fsck.reiser4 can't fix 1 fixable corruption

fsck.reiser4 can't fix 1 fixable corruption

[Raymond A. Meijer]

Hello,

I'm running Debian with these packages installed:

ii  reiser4progs                  0.5.3-1
ii  libc6                         2.3.2.ds1-11
ii  libncurses5                   5.4-3
ii  libreadline4                  4.3-10
ii  libuuid1                      1.35-4

The kernel is a kernel.org 2.6.5 kernel patched with the 
http://thebsh.namesys.com/snapshots/LATEST/all.diff.gz patch and an 
additional patch that was sent to this list (Subject: "Re: snapshot 
2004.03.26 + linux 2.6.5 = build error").

I decided to fsck my Reiser4 backup partition, which happens to be on an LVM2 
Logical Volume:

  --- Logical volume ---
  LV Name                /dev/sys/backup
  VG Name                sys
  LV UUID                boifcf-z605-EHXl-0zi7-CdW6-vpwk-35XsJ2
  LV Write Access        read/write
  LV Status              available
  # open                 0
  LV Size                8.00 GB
  Current LE             256
  Segments               2
  Allocation             next free (default)
  Read ahead sectors     0
  Block device           254:7

fsck failed however, with the following warning (sorry about the line 
wrapping):


==============================================================================
[root@rmeijer-noc ~] # fsck /dev/sys/backup
fsck 1.35 (28-Feb-2004)
*******************************************************************
This is an EXPERIMENTAL version of fsck.reiser4. Read REAMDE first.
*******************************************************************

Fscking the /dev/sys/backup block device.
Will check the consistency of the Reiser4 SuperBlock.
Will check the consistency of the Reiser4 FileSystem.
Continue?
(Yes/No): Yes
***** Openning the fs.
Reiser4 journal (journal40) on /dev/sys/backup: 0 transactions replaied of the 
total 0 blocks.
Reiser4 fs was detected on the /dev/sys/backup.
Master super block (16):
magic:          ReIsEr4
blksize:        4096
format:         0x0 (format40)
uuid:           f6a57152-e546-4a2a-b833-c0bce2a62caa
label:          <none>

Format super block (17):
plugin:         format40
description:    Disk-format for reiser4, ver. 0.5.3
magic:          ReIsEr40FoRmAt
flushes:        0
mkfs id:        0x6909edad
blocks:         2097152
free blocks:    420905
root block:     663
tail policy:    0x4 (smart)
next oid:       0x6f6ae
file count:     262871
tree height:    4
key policy:     LARGE

***** Tree Traverse Pass: scanning the reiser4 internal tree.
        Read nodes 141159
        Nodes left in the tree 141159
                Leaves of them 138592, Twigs of them 2509
        Time interval: Tue Apr 20 17:00:59 2004 - Tue Apr 20 17:03:13 2004
***** TwigScan Pass: checking extent pointers of all twigs.
        Read twigs 2509
        Time interval: Tue Apr 20 17:03:13 2004 - Tue Apr 20 17:03:20 2004
***** Semantic Traverse Pass: reiser4 semantic tree checking.
Error: Node (1444613), item (2): StatData of the file [2a:0:ffff] has the 
wrong bytes (0), Should be (188). Plugin (stat40).
        Time interval: Tue Apr 20 17:03:20 2004 - Tue Apr 20 17:07:02 2004
***** fsck.reiser4 finished at Tue Apr 20 17:07:02 2004
Closing fs...done

1 fixable corruptions were detected in the FileSystem. Run with --fix option 
to fix them.
==============================================================================


But running fsck.reiser4 with (among others) the --fix option results in the 
following:


==============================================================================
[root@rmeijer-noc ~] # fsck.reiser4 -v --build-sb --build-fs 
--fix /dev/sys/backup
*******************************************************************
This is an EXPERIMENTAL version of fsck.reiser4. Read REAMDE first.
*******************************************************************

Fscking the /dev/sys/backup block device.
Will build the Reiser4 SuperBlock.
Will build the Reiser4 FileSystem.
Continue?
(Yes/No): Yes
***** Openning the fs.
Reiser4 journal (journal40) on /dev/sys/backup: 0 transactions replaied of the 
total 0 blocks.
Reiser4 fs was detected on the /dev/sys/backup.
Master super block (16):
magic:          ReIsEr4
blksize:        4096
format:         0x0 (format40)
uuid:           f6a57152-e546-4a2a-b833-c0bce2a62caa
label:          <none>

Format super block (17):
plugin:         format40
description:    Disk-format for reiser4, ver. 0.5.3
magic:          ReIsEr40FoRmAt
flushes:        0
mkfs id:        0x6909edad
blocks:         2097152
free blocks:    420905
root block:     663
tail policy:    0x4 (smart)
next oid:       0x6f6ae
file count:     262871
tree height:    4
key policy:     LARGE

***** Tree Traverse Pass: scanning the reiser4 internal tree.
        Read nodes 141159
        Nodes left in the tree 141159
                Leaves of them 138592, Twigs of them 2509
        Time interval: Tue Apr 20 16:46:16 2004 - Tue Apr 20 16:48:19 2004
***** TwigScan Pass: checking extent pointers of all twigs.
        Read twigs 2509
        Time interval: Tue Apr 20 16:48:19 2004 - Tue Apr 20 16:48:27 2004
***** DiskScan Pass: scanning the partition for unconnected nodes.
        Read nodes 0
        Good nodes 0
                Leaves of them 0, Twigs of them 0
        Time interval: Tue Apr 20 16:48:27 2004 - Tue Apr 20 16:48:27 2004
***** TwigScan Pass: checking extent pointers of all twigs.
        Read twigs 0
        Time interval: Tue Apr 20 16:48:27 2004 - Tue Apr 20 16:48:27 2004
***** AddMissing Pass: inserting unconnected nodes into the tree.
        Twigs: read 0, inserted 0, by items 0
        Leaves: read 0, inserted 0, by items 0
        Time interval: Tue Apr 20 16:48:27 2004 - Tue Apr 20 16:48:27 2004
***** Semantic Traverse Pass: reiser4 semantic tree checking.
Error: Node (1444613), item (2): StatData of the file [2a:0:ffff] has the 
wrong bytes (0), Fixed to (188). Plugin (stat40).
        Time interval: Tue Apr 20 16:48:27 2004 - Tue Apr 20 16:55:48 2004
***** Cleanup Pass: cleaning up the reiser4 storage tree.
        Removed items 0
        Time interval: Tue Apr 20 16:55:48 2004 - Tue Apr 20 16:59:09 2004
***** fsck.reiser4 finished at Tue Apr 20 16:59:09 2004
Closing fs...done

No corruption found.
==============================================================================


Looks good, but when I run "fsck.reiser4 /dev/sys/backup" again, it tells me 
that it has found 1 fixable corruption and that I need to run it with the 
--fix parameter...

