email =3D "Please email "
+static const char ntfs_please_email[] =3D "Please email "
 		"lin...@li... and say that you saw "
 		"this message.  Thank you.";
=20
@@ -1106,7 +1106,7 @@ bool ntfs_may_write_mft_record(ntfs_volume *vol,
const unsigned long mft_no,
 	return true;
 }
=20
-static const char *es =3D "  Leaving inconsistent metadata.  Unmount and
run "
+static const char es[] =3D "  Leaving inconsistent metadata.  Unmount and
run "
 		"chkdsk.";
=20
 /**
diff --git a/fs/ntfs/super.c b/fs/ntfs/super.c
index 1594c90..1fe6ddd 100644
--- a/fs/ntfs/super.c
+++ b/fs/ntfs/super.c
@@ -647,7 +647,7 @@ not_ntfs:
 static struct buffer_head *read_ntfs_boot_sector(struct super_block
*sb,
 		const int silent)
 {
-	const char *read_err_str =3D "Unable to read %s boot sector.";
+	const char read_err_str[] =3D "Unable to read %s boot sector.";
 	struct buffer_head *bh_primary, *bh_backup;
 	sector_t nr_blocks =3D NTFS_SB(sb)->nr_blocks;
=20
@@ -1756,9 +1756,9 @@ static bool load_system_files(ntfs_volume *vol)
 #ifdef NTFS_RW
 	/* Get mft mirror inode compare the contents of $MFT and $MFTMirr. */
 	if (!load_and_init_mft_mirror(vol) || !check_mft_mirror(vol)) {
-		static const char *es1 =3D "Failed to load $MFTMirr";
-		static const char *es2 =3D "$MFTMirr does not match $MFT";
-		static const char *es3 =3D ".  Run ntfsfix and/or chkdsk.";
+		static const char es1[] =3D "Failed to load $MFTMirr";
+		static const char es2[] =3D "$MFTMirr does not match $MFT";
+		static const char es3[] =3D ".  Run ntfsfix and/or chkdsk.";
=20
 		/* If a read-write mount, convert it to a read-only mount. */
 		if (!(sb->s_flags & MS_RDONLY)) {
@@ -1880,11 +1880,11 @@ get_ctx_vol_failed:
 #ifdef NTFS_RW
 	/* Make sure that no unsupported volume flags are set. */
 	if (vol->vol_flags & VOLUME_MUST_MOUNT_RO_MASK) {
-		static const char *es1a =3D "Volume is dirty";
-		static const char *es1b =3D "Volume has been modified by chkdsk";
-		static const char *es1c =3D "Volume has unsupported flags set";
-		static const char *es2a =3D ".  Run chkdsk and mount in Windows.";
-		static const char *es2b =3D ".  Mount in Windows.";
+		static const char es1a[] =3D "Volume is dirty";
+		static const char es1b[] =3D "Volume has been modified by chkdsk";
+		static const char es1c[] =3D "Volume has unsupported flags set";
+		static const char es2a[] =3D ".  Run chkdsk and mount in Windows.";
+		static const char es2b[] =3D ".  Mount in Windows.";
 		const char *es1, *es2;
=20
 		es2 =3D es2a;
@@ -1926,9 +1926,9 @@ get_ctx_vol_failed:
 	rp =3D NULL;
 	if (!load_and_check_logfile(vol, &rp) ||
 			!ntfs_is_logfile_clean(vol->logfile_ino, rp)) {
-		static const char *es1a =3D "Failed to load $LogFile";
-		static const char *es1b =3D "$LogFile is not clean";
-		static const char *es2 =3D ".  Mount in Windows.";
+		static const char es1a[] =3D "Failed to load $LogFile";
+		static const char es1b[] =3D "$LogFile is not clean";
+		static const char es2[] =3D ".  Mount in Windows.";
 		const char *es1;
=20
 		es1 =3D !vol->logfile_ino ? es1a : es1b;
@@ -1974,10 +1974,10 @@ get_ctx_vol_failed:
 	 */
 	err =3D check_windows_hibernation_status(vol);
 	if (unlikely(err)) {
-		static const char *es1a =3D "Failed to determine if Windows is "
+		static const char es1a[] =3D "Failed to determine if Windows is "
 				"hibernated";