What can I do to fix this? Or isn't my filesystem broken in the first place? 
Could it be a problem with fsck.reiser4?

If you need more information, please let me know.


Thanks,

Ray

P.S. How about all these spelling errors in fsck's output? ;)

Re: fsck.reiser4 can't fix 1 fixable corruption

[Sander Sweers]

Raymond A. Meijer wrote:

>Hello,
>
>I'm running Debian with these packages installed:
>
>ii  reiser4progs                  0.5.3-1
>ii  libc6                         2.3.2.ds1-11
>ii  libncurses5                   5.4-3
>ii  libreadline4                  4.3-10
>ii  libuuid1                      1.35-4
>
>*snip*
>
>==============================================================================
>[root@rmeijer-noc ~] # fsck.reiser4 -v --build-sb --build-fs 
>--fix /dev/sys/backup
>*******************************************************************
>This is an EXPERIMENTAL version of fsck.reiser4. Read REAMDE first.
>*******************************************************************
>
>Fscking the /dev/sys/backup block device.
>Will build the Reiser4 SuperBlock.
>Will build the Reiser4 FileSystem.
>Continue?
>(Yes/No): Yes
>***** Openning the fs.
>Reiser4 journal (journal40) on /dev/sys/backup: 0 transactions replaied of the 
>total 0 blocks.
>Reiser4 fs was detected on the /dev/sys/backup.
>Master super block (16):
>magic:          ReIsEr4
>blksize:        4096
>format:         0x0 (format40)
>uuid:           f6a57152-e546-4a2a-b833-c0bce2a62caa
>label:          <none>
>
>Format super block (17):
>plugin:         format40
>description:    Disk-format for reiser4, ver. 0.5.3
>magic:          ReIsEr40FoRmAt
>flushes:        0
>mkfs id:        0x6909edad
>blocks:         2097152
>free blocks:    420905
>root block:     663
>tail policy:    0x4 (smart)
>next oid:       0x6f6ae
>file count:     262871
>tree height:    4
>key policy:     LARGE
>
>***** Tree Traverse Pass: scanning the reiser4 internal tree.
>        Read nodes 141159
>        Nodes left in the tree 141159
>                Leaves of them 138592, Twigs of them 2509
>        Time interval: Tue Apr 20 16:46:16 2004 - Tue Apr 20 16:48:19 2004
>***** TwigScan Pass: checking extent pointers of all twigs.
>        Read twigs 2509
>        Time interval: Tue Apr 20 16:48:19 2004 - Tue Apr 20 16:48:27 2004
>***** DiskScan Pass: scanning the partition for unconnected nodes.
>        Read nodes 0
>        Good nodes 0
>                Leaves of them 0, Twigs of them 0
>        Time interval: Tue Apr 20 16:48:27 2004 - Tue Apr 20 16:48:27 2004
>***** TwigScan Pass: checking extent pointers of all twigs.
>        Read twigs 0
>        Time interval: Tue Apr 20 16:48:27 2004 - Tue Apr 20 16:48:27 2004
>***** AddMissing Pass: inserting unconnected nodes into the tree.
>        Twigs: read 0, inserted 0, by items 0
>        Leaves: read 0, inserted 0, by items 0
>        Time interval: Tue Apr 20 16:48:27 2004 - Tue Apr 20 16:48:27 2004
>***** Semantic Traverse Pass: reiser4 semantic tree checking.
>Error: Node (1444613), item (2): StatData of the file [2a:0:ffff] has the 
>wrong bytes (0), Fixed to (188). Plugin (stat40).
>        Time interval: Tue Apr 20 16:48:27 2004 - Tue Apr 20 16:55:48 2004
>***** Cleanup Pass: cleaning up the reiser4 storage tree.
>        Removed items 0
>        Time interval: Tue Apr 20 16:55:48 2004 - Tue Apr 20 16:59:09 2004
>***** fsck.reiser4 finished at Tue Apr 20 16:59:09 2004
>Closing fs...done
>
>No corruption found.
>==============================================================================
>
>
>Looks good, but when I run "fsck.reiser4 /dev/sys/backup" again, it tells me 
>that it has found 1 fixable corruption and that I need to run it with the 
>--fix parameter...
>
>What can I do to fix this? Or isn't my filesystem broken in the first place? 
>Could it be a problem with fsck.reiser4?
>
>If you need more information, please let me know.
>
>
>Thanks,
>
>Ray
>
>P.S. How about all these spelling errors in fsck's output? ;)
>
>
>  
>

------------------------------------------------------------------------
Error: Node (3978247), item (1): StatData of the file [2a:0:ffff] has 
the wrong
bytes (0), Should be (185). Plugin (stat40).
        Time interval: Tue Apr 20 20:46:31 2004 - Tue Apr 20 20:46:31 
2004   
***** fsck.reiser4 finished at Tue Apr 20 20:46:31 2004
Closing fs...done
------------------------------------------------------------------------

Same problem here, same kernel and reiser4progs but this is not on lvm 
but a 25G partition.

The way the 2 errors look the same is scary.

Sander

Re: fsck.reiser4 can't fix 1 fixable corruption

[Raymond A. Meijer]

On Tue 20 April 2004 21:54, Sander Sweers wrote:

> >Error: Node (1444613), item (2): StatData of the file [2a:0:ffff] has the 
> >wrong bytes (0), Should be (188). Plugin (stat40).
> >        Time interval: Tue Apr 20 17:03:20 2004 - Tue Apr 20 17:07:02 2004
> >***** fsck.reiser4 finished at Tue Apr 20 17:07:02 2004
> >Closing fs...done

> ------------------------------------------------------------------------
> Error: Node (3978247), item (1): StatData of the file [2a:0:ffff] has
> the wrong bytes (0), Should be (185). Plugin (stat40).
>         Time interval: Tue Apr 20 20:46:31 2004 - Tue Apr 20 20:46:31 2004
> ***** fsck.reiser4 finished at Tue Apr 20 20:46:31 2004
> Closing fs...done
> ------------------------------------------------------------------------

> Same problem here, same kernel and reiser4progs but this is not on lvm
> but a 25G partition.

> The way the 2 errors look the same is scary.

It is indeed...

Is there any way we can come up with more debug information to pinpoint the 
cause of this problem?


Ray

Re: fsck.reiser4 can't fix 1 fixable corruption

[Vitaly Fertman]

[-- Attachment #1: Type: text/plain, Size: 598 bytes --]

Hello, 

would you try this patch, it should fix the problem,
email me about the result please.

> ***** Semantic Traverse Pass: reiser4 semantic tree checking.
> Error: Node (1444613), item (2): StatData of the file [2a:0:ffff] has the
> wrong bytes (0), Fixed to (188). Plugin (stat40).
>
> Looks good, but when I run "fsck.reiser4 /dev/sys/backup" again, it tells
> me that it has found 1 fixable corruption and that I need to run it with
> the --fix parameter...
> P.S. How about all these spelling errors in fsck's output? ;)

all spelling fixes are appreciated :)

-- 
Thanks,
Vitaly Fertman

[-- Attachment #2: reiser4progs-0.5.3.patch --]
[-- Type: text/x-diff, Size: 17181 bytes --]

diff -Nru a/librepair/repair.c b/librepair/repair.c
--- a/librepair/repair.c	Fri Apr  2 17:41:46 2004
+++ b/librepair/repair.c	Fri Apr  2 17:41:46 2004
@@ -617,7 +617,7 @@
 	
 	/* FIXME: This is oid40 specific fix, not correct. To be rewritten when
 	   shared oid handling will be realy. */
-	if (control->oid && control->oid != val) {
+	if (control->oid && control->oid > val) {
 		if (mode != RM_BUILD) {
 			aal_mess("First not used oid %llu is wrong. %s %llu.",
 				 val, mode == RM_CHECK ? "Sould be" : 
diff -Nru a/plugin/key/key_large/key_large.c b/plugin/key/key_large/key_large.c
--- a/plugin/key/key_large/key_large.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/key/key_large/key_large.c	Fri Apr  2 17:41:46 2004
@@ -207,12 +207,12 @@
 	if (key_large_get_type(key1) == KEY_FILENAME_TYPE)
 		return 0;
 
-	ord1 = key_large_get_ordering(key1);
-	ord2 = key_large_get_ordering(key2);
-	
 	/* Checking ordering. */
-	if ((res = aal_memcmp(&ord1, &ord2, sizeof(ord1))))
+	if ((res = kl_comp_el((key_large_t *)key1->body,
+			      (key_large_t *)key2->body, 1)))
+	{
 		return res;
+	}
 	
 	/* Checking objectid  */
 	return kl_comp_el((key_large_t *)key1->body,
@@ -295,9 +295,8 @@
 				   len - INLINE_CHARS);
 	}
 
-	ordering |= ((uint64_t)plug_call(fibre->o.fibre_ops, 
-					 build, name + INLINE_CHARS,
-					 len - INLINE_CHARS) << FIBRE_SHIFT);
+	ordering |= ((uint64_t)plug_call(fibre->o.fibre_ops, build, 
+					 name, len) << FIBRE_SHIFT);
 	
 	/* Setting up objectid and offset */
 	key_large_set_ordering(key, ordering);
diff -Nru a/plugin/key/key_short/key_short.c b/plugin/key/key_short/key_short.c
--- a/plugin/key/key_short/key_short.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/key/key_short/key_short.c	Fri Apr  2 17:41:46 2004
@@ -276,9 +276,8 @@
 				   len - OBJECTID_CHARS);
 	}
 	
-	objectid |= ((uint64_t)plug_call(fibre->o.fibre_ops, 
-					 build, name + OBJECTID_CHARS,
-					 len - OBJECTID_CHARS) << FIBRE_SHIFT);
+	objectid |= ((uint64_t)plug_call(fibre->o.fibre_ops, build, 
+					 name, len) << FIBRE_SHIFT);
 	
 	/* Objectid must occupie 60 bits. If it takes more, then we have broken
 	   key, or objectid allocator reached this value, that impossible in
diff -Nru a/plugin/object/dir40/dir40.c b/plugin/object/dir40/dir40.c
--- a/plugin/object/dir40/dir40.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/object/dir40/dir40.c	Fri Apr  2 17:41:46 2004
@@ -464,7 +476,6 @@
 {
 	rid_t pid;
 	dir40_t *dir;
-	uint64_t mask;
 	entry_hint_t entry;
 	trans_hint_t body_hint;
 	reiser4_plug_t *body_plug;
@@ -556,9 +567,7 @@
 	}
 
 	/* Create stat data item. */
-	mask = (1 << SDEXT_UNIX_ID | 1 << SDEXT_LW_ID);
-	
-	if (obj40_create_stat(&dir->obj, hint->label.statdata, mask,
+	if (obj40_create_stat(&dir->obj, hint->label.statdata, 
 			      1, body_hint.len, 0, 1, S_IFDIR, NULL))
 	{
 	
diff -Nru a/plugin/object/dir40/dir40_repair.c b/plugin/object/dir40/dir40_repair.c
--- a/plugin/object/dir40/dir40_repair.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/object/dir40/dir40_repair.c	Fri Apr  2 17:41:46 2004
@@ -7,7 +7,11 @@
 #include "dir40.h"
 #include "repair/plugin.h"
 