-		static const char *es1b =3D "Windows is hibernated";
-		static const char *es2 =3D ".  Run chkdsk.";
+		static const char es1b[] =3D "Windows is hibernated";
+		static const char es2[] =3D ".  Run chkdsk.";
 		const char *es1;
=20
 		es1 =3D err < 0 ? es1a : es1b;
@@ -2049,8 +2049,8 @@ get_ctx_vol_failed:
 	/* If (still) a read-write mount, empty the logfile. */
 	if (!(sb->s_flags & MS_RDONLY) &&
 			!ntfs_empty_logfile(vol->logfile_ino)) {
-		static const char *es1 =3D "Failed to empty $LogFile";
-		static const char *es2 =3D ".  Mount in Windows.";
+		static const char es1[] =3D "Failed to empty $LogFile";
+		static const char es2[] =3D ".  Mount in Windows.";
=20
 		/* Convert to a read-only mount. */
 		if (!(vol->on_errors & (ON_ERRORS_REMOUNT_RO |
@@ -2089,8 +2089,8 @@ get_ctx_vol_failed:
 #ifdef NTFS_RW
 	/* Find the quota file, load it if present, and set it up. */
 	if (!load_and_init_quota(vol)) {
-		static const char *es1 =3D "Failed to load $Quota";
-		static const char *es2 =3D ".  Run chkdsk.";
+		static const char es1[] =3D "Failed to load $Quota";
+		static const char es2[] =3D ".  Run chkdsk.";
=20
 		/* If a read-write mount, convert it to a read-only mount. */
 		if (!(sb->s_flags & MS_RDONLY)) {
@@ -2113,8 +2113,8 @@ get_ctx_vol_failed:
 	/* If (still) a read-write mount, mark the quotas out of date. */
 	if (!(sb->s_flags & MS_RDONLY) &&
 			!ntfs_mark_quotas_out_of_date(vol)) {
-		static const char *es1 =3D "Failed to mark quotas out of date";
-		static const char *es2 =3D ".  Run chkdsk.";
+		static const char es1[] =3D "Failed to mark quotas out of date";
+		static const char es2[] =3D ".  Run chkdsk.";
=20
 		/* Convert to a read-only mount. */
 		if (!(vol->on_errors & (ON_ERRORS_REMOUNT_RO |
@@ -2133,8 +2133,8 @@ get_ctx_vol_failed:
 	 * it, and set it up.
 	 */
 	if (!load_and_init_usnjrnl(vol)) {
-		static const char *es1 =3D "Failed to load $UsnJrnl";
-		static const char *es2 =3D ".  Run chkdsk.";
+		static const char es1[] =3D "Failed to load $UsnJrnl";
+		static const char es2[] =3D ".  Run chkdsk.";
=20
 		/* If a read-write mount, convert it to a read-only mount. */
 		if (!(sb->s_flags & MS_RDONLY)) {
@@ -2156,9 +2156,9 @@ get_ctx_vol_failed:
 	}
 	/* If (still) a read-write mount, stamp the transaction log. */
 	if (!(sb->s_flags & MS_RDONLY) && !ntfs_stamp_usnjrnl(vol)) {
-		static const char *es1 =3D "Failed to stamp transaction log "
+		static const char []es1 =3D "Failed to stamp transaction log "
 				"($UsnJrnl)";
-		static const char *es2 =3D ".  Run chkdsk.";
+		static const char []es2 =3D ".  Run chkdsk.";
=20
 		/* Convert to a read-only mount. */
 		if (!(vol->on_errors & (ON_ERRORS_REMOUNT_RO |
@@ -2389,13 +2389,13 @@ static void ntfs_put_super(struct super_block
*sb)
 		mutex_unlock(&vol->mft_ino->i_mutex);
 		write_inode_now(vol->mft_ino, 1);
 		if (!list_empty(&sb->s_dirty)) {
-			static const char *_s1 =3D "inodes";
-			static const char *_s2 =3D "";
+			static const char _s1[] =3D "inodes";
+			static const char _s2[] =3D "";
 			s1 =3D _s1;
 			s2 =3D _s2;
 		} else {
-			static const char *_s1 =3D "mft pages";
-			static const char *_s2 =3D "They have been thrown "
+			static const char _s1[] =3D "mft pages";
+			static const char _s2[] =3D "They have been thrown "
 					"away.  ";
 			s1 =3D _s1;
 			s2 =3D _s2;