-#define dir40_exts ((uint64_t)1 << SDEXT_LW_ID)
+/* Set of extentions that must present. */
+#define DIR40_EXTS_MUST ((uint64_t)1 << SDEXT_LW_ID)
+
+/* Set of unknown extentions. */
+#define DIR40_EXTS_UNKN ((uint64_t)1 << SDEXT_SYMLINK_ID)
 
 static errno_t dir40_extensions(reiser4_place_t *stat) {
 	uint64_t extmask;
@@ -15,12 +19,11 @@
 	extmask = obj40_extmask(stat);
 	
 	/* Check that there is no one unknown extension. */
-	/*
-	if (extmask & ~(dir40_exts | 1 << SDEXT_PLUG_ID))
+	if (extmask & DIR40_EXTS_UNKN)
 		return RE_FATAL;
-	*/
+	
 	/* Check that LW and UNIX extensions exist. */
-	return ((extmask & dir40_exts) == dir40_exts) ? 0 : RE_FATAL;
+	return ((extmask & DIR40_EXTS_MUST) == DIR40_EXTS_MUST) ? 0 : RE_FATAL;
 }
 
 /* Check SD extensions and that mode in LW extension is DIRFILE. */
@@ -153,8 +156,7 @@
 	
 	info = &dir->obj.info;
 	
-	if ((res = obj40_launch_stat(&dir->obj, NULL, dir40_exts, 1, 
-				     S_IFDIR, mode)))
+	if ((res = obj40_launch_stat(&dir->obj, NULL, 1, S_IFDIR, mode)))
 		return res;
 	
 	/* Try to register SD as an item of this file. */
diff -Nru a/plugin/object/obj40/obj40.c b/plugin/object/obj40/obj40.c
--- a/plugin/object/obj40/obj40.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/object/obj40/obj40.c	Fri Apr  2 17:41:46 2004
@@ -189,9 +189,9 @@
 
 /* Create stat data item basing on passed extensions @mask, @size, @bytes,
    @nlinks, @mode and @path for symlinks. Returns error or zero for success. */
-errno_t obj40_create_stat(obj40_t *obj, rid_t pid, uint64_t mask,
-			  uint64_t size, uint64_t bytes, uint64_t rdev,
-			  uint32_t nlink, uint16_t mode, char *path)
+errno_t obj40_create_stat(obj40_t *obj, rid_t pid, uint64_t size, 
+			  uint64_t bytes, uint64_t rdev, uint32_t nlink, 
+			  uint16_t mode, char *path)
 {
 	int64_t res;
 	lookup_t lookup;
@@ -218,7 +218,7 @@
 		  &hint.offset, &obj->info.object);
     
 	/* Initializing stat data item hint. */
-	stat.extmask = mask;
+	stat.extmask = (1 << SDEXT_UNIX_ID | 1 << SDEXT_LW_ID);
     	
 	/* Light weight hint initializing. */
 	lw_ext.size = size;
@@ -247,8 +247,10 @@
 	stat.ext[SDEXT_LW_ID] = &lw_ext;
 	stat.ext[SDEXT_UNIX_ID] = &unix_ext;
 
-	if ((1 << SDEXT_SYMLINK_ID) & mask)
+	if (path) {
+		stat.extmask |= (1 << SDEXT_SYMLINK_ID);
 		stat.ext[SDEXT_SYMLINK_ID] = path;
+	}
 	
 	hint.specific = &stat;
 
diff -Nru a/plugin/object/obj40/obj40.h b/plugin/object/obj40/obj40.h
--- a/plugin/object/obj40/obj40.h	Fri Apr  2 17:41:46 2004
+++ b/plugin/object/obj40/obj40.h	Fri Apr  2 17:41:46 2004
@@ -130,14 +130,12 @@
 
 #ifndef ENABLE_STAND_ALONE
 extern errno_t obj40_create_stat(obj40_t *obj, rid_t pid,
-				 uint64_t mask, uint64_t size,
-				 uint64_t bytes, uint64_t rdev,
-				 uint32_t nlink, uint16_t mode,
-				 char *path);
+				 uint64_t size, uint64_t bytes, 
+				 uint64_t rdev, uint32_t nlink, 
+				 uint16_t mode, char *path);
 
 extern errno_t obj40_launch_stat(obj40_t *obj, stat_func_t stat_func, 
-				 uint64_t mask, uint32_t nlink, 
-				 uint16_t objmode, uint8_t mode);
+				 uint32_t nlink, uint16_t objmode, uint8_t mode);
 
 extern errno_t obj40_check_stat(obj40_t *obj, nlink_func_t nlink_func,
 				mode_func_t mode_func, size_func_t size_func,
diff -Nru a/plugin/object/obj40/obj40_repair.c b/plugin/object/obj40/obj40_repair.c
--- a/plugin/object/obj40/obj40_repair.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/object/obj40/obj40_repair.c	Fri Apr  2 17:41:46 2004
@@ -6,6 +6,7 @@
 #ifndef ENABLE_STAND_ALONE
 #include "obj40.h"
 #include <repair/plugin.h>
+#include <sys/stat.h>
 
 /* Obtains the plugin of the type @type from SD if stored there, otherwise
    obtains the default one from the params. This differs from obj40_plug as it
@@ -60,13 +61,10 @@
 	if (info->start.plug->id.group != STATDATA_ITEM)
 		return RE_FATAL;
 	
-	/* Is @info->start SD of the wanted file? If some fields are broken, 
-	   like offset != 0, fix it at check_struct time. */
-	if (info->object.plug->o.key_ops->compshort(&info->object, 
-						    &info->start.key))
-	{
+	/* Compare the correct key with the place key. */
+	if (plug_call(info->object.plug->o.key_ops, compfull,
+		      &info->object, &info->start.key))
 		return RE_FATAL;
-	}
 	
 	/* Some SD is recognized. Check that this is our SD. */
 	return stat_func ? stat_func(&info->start) : 0;
@@ -79,48 +77,28 @@
 	reiser4_key_t key;
 
 	aal_assert("vpf-1121", obj->info.tree != NULL);
-	aal_assert("vpf-1127", obj->info.object.plug || obj->info.start.plug);
+	aal_assert("vpf-1127", obj->info.object.plug != NULL);
 	
 	info = &obj->info;
 	