--=20
surya.
16/04/2007

Re: [Linux-NTFS-Dev] New LDM Vista patch! - was: Re: LDM vista patch testing

[Kazz <ka...@st...>]

On Wed, 2007-04-11 at 15:46 +0100, Anton Altaparmakov wrote:
> Hi again,
> 
> On 11 Apr 2007, at 14:09, Anton Altaparmakov wrote:
> > On 11 Apr 2007, at 00:18, Kazz wrote:
> >> One more question since if anyone can answer it, you probably can.  I
> >> planned to use windows software RAID 0 and RAID 5 on different
> >> partitions since I'd read that Linux was capable of using those  
> >> too, but
> >> when I finally went to set it up, I discovered that vista "ultimate"
> >> wasn't quite so ultimate when it came to software RAID, and it had no
> >> options for RAID 5 or even RAID 1.
> >
> > Really?  /me tries it..  Yes you are right!  Even using the command  
> > line I get (note disk 1 is contains a single simple volume):
> >
> > <quote>
> > DISKPART> select disk=1
> >
> > Disk 1 is now the selected disk.
> >
> > DISKPART> add disk=2
> >
> > The command you selected is not available with this version of  
> > Windows.
> > </quote>
> >
> > And I get the same error message trying to create a RAID 5 set...
> >
> > However, I thought of a cool trick.  Boot with the Vista  
> > installation CD and then when you have started the installer (it  
> > does not need to be installing yet, the CD just needs to have  
> > booted into the full installation environment, i.e. you can for  
> > example do it at the screen where it asks to enter the CD Key)  
> > press SHIFT+F10.  You then get a command prompt up running with  
> > full administrative privileges.  Then type "diskpart" and you are  
> > now running diskpart command line utility.  You can now create both  
> > RAID 5 and Mirror sets!  (-:
> [snip]
> > I am doing this right now to see whether it works...  I will let  
> > you know when I have finished whether it worked or not...
> 
> Having finished installing Vista Ultimate in VMware Fusion I have to  
> say my trick was not so clever...  It does not work!  )-:
> 

It was still clever, they just must have realized that some people would
be clever, and they can't have people using unauthorized functionality
that's already built in, now can they? ;)

> Booting into Vista Ultimate it showed the disks as being a "foreign  
> diskset" and all I could do was import the diskset.  Once I had done  
> that both the mirror and raid5 arrays showed up as "Failed" and  
> neither in the GUI nor in diskpart could I do anything to bring them  
> back online...
> 
> So whilst you can create them with the installer (and use them! - I  
> also formatted them in the installer in diskpart to NTFS and even  
> copied some files from the install dvd onto the arrays...) Vista  
> Ultimate cannot use them.  )-:
> 
> How very annoying of it!
> 

Yes it is... :\  That's what I was afraid of, and it's the same kind of
thing that it did when I made dynamic disks in xp and tried to read
them.

> I wonder which version of Vista is needed to use Fault Tolerant disks  
> sets...

They seem to reserve that for servers.  I guess they need something to
make people buy those versions, and unnecessarily reducing functionality
in other versions is a common way for them to do it.  The "home"
versions for example are relatively crippled even compared to
"ultimate" (which is less "ultimate" than we are led to believe).

You may not be interested enough in this to try, but I read that in xp
there was a way to copy certain files from the server versions into the
other versions, I assume replacing some or all of the disk access
drivers/programs, and that would allow it to both create and use fault
tolerant arrays.