-	if (info->object.plug) {
-		locality = plug_call(info->object.plug->o.key_ops,
-				     get_locality, &info->object);
-
-		objectid = plug_call(info->object.plug->o.key_ops,
-				     get_objectid, &info->object);
-
-		ordering = plug_call(info->object.plug->o.key_ops,
-				     get_ordering, &info->object);
-
-	
-		plug_call(info->object.plug->o.key_ops, build_generic, &key,
-			  KEY_STATDATA_TYPE, locality, ordering, objectid, 0);
-		
-		/* Realizing on the key: SD is not found. Check if the key 
-		   pointer is correct. */
-		if (plug_call(info->object.plug->o.key_ops, compfull, 
-			      &key, &info->object))
-			return RE_FATAL;
-	} else {
-		/* Realizing on the SD. */
-		aal_assert("vpf-1204",  info->start.plug->id.group == 
-			   		STATDATA_ITEM);
-
-		locality = plug_call(info->object.plug->o.key_ops,
-				     get_locality, &info->start.key);
-
-		objectid = plug_call(info->object.plug->o.key_ops,
-				     get_objectid, &info->start.key);
-
-		ordering = plug_call(info->object.plug->o.key_ops,
-				     get_ordering, &info->start.key);
-
-		/* Build the SD key into @info->object. */
-		plug_call(info->start.key.plug->o.key_ops, build_generic, 
-			  &info->object, KEY_STATDATA_TYPE, locality, 
-			  ordering, objectid, 0);
-	}
+	/* Check if the key pointer is correct and then check the found item 
+	   if it is SD with the proper key. */
+	locality = plug_call(info->object.plug->o.key_ops,
+			     get_locality, &info->object);
+
+	objectid = plug_call(info->object.plug->o.key_ops,
+			     get_objectid, &info->object);
+
+	ordering = plug_call(info->object.plug->o.key_ops,
+			     get_ordering, &info->object);
+
+	plug_call(info->object.plug->o.key_ops, build_generic, &key,
+		  KEY_STATDATA_TYPE, locality, ordering, objectid, 0);
+
+	/* Compare the correct key with the search one. */
+	if (plug_call(info->object.plug->o.key_ops, compfull, 
+		      &key, &info->object))
+		return RE_FATAL;
 	
 	/* @info->object is the key of SD for now and @info->start is the 
 	   result of tree lookup by @info->object -- skip objects w/out SD. */
@@ -255,8 +233,7 @@
 }
 
 errno_t obj40_launch_stat(obj40_t *obj, stat_func_t stat_func, 
-			  uint64_t mask, uint32_t nlink, 
-			  uint16_t objmode, uint8_t mode)
+			  uint32_t nlink, uint16_t objmode, uint8_t mode)
 {
 	reiser4_key_t *key;
 	lookup_t lookup;
@@ -309,8 +286,8 @@
 	if ((pid = obj->core->param_ops.value("statdata") == INVAL_PID))
 		return -EINVAL;
 
-	if ((res = obj40_create_stat(obj, pid, mask, 0, 0,
-				     0, nlink, objmode, NULL)))
+	if ((res = obj40_create_stat(obj, pid, 0, 0, 0, nlink, objmode, 
+				     objmode == S_IFLNK ? "FAKE_LINK" : NULL)))
 	{
 		aal_error("The file [%s] failed to create a "
 			  "StatData item. Plugin %s.", 
diff -Nru a/plugin/object/reg40/reg40.c b/plugin/object/reg40/reg40.c
--- a/plugin/object/reg40/reg40.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/object/reg40/reg40.c	Fri Apr  2 17:41:46 2004
@@ -195,7 +195,6 @@
 				     object_hint_t *hint)
 {
 	reg40_t *reg;
-	uint64_t mask;
 	
 	aal_assert("umka-1169", info != NULL);
 	aal_assert("umka-1738", hint != NULL);
@@ -229,12 +228,9 @@
 		goto error_free_reg;
 	}
 
-	/* Create stat data item with size, bytes, nlinks equal to zero and mask
-	   equal to @mask. */
-	mask = (1 << SDEXT_UNIX_ID | 1 << SDEXT_LW_ID);
-
+	/* Create stat data item with size, bytes, nlinks equal to zero. */
 	if (obj40_create_stat(&reg->obj, hint->label.statdata,
-			      mask, 0, 0, 0, 0, S_IFREG, NULL))
+			      0, 0, 0, 0, S_IFREG, NULL))
 	{
 		goto error_free_reg;
 	}
diff -Nru a/plugin/object/reg40/reg40_repair.c b/plugin/object/reg40/reg40_repair.c
--- a/plugin/object/reg40/reg40_repair.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/object/reg40/reg40_repair.c	Fri Apr  2 17:41:46 2004
@@ -7,7 +7,11 @@
 #include "reg40.h"
 #include "repair/plugin.h"
 
-#define reg40_exts ((uint64_t)1 << SDEXT_LW_ID)
+/* Set of extentions that must present. */
+#define REG40_EXTS_MUST ((uint64_t)1 << SDEXT_LW_ID)
+
+/* Set of unknown extentions. */
+#define REG40_EXTS_UNKN ((uint64_t)1 << SDEXT_SYMLINK_ID)
 
 static errno_t reg40_extensions(reiser4_place_t *stat) {
 	uint64_t extmask;
@@ -15,12 +19,12 @@
 	/* Check that there is no one unknown extension. */
 	extmask = obj40_extmask(stat);
 	
-	/*
-	if (extmask & ~(reg40_exts | 1 << SDEXT_PLUG_ID))
+	/* Check that there is no one unknown extension. */
+	if (extmask & REG40_EXTS_UNKN)
 		return RE_FATAL;
-	*/
+
 	/* Check that LW and UNIX extensions exist. */
-	return ((extmask & reg40_exts) == reg40_exts) ? 0 : RE_FATAL;
+	return ((extmask & REG40_EXTS_MUST) == REG40_EXTS_MUST) ? 0 : RE_FATAL;
 }
 
 /* Check SD extensions and that mode in LW extension is REGFILE. */
@@ -389,18 +393,12 @@
 	info = &reg->obj.info;
 	
 	if ((res = obj40_launch_stat(&reg->obj, callback_stat, 
-				     reg40_exts, 1, S_IFREG, mode)))
-	{
+				     1, S_IFREG, mode)))
 		return res;
-	}
 
 	/* Try to register SD as an item of this file. */
 	if (func && func(&info->start, data))
 		return -EINVAL;
-	
-	/* Fix SD's key if differs. */
-	if ((res = obj40_fix_key(&reg->obj, &info->start, &info->object, mode)))
-		return res;
 	
 	/* Get the reg file tail policy. */
 	if (!(reg->policy = obj40_plug(&reg->obj, POLICY_PLUG_TYPE, "policy")))
diff -Nru a/plugin/object/spl40/spl40.c b/plugin/object/spl40/spl40.c
--- a/plugin/object/spl40/spl40.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/object/spl40/spl40.c	Fri Apr  2 17:41:46 2004
@@ -61,7 +61,6 @@
 				     object_hint_t *hint)
 {
 	spl40_t *spl;
-	uint64_t mask;
 	
 	aal_assert("umka-2531", info != NULL);
 	aal_assert("umka-2533", hint != NULL);
@@ -73,10 +72,8 @@
 	/* Inizializes file handle */
 	obj40_init(&spl->obj, &spl40_plug, spl40_core, info);
 
-	mask = (1 << SDEXT_UNIX_ID | 1 << SDEXT_LW_ID);
-	
 	if (obj40_create_stat(&spl->obj, hint->label.statdata,
-			      mask, 0, 0, hint->body.spl.rdev,
+			      0, 0, hint->body.spl.rdev,
 			      0, hint->label.mode, NULL))
 	{
 		goto error_free_spl;
diff -Nru a/plugin/object/spl40/spl40_repair.c b/plugin/object/spl40/spl40_repair.c
--- a/plugin/object/spl40/spl40_repair.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/object/spl40/spl40_repair.c	Fri Apr  2 17:41:46 2004
@@ -7,7 +7,11 @@
 #include "spl40.h"
 #include "repair/plugin.h"
 
-#define spl40_exts ((uint64_t)1 << SDEXT_LW_ID)
+/* Set of extentions that must present. */
+#define SPL40_EXTS_MUST ((uint64_t)1 << SDEXT_LW_ID)
+
+/* Set of unknown extentions. */
+#define SPL40_EXTS_UNKN ((uint64_t)1 << SDEXT_SYMLINK_ID)
 
 static errno_t spl40_extensions(reiser4_place_t *stat) {
 	uint64_t extmask;
@@ -15,12 +19,11 @@
 	extmask = obj40_extmask(stat);
 	
 	/* Check that there is no one unknown extension. */
-	/*
-	if (extmask & ~(sym40_exts | 1 << SDEXT_PLUG_ID))
+	if (extmask & SPL40_EXTS_UNKN)
 		return RE_FATAL;
-	*/
+	
 	/* Check that LW and UNIX extensions exist. */
-	return ((extmask & spl40_exts) == spl40_exts) ? 0 : RE_FATAL;
+	return ((extmask & SPL40_EXTS_MUST) == SPL40_EXTS_MUST) ? 0 : RE_FATAL;
 }
 
 /* Check SD extensions and that mode in LW extension is DIRFILE. */
@@ -90,8 +93,7 @@
 	aal_assert("vpf-1358", spl->obj.info.tree != NULL);
 	aal_assert("vpf-1359", spl->obj.info.object.plug != NULL);
 
-	if ((res = obj40_launch_stat(&spl->obj, spl40_extensions, 
-				     0, 1, 0, mode)))
+	if ((res = obj40_launch_stat(&spl->obj, spl40_extensions, 1, 0, mode)))
 		return res;
 	
 	/* Try to register SD as an item of this file. */
diff -Nru a/plugin/object/sym40/sym40.c b/plugin/object/sym40/sym40.c
--- a/plugin/object/sym40/sym40.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/object/sym40/sym40.c	Fri Apr  2 17:41:46 2004
@@ -89,7 +89,6 @@
 {
 	sym40_t *sym;
 	uint32_t len;
-	uint64_t mask;
 	
 	aal_assert("umka-1741", info != NULL);
 	aal_assert("vpf-1094",  info->tree != NULL);
@@ -101,14 +100,10 @@
 	/* Inizializes symlink file handle. */
 	obj40_init(&sym->obj, &sym40_plug, sym40_core, info);
 
-	/* Initializing stat data extensions mask */
-	mask = (1 << SDEXT_UNIX_ID | 1 << SDEXT_LW_ID |
-		1 << SDEXT_SYMLINK_ID);
-
 	len = aal_strlen(hint->body.sym);
 
 	/* Create symlink sta data item. */
-	if (obj40_create_stat(&sym->obj, hint->label.statdata, mask,
+	if (obj40_create_stat(&sym->obj, hint->label.statdata,
 			      len, len, 0, 0, S_IFLNK, hint->body.sym))
 	{
 		goto error_free_sym;
diff -Nru a/plugin/object/sym40/sym40_repair.c b/plugin/object/sym40/sym40_repair.c
--- a/plugin/object/sym40/sym40_repair.c	Fri Apr  2 17:41:46 2004
+++ b/plugin/object/sym40/sym40_repair.c	Fri Apr  2 17:41:46 2004
@@ -8,20 +8,15 @@
 #include "sym40.h"
 #include "repair/plugin.h"
 
-#define sym40_exts ((uint64_t)1 << SDEXT_LW_ID | 1 << SDEXT_SYMLINK_ID)
+#define SYM40_EXTS_MUST ((uint64_t)1 << SDEXT_LW_ID | 1 << SDEXT_SYMLINK_ID)
 
 static errno_t sym40_extensions(reiser4_place_t *stat) {
 	uint64_t extmask;
 	
 	extmask = obj40_extmask(stat);
 	
-	/* Check that there is no one unknown extension. */
-	/*
-	if (extmask & ~(sym40_exts | 1 << SDEXT_PLUG_ID))
-		return RE_FATAL;
-	*/
 	/* Check that LW, UNIX and SYMLINK extensions exist. */
-	return ((extmask & sym40_exts) == sym40_exts) ? 0 : RE_FATAL;
+	return ((extmask & SYM40_EXTS_MUST) == SYM40_EXTS_MUST) ? 0 : RE_FATAL;
 }
 
 /* Check SD extensions and that mode in LW extension is DIRFILE. */
@@ -92,7 +87,7 @@
 	place = STAT_PLACE(&sym->obj);
 	
 	if ((res = obj40_launch_stat(&sym->obj, sym40_extensions, 
-				     sym40_exts, 1, S_IFLNK, mode)))
+				     1, S_IFLNK, mode)))
 		return res;
 	
 	/* Try to register SD as an item of this file. */

Re: fsck.reiser4 can't fix 1 fixable corruption

[Raymond A. Meijer]

On Wed 21 April 2004 11:55, Vitaly Fertman wrote:

Hi Vitaly,

Thanks for your reply!