Thanks for giving it a shot anyway,
	Kazz

Re: [Linux-NTFS-Dev] New LDM Vista patch! - was: Re: LDM vista patch testing

[Anton A. <ai...@ca...>]

Hi again,

On 11 Apr 2007, at 14:09, Anton Altaparmakov wrote:
> On 11 Apr 2007, at 00:18, Kazz wrote:
>> One more question since if anyone can answer it, you probably can.  I
>> planned to use windows software RAID 0 and RAID 5 on different
>> partitions since I'd read that Linux was capable of using those  
>> too, but
>> when I finally went to set it up, I discovered that vista "ultimate"
>> wasn't quite so ultimate when it came to software RAID, and it had no
>> options for RAID 5 or even RAID 1.
>
> Really?  /me tries it..  Yes you are right!  Even using the command  
> line I get (note disk 1 is contains a single simple volume):
>
> <quote>
> DISKPART> select disk=1
>
> Disk 1 is now the selected disk.
>
> DISKPART> add disk=2
>
> The command you selected is not available with this version of  
> Windows.
> </quote>
>
> And I get the same error message trying to create a RAID 5 set...
>
> However, I thought of a cool trick.  Boot with the Vista  
> installation CD and then when you have started the installer (it  
> does not need to be installing yet, the CD just needs to have  
> booted into the full installation environment, i.e. you can for  
> example do it at the screen where it asks to enter the CD Key)  
> press SHIFT+F10.  You then get a command prompt up running with  
> full administrative privileges.  Then type "diskpart" and you are  
> now running diskpart command line utility.  You can now create both  
> RAID 5 and Mirror sets!  (-:
[snip]
> I am doing this right now to see whether it works...  I will let  
> you know when I have finished whether it worked or not...

Having finished installing Vista Ultimate in VMware Fusion I have to  
say my trick was not so clever...  It does not work!  )-:

Booting into Vista Ultimate it showed the disks as being a "foreign  
diskset" and all I could do was import the diskset.  Once I had done  
that both the mirror and raid5 arrays showed up as "Failed" and  
neither in the GUI nor in diskpart could I do anything to bring them  
back online...

So whilst you can create them with the installer (and use them! - I  
also formatted them in the installer in diskpart to NTFS and even  
copied some files from the install dvd onto the arrays...) Vista  
Ultimate cannot use them.  )-:

How very annoying of it!

I wonder which version of Vista is needed to use Fault Tolerant disks  
sets...

Best regards,

	Anton
-- 
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer, http://www.linux-ntfs.org/

Re: [Linux-NTFS-Dev] New LDM Vista patch! - was: Re: LDM vista patch testing

[Anton A. <ai...@ca...>]

Hi,

On 11 Apr 2007, at 00:18, Kazz wrote:
> Hey, sorry I didn't respond earlier, but I've been pretty busy lately
> and I just got this.

No problem.

> Yes, I am still interested, however I was forced to change my planned
> setup somewhat due to LDM not being a working option at the time, so I
> would have to risk making all of my partitions (windows and Linux)
> inaccessible to really try it right now.
>
> I'm not sure when I'll get a chance, but if I can clear off another  
> hard
> drive to test it on then I'll give it a shot as soon as I can.
>
> Just to be sure, if it does work on one drive (probably an old
> PATA/100), there should be no problems with my other SATA2 drives,
> right?

Yes.  The type of disk does not matter at all.

> The way I understand it, it reads the partition table or it  
> doesn't, and
> there shouldn't be a reason for it to fail on one if it works on
> another, ...

Correct.

> One more question since if anyone can answer it, you probably can.  I
> planned to use windows software RAID 0 and RAID 5 on different
> partitions since I'd read that Linux was capable of using those  
> too, but
> when I finally went to set it up, I discovered that vista "ultimate"
> wasn't quite so ultimate when it came to software RAID, and it had no
> options for RAID 5 or even RAID 1.

Really?  /me tries it..  Yes you are right!  Even using the command  
line I get (note disk 1 is contains a single simple volume):

<quote>
DISKPART> select disk=1

Disk 1 is now the selected disk.

DISKPART> add disk=2

The command you selected is not available with this version of Windows.
</quote>

And I get the same error message trying to create a RAID 5 set...