> would you try this patch, it should fix the problem,
> email me about the result please.

I got several rejects:

patching file librepair/repair.c
Reversed (or previously applied) patch detected!  Assume -R? [n]
Apply anyway? [n]
Skipping patch.
1 out of 1 hunk ignored -- saving rejects to file librepair/repair.c.rej
patching file plugin/key/key_large/key_large.c
Hunk #2 FAILED at 295.
1 out of 2 hunks FAILED -- saving rejects to file 
plugin/key/key_large/key_large.c.rej
patching file plugin/key/key_short/key_short.c
Reversed (or previously applied) patch detected!  Assume -R? [n]
Apply anyway? [n]
Skipping patch.
1 out of 1 hunk ignored -- saving rejects to file 
plugin/key/key_short/key_short.c.rej
patching file plugin/object/dir40/dir40.c
Hunk #1 succeeded at 488 (offset 12 lines).
Hunk #2 succeeded at 579 (offset 12 lines).
patching file plugin/object/dir40/dir40_repair.c
patching file plugin/object/obj40/obj40.c
patching file plugin/object/obj40/obj40.h
patching file plugin/object/obj40/obj40_repair.c
patching file plugin/object/reg40/reg40.c
patching file plugin/object/reg40/reg40_repair.c
patching file plugin/object/spl40/spl40.c
patching file plugin/object/spl40/spl40_repair.c
patching file plugin/object/sym40/sym40.c
patching file plugin/object/sym40/sym40_repair.c

> > P.S. How about all these spelling errors in fsck's output? ;)

> all spelling fixes are appreciated :)

I'll let you know :)


Cheers,

Ray

Re: fsck.reiser4 can't fix 1 fixable corruption

[Raymond A. Meijer]

On Wed 21 April 2004 11:55, Vitaly Fertman wrote:

> would you try this patch, it should fix the problem,
> email me about the result please.

No luck... :(


fsck.reiser4 --fix /dev/sys/backup:

==============================================================================
*******************************************************************
This is an EXPERIMENTAL version of fsck.reiser4. Read REAMDE first.
*******************************************************************

Fscking the /dev/sys/backup block device.
Will fix minor corruptions of the Reiser4 SuperBblock.
Will fix minor corruptions of the Reiser4 FileSystem.
Continue?
(Yes/No): Yes
***** Openning the fs.
Reiser4 journal (journal40) on /dev/sys/backup: 0 transactions replaied of the 
total 0 blocks.
Reiser4 fs was detected on the /dev/sys/backup.
Master super block (16):
magic:          ReIsEr4
blksize:        4096
format:         0x0 (format40)
uuid:           f6a57152-e546-4a2a-b833-c0bce2a62caa
label:          <none>

Format super block (17):
plugin:         format40
description:    Disk-format for reiser4, ver. 0.5.3
magic:          ReIsEr40FoRmAt
flushes:        0
mkfs id:        0x6909edad
blocks:         2097152
free blocks:    394814
root block:     171208
tail policy:    0x4 (smart)
next oid:       0x78512
file count:     291088
tree height:    4
key policy:     LARGE

***** Tree Traverse Pass: scanning the reiser4 internal tree.
        Read nodes 151585
        Nodes left in the tree 151585
                Leaves of them 148854, Twigs of them 2674
        Time interval: Wed Apr 21 12:37:54 2004 - Wed Apr 21 12:40:20 2004
***** TwigScan Pass: checking extent pointers of all twigs.
        Read twigs 2674
        Time interval: Wed Apr 21 12:40:20 2004 - Wed Apr 21 12:40:28 2004
***** Semantic Traverse Pass: reiser4 semantic tree checking.
Error: Node (25), item (2): StatData of the file [2a:0:ffff] has the wrong 
bytes (0), Fixed to (188). Plugin (stat40).
        Time interval: Wed Apr 21 12:40:29 2004 - Wed Apr 21 12:45:19 2004
***** fsck.reiser4 finished at Wed Apr 21 12:45:19 2004
Closing fs...done

No corruption found.
==============================================================================


fsck /dev/sys/backup:

==============================================================================
fsck 1.35 (28-Feb-2004)
*******************************************************************
This is an EXPERIMENTAL version of fsck.reiser4. Read REAMDE first.
*******************************************************************

Fscking the /dev/sys/backup block device.
Will check the consistency of the Reiser4 SuperBlock.
Will check the consistency of the Reiser4 FileSystem.
Continue?
(Yes/No): Yes
***** Openning the fs.
Reiser4 journal (journal40) on /dev/sys/backup: 0 transactions replaied of the 
total 0 blocks.
Reiser4 fs was detected on the /dev/sys/backup.
Master super block (16):
magic:          ReIsEr4
blksize:        4096
format:         0x0 (format40)
uuid:           f6a57152-e546-4a2a-b833-c0bce2a62caa
label:          <none>

Format super block (17):
plugin:         format40
description:    Disk-format for reiser4, ver. 0.5.3
magic:          ReIsEr40FoRmAt
flushes:        0
mkfs id:        0x6909edad
blocks:         2097152
free blocks:    394814
root block:     171208
tail policy:    0x4 (smart)
next oid:       0x78512
file count:     291088
tree height:    4
key policy:     LARGE