However, I thought of a cool trick.  Boot with the Vista installation  
CD and then when you have started the installer (it does not need to  
be installing yet, the CD just needs to have booted into the full  
installation environment, i.e. you can for example do it at the  
screen where it asks to enter the CD Key) press SHIFT+F10.  You then  
get a command prompt up running with full administrative privileges.   
Then type "diskpart" and you are now running diskpart command line  
utility.  You can now create both RAID 5 and Mirror sets!  (-:

Use "help" and "help create volume raid" and "help create volume  
simple" and "help select" and help add" to see how to do it...

Basically for a raid 5 you need to do for example:

create volume raid disk=1,2,3,...

Where 1,2,3,... are the disk numbers of the dynamic disks you want to  
use.

To see which disk has which number use "list disk"...

If the disks are not dynamic yet use "select disk=1" where 1 is the  
disk you want then "convert dynamic" to convert the select disk to be  
a dynamic disk.

To create a mirror you need to first create a simple volume using  
"create volume simple disk=1" where 1 is the disk to do it on, then  
to add a dynamic disk with sufficient free space to it as a mirror  
using the command "select volume=1" followed by "add disk=2" where 1  
is the existing simple volume (use "list volume" to see which number  
you need) and 2 is the disk which you want to have the mirror created  
on.

When you are done, type "exit" to exit diskpart then again "exit" to  
exit the command prompt.  Then shutdown your computer and start it  
from the HD and presto you are running Vista Ultimate with a RAID5  
and a Mirror set.  (-:

I am doing this right now to see whether it works...  I will let you  
know when I have finished whether it worked or not...

> Since they can be read in Linux, do you know if there is any way to
> *create* those types of arrays in Linux (or anywhere else) and get
> windows to use them?  If not I'll just use RAID 0 in windows, but I am
> very annoyed (and unfortunately not surprised) that they are keeping
> redundant arrays from home users who probably need it the most since
> servers are usually backed up externally and most home computers  
> aren't.

No there is no way that I know to write to LDM from Linux.

You should remember that redundant RAID is no substitute for back  
up!  We have had many RAID-5 sets fail because 2 disks failed at the  
same time for example...

> Enough ranting I guess, I'll let you know when I have a chance to test
> the patch, and please let me know if someone else tests it first. :)

Ok, great, thanks!  And will do.

Best regards,

	Anton

> Thanks again!
>   Kazz
>
> On Fri, 2007-03-30 at 22:24 +0100, Anton Altaparmakov wrote:
>> Hi,
>>
>> Are you still interested in getting Vista LDM to work on Linux?
>>
>> I have investigated the Vista LDM and have come up with a new patch
>> that may well work.  It is attached to this email.  To apply it save
>> the patch in your home directory (or wherever) then "cd" into the
>> kernel sources directory (the patch is against the latest 2.6.20
>> kernel, actually against the current kernel git tree but I have
>> tested that it applies cleanly to the 2.6.20 kernel also) and then  
>> do:
>>
>> patch -p1 < ~/ldm-vista.diff
>>
>> Obviously if it is not in your home directory you need to replace
>> "~/" with the path to the patch...
>>
>> You should see this output:
>>
>> $ patch -p1 < ~/ldm-vista.diff patching file Documentation/ldm.txt
>> patching file fs/partitions/Kconfig
>> patching file fs/partitions/ldm.c
>> patching file fs/partitions/ldm.h
>>
>> Then recompile the kernel with LDM enabled and try it again with the
>> Vista LDM.  Hopefully it will now work.
>>
>> If you (or anyone) tries it please let me know how it works out.  If
>> it works I will submit it to Linus for inclusion into mainline...
>>
>> Best regards,
>>
>> 	Anton

-- 
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer, http://www.linux-ntfs.org/

Re: [Linux-NTFS-Dev] New LDM Vista patch! - was: Re: LDM vista patch testing

[Kazz <ka...@st...>]

Hey, sorry I didn't respond earlier, but I've been pretty busy lately
and I just got this.

Yes, I am still interested, however I was forced to change my planned
setup somewhat due to LDM not being a working option at the time, so I
would have to risk making all of my partitions (windows and Linux)
inaccessible to really try it right now.

I'm not sure when I'll get a chance, but if I can clear off another hard
drive to test it on then I'll give it a shot as soon as I can.

Just to be sure, if it does work on one drive (probably an old
PATA/100), there should be no problems with my other SATA2 drives,
right?

The way I understand it, it reads the partition table or it doesn't, and
there shouldn't be a reason for it to fail on one if it works on
another, but if I'm wrong I don't want to risk having to go through
something weird and scary to try to convert them back to regular
partition tables before I'm pretty sure I won't have to.

One more question since if anyone can answer it, you probably can.  I
planned to use windows software RAID 0 and RAID 5 on different
partitions since I'd read that Linux was capable of using those too, but
when I finally went to set it up, I discovered that vista "ultimate"
wasn't quite so ultimate when it came to software RAID, and it had no
options for RAID 5 or even RAID 1.

Since they can be read in Linux, do you know if there is any way to
*create* those types of arrays in Linux (or anywhere else) and get
windows to use them?  If not I'll just use RAID 0 in windows, but I am
very annoyed (and unfortunately not surprised) that they are keeping
redundant arrays from home users who probably need it the most since
servers are usually backed up externally and most home computers aren't.

Enough ranting I guess, I'll let you know when I have a chance to test
the patch, and please let me know if someone else tests it first. :)

Thanks again!
  Kazz

On Fri, 2007-03-30 at 22:24 +0100, Anton Altaparmakov wrote:
> Hi,
> 
> Are you still interested in getting Vista LDM to work on Linux?
> 
> I have investigated the Vista LDM and have come up with a new patch  
> that may well work.  It is attached to this email.  To apply it save  
> the patch in your home directory (or wherever) then "cd" into the  
> kernel sources directory (the patch is against the latest 2.6.20  
> kernel, actually against the current kernel git tree but I have  
> tested that it applies cleanly to the 2.6.20 kernel also) and then do:
> 
> patch -p1 < ~/ldm-vista.diff
> 
> Obviously if it is not in your home directory you need to replace  
> "~/" with the path to the patch...
> 
> You should see this output:
> 
> $ patch -p1 < ~/ldm-vista.diff patching file Documentation/ldm.txt
> patching file fs/partitions/Kconfig
> patching file fs/partitions/ldm.c
> patching file fs/partitions/ldm.h
> 
> Then recompile the kernel with LDM enabled and try it again with the  
> Vista LDM.  Hopefully it will now work.
> 
> If you (or anyone) tries it please let me know how it works out.  If  
> it works I will submit it to Linus for inclusion into mainline...
> 
> Best regards,
> 
> 	Anton

Re: [Linux-NTFS-Dev] NTFS Drive Can't Be Read/Mounted in 2.6 Kernel

[Anton A. <ai...@ca...>]

Hi,

On 4 Apr 2007, at 16:10, Fresh Fish wrote:
> It worked! I can now enjoy the new 2.6 kernel! Thanks so much!

You are welcome.  (-:

> BTW if you don't mind, how did you know that my hdb disk uses  
> OnTrackDisk Manager? Also how did you find the solution?
> I just want to learn how I'm not experienced at Linux and I'm  
> trying to learn.

 From your dmesg output.  Here are the relevant line:

hdb: hdb1[DM]

I searched linux-2.6/fs/partitions/* for the string "DM" and I found  
it in those lines:



                 if (SYS_IND(p) == DM6_PARTITION)
                         printk("[DM]");

in the file msdos.c.

I then looked up DM6_PARTITION in include/linux/* and found it in  
those lines:

         DM6_PARTITION = 0x54,           /* has DDO: use xlated geom  
& offset */

in the file genhd.h.

I then used Google to search for "DDO DM6" and the very first hit was  
this one:

http://kerneltrap.org/node/3417

Which explained it all...

Also I found the large disk HOWTO which also explained it:

http://tldp.org/HOWTO/Large-Disk-HOWTO-8.html#ss8.7

Finally in the kernel sources, in Documentation/ide.txt you also find  
the string "DM" when describing the remap63 option:

"hdx=remap63"          : remap the drive: add 63 to all sector numbers
                           (for DM OnTrack)

So lots of places to find the info if you know what you are looking  
for.  (-:

Best regards,

	Anton
-- 
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer, http://www.linux-ntfs.org/

Re: [Linux-NTFS-Dev] ntfs-3g problem

[Szakacsits S. <sz...@si...>]

Hi,

On Mon, 2 Apr 2007, Udrescu Gabriel sau G@By wrote:

> 2 days ago I hade some power failure and my ubuntu dapper drake box
> rebooted. I use a ntfs formatted partition and after the problems were
> solved i've switched on my PC again. But surprize, the partition that is
> mounted via ntfs-3g was read-only. i've tried to remount it, but i couldn't,
> because i've recived some nasty errors =>
> 
> root@pc1:/media/magazie/Downloads# umount /dev/hda5
> umount: /media/magazie: device is busy
> umount: /media/magazie: device is busy

This is not nasty but totally normal if files are in use from the disk.
Use fuser or lsof to find out what softwares are using the volume (file 
managers and shells most typically).
 
> root@pc1:/media/magazie/Downloads# fusermount -u /dev/hda5
> fusermount: entry for /dev/hda5 not found in /etc/mtab
> (even if there is an entry of /dev/hda5 in /etc/mtab, you can see below)

fusermount is buggy, just forget it completely. It will be removed from the 
FUSE package in the future.
 
> root@pc1:/var/log# cat syslog|grep ntfs
> Apr  1 21:27:48 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
> bitmap but is marked used itself.  Corrupt filesystem or library bug!  Run
> chkdsk immediately!
> Apr  1 21:27:48 pc1 ntfs-3g[3472]: Could not allocate new MFT record

This is a consequence of the power failure. It's repeated in your log 
whenever you wanted to create a new file.

> trying to fix this problem with "chkdsk", i've recived this message :
> root@pc1:/var/log# chkdsk
> bash: chkdsk: command not found
> isn't chkdsk included in ntfs-3g? or it's under a different name ?

Chkdsk is the Windows file system checker:
http://www.microsoft.com/resources/documentation/windows/xp/all/proddocs/en-us/chkdsk.mspx

But I very seriously suggest to make a full backup of you ntfs partition 
with your favourite backup software or ntfsclone first before you would use 
it.

> It might be useful for you if I tell you I'm not using the latest ntfs-3g
> version, but the older 1.0 release. I'm a bit afraid to uninstall this
> version and install the newest release (on March 31, 2007) because of that
> problems. 

These issues are file system independent. If you have a power failure then 
you always lose data if write was ongoing. Some file system can restore a 
consistent state (not the data of course, only the fs consistency), some 
don't.

> Really I have important data on the ntfs partition and I don't want to 
> loose them

It's maybe late. You must always keep backups if you value your data,
especially if power outages are expected without a well configured UPS. 
Plenty disks die every seconds in the world and only those believe it won't 
happen with them who never lost data -- but only until they do (hardware 
problem, software bug, user error, etc).
 
Regards,
	Szaka

Re: [Linux-NTFS-Dev] NTFS Drive Can't Be Read/Mounted in 2.6 Kernel

[Anton A. <ai...@ca...>]

Hi,

On 3 Apr 2007, at 10:02, Fresh Fish wrote:
> Hi there! This is fishda from the linux ntfs forums and attached is  
> the dmesg of my 2.6 kernel.
> If you're wondering here's a link to my post in which you've replied:
> http://forum.linux-ntfs.org/viewtopic.php?t=642
>
> The drive is really a NTFS one. It's been read in the 2.4 kernel  
> and in windows xp.
>
> I hope you can help me because it's been a long time problem and  
> I'd really like to access my files.

Your dmesg output shows that your hdb disk uses the OnTrackDisk  
Manager on the disk.  2.6 kernels do not support this kind of disk  
manager any more automatically.

According to the Large disk howto (http://tldp.org/HOWTO/html_single/ 
Large-Disk-HOWTO/#s8) you need to boot with the "hda=remap63" option  
added to your kernel command line to get your disk to work.

This is nothing to do with the NTFS driver but with the kernel's  
support for your disk...

Best regards,

	Anton
-- 
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer, http://www.linux-ntfs.org/

[Linux-NTFS-Dev] ntfs-3g problem

[Udrescu G. s. G@B. <gab...@gm...>]

Hello!
2 days ago I hade some power failure and my ubuntu dapper drake box
rebooted. I use a ntfs formatted partition and after the problems were
solved i've switched on my PC again. But surprize, the partition that is
mounted via ntfs-3g was read-only. i've tried to remount it, but i couldn't,
because i've recived some nasty errors =>

root@pc1:/media/magazie/Downloads# umount /dev/hda5
umount: /media/magazie: device is busy
umount: /media/magazie: device is busy

root@pc1:/media/magazie/Downloads# fusermount -u /dev/hda5
fusermount: entry for /dev/hda5 not found in /etc/mtab
(even if there is an entry of /dev/hda5 in /etc/mtab, you can see below)

/dev/hda2 / ext3 rw,errors=remount-ro 0 0
proc /proc proc rw 0 0
/sys /sys sysfs rw 0 0
varrun /var/run tmpfs rw 0 0
varlock /var/lock tmpfs rw 0 0
procbususb /proc/bus/usb usbfs rw 0 0
udev /dev tmpfs rw 0 0
devpts /dev/pts devpts rw,gid=5,mode=620 0 0
devshm /dev/shm tmpfs rw 0 0
lrm /lib/modules/2.6.15-27-386/volatile tmpfs rw 0 0
/dev/hda3 /home ext3 rw 0 0
/dev/hda5 /media/magazie fuse
rw,nosuid,nodev,noatime,allow_other,default_permi
binfmt_misc /proc/sys/fs/binfmt_misc binfmt_misc rw 0 0

then, my first thought was to see the kernel logs, so i've tried cat
/var/log/syslog|grep ntfs and i've found this =>

root@pc1:/var/log# cat syslog|grep ntfs
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im
mediately!
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im
mediately!
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im
mediately!
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im
mediately!
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im
mediately!
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im
mediately!
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im
mediately!
Apr  1 21:27:48 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:27:52 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im
mediately!
Apr  1 21:27:52 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:27:55 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im
mediately!
Apr  1 21:27:55 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:27:59 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im
mediately!
Apr  1 21:27:59 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:28:29 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im
mediately!
Apr  1 21:28:29 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:51:51 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm
ap but is marked used itself.  Corrupt filesystem or library bug!  Run
chkdsk im                              mediately!
Apr  1 21:51:51 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:51:51 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm                              ap but is marked used itself.  Corrupt
filesystem or library bug!  Run chkdsk im
mediately!
Apr  1 21:51:51 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 21:51:51 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm                              ap but is marked used itself.  Corrupt
filesystem or library bug!  Run chkdsk im
mediately!
Apr  1 21:51:51 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 22:57:00 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm                              ap but is marked used itself.  Corrupt
filesystem or library bug!  Run chkdsk im
mediately!
Apr  1 22:57:00 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 22:57:42 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm                              ap but is marked used itself.  Corrupt
filesystem or library bug!  Run chkdsk im
mediately!
Apr  1 22:57:42 pc1 ntfs-3g[3472]: Could not allocate new MFT record
Apr  1 22:58:30 pc1 ntfs-3g[3472]: Mft record 0xa4 was marked unused in mft
bitm                              ap but is marked used itself.  Corrupt
filesystem or library bug!  Run chkdsk im
mediately!
Apr  1 22:58:30 pc1 ntfs-3g[3472]: Could not allocate new MFT record

trying to fix this problem with "chkdsk", i've recived this message :
root@pc1:/var/log# chkdsk
bash: chkdsk: command not found
isn't chkdsk included in ntfs-3g? or it's under a different name ?

It might be useful for you if I tell you I'm not using the latest ntfs-3g
version, but the older 1.0 release. I'm a bit afraid to uninstall this
version and install the newest release (on March 31, 2007) because of that
problems. Really I have important data on the ntfs partition and I don't
want to loose them

Big thanks in advance,
Your ntfs-3g Romanian user,
Udrescu Gabriel
-- 
http://eurorevelion.googlepages.com
http://afaith.blogspot.com
http://www.campkalivas.lx.ro
__________________________________
site-uri facute de  mine / site that i have created