***** Tree Traverse Pass: scanning the reiser4 internal tree.
        Read nodes 151585
        Nodes left in the tree 151585
                Leaves of them 148854, Twigs of them 2674
        Time interval: Wed Apr 21 12:46:49 2004 - Wed Apr 21 12:49:14 2004
***** TwigScan Pass: checking extent pointers of all twigs.
        Read twigs 2674
        Time interval: Wed Apr 21 12:49:14 2004 - Wed Apr 21 12:49:22 2004
***** Semantic Traverse Pass: reiser4 semantic tree checking.
Error: Node (25), item (2): StatData of the file [2a:0:ffff] has the wrong 
bytes (0), Should be (188). Plugin (stat40).
        Time interval: Wed Apr 21 12:49:22 2004 - Wed Apr 21 12:55:59 2004
***** fsck.reiser4 finished at Wed Apr 21 12:55:59 2004
Closing fs...done

1 fixable corruptions were detected in the FileSystem. Run with --fix option 
to fix them.
==============================================================================


Thanks,

Ray

Re: fsck.reiser4 can't fix 1 fixable corruption

[Vitaly Fertman]

On Wednesday 21 April 2004 13:59, Raymond A. Meijer wrote:
> On Wed 21 April 2004 11:55, Vitaly Fertman wrote:
> > would you try this patch, it should fix the problem,
> > email me about the result please.
>
> No luck... :(

ok, that patch avoids such problems if you have not fsck'ed yet.

I have put another internal reiser4progs snapshot with some more 
fixes into the www.namesys.com/snapshots/2004.04.21-internal.testing/
I have no problems with it, try it please.

-- 
Thanks,
Vitaly Fertman


> fsck.reiser4 --fix /dev/sys/backup:
>
> ===========================================================================
>=== ******************************************************************* This
> is an EXPERIMENTAL version of fsck.reiser4. Read REAMDE first.
> *******************************************************************
>
> Fscking the /dev/sys/backup block device.
> Will fix minor corruptions of the Reiser4 SuperBblock.
> Will fix minor corruptions of the Reiser4 FileSystem.
> Continue?
> (Yes/No): Yes
> ***** Openning the fs.
> Reiser4 journal (journal40) on /dev/sys/backup: 0 transactions replaied of
> the total 0 blocks.
> Reiser4 fs was detected on the /dev/sys/backup.
> Master super block (16):
> magic:          ReIsEr4
> blksize:        4096
> format:         0x0 (format40)
> uuid:           f6a57152-e546-4a2a-b833-c0bce2a62caa
> label:          <none>
>
> Format super block (17):
> plugin:         format40
> description:    Disk-format for reiser4, ver. 0.5.3
> magic:          ReIsEr40FoRmAt
> flushes:        0
> mkfs id:        0x6909edad
> blocks:         2097152
> free blocks:    394814
> root block:     171208
> tail policy:    0x4 (smart)
> next oid:       0x78512
> file count:     291088
> tree height:    4
> key policy:     LARGE
>
> ***** Tree Traverse Pass: scanning the reiser4 internal tree.
>         Read nodes 151585
>         Nodes left in the tree 151585
>                 Leaves of them 148854, Twigs of them 2674
>         Time interval: Wed Apr 21 12:37:54 2004 - Wed Apr 21 12:40:20 2004
> ***** TwigScan Pass: checking extent pointers of all twigs.
>         Read twigs 2674
>         Time interval: Wed Apr 21 12:40:20 2004 - Wed Apr 21 12:40:28 2004
> ***** Semantic Traverse Pass: reiser4 semantic tree checking.
> Error: Node (25), item (2): StatData of the file [2a:0:ffff] has the wrong
> bytes (0), Fixed to (188). Plugin (stat40).
>         Time interval: Wed Apr 21 12:40:29 2004 - Wed Apr 21 12:45:19 2004
> ***** fsck.reiser4 finished at Wed Apr 21 12:45:19 2004
> Closing fs...done
>
> No corruption found.
> ===========================================================================
>===
>
>
> fsck /dev/sys/backup:
>
> ===========================================================================
>=== fsck 1.35 (28-Feb-2004)
> *******************************************************************
> This is an EXPERIMENTAL version of fsck.reiser4. Read REAMDE first.
> *******************************************************************
>
> Fscking the /dev/sys/backup block device.
> Will check the consistency of the Reiser4 SuperBlock.
> Will check the consistency of the Reiser4 FileSystem.
> Continue?
> (Yes/No): Yes
> ***** Openning the fs.
> Reiser4 journal (journal40) on /dev/sys/backup: 0 transactions replaied of
> the total 0 blocks.
> Reiser4 fs was detected on the /dev/sys/backup.
> Master super block (16):
> magic:          ReIsEr4
> blksize:        4096
> format:         0x0 (format40)
> uuid:           f6a57152-e546-4a2a-b833-c0bce2a62caa
> label:          <none>
>
> Format super block (17):
> plugin:         format40
> description:    Disk-format for reiser4, ver. 0.5.3
> magic:          ReIsEr40FoRmAt
> flushes:        0
> mkfs id:        0x6909edad
> blocks:         2097152
> free blocks:    394814
> root block:     171208
> tail policy:    0x4 (smart)
> next oid:       0x78512
> file count:     291088
> tree height:    4
> key policy:     LARGE
>
> ***** Tree Traverse Pass: scanning the reiser4 internal tree.
>         Read nodes 151585
>         Nodes left in the tree 151585
>                 Leaves of them 148854, Twigs of them 2674
>         Time interval: Wed Apr 21 12:46:49 2004 - Wed Apr 21 12:49:14 2004
> ***** TwigScan Pass: checking extent pointers of all twigs.
>         Read twigs 2674
>         Time interval: Wed Apr 21 12:49:14 2004 - Wed Apr 21 12:49:22 2004
> ***** Semantic Traverse Pass: reiser4 semantic tree checking.
> Error: Node (25), item (2): StatData of the file [2a:0:ffff] has the wrong
> bytes (0), Should be (188). Plugin (stat40).
>         Time interval: Wed Apr 21 12:49:22 2004 - Wed Apr 21 12:55:59 2004
> ***** fsck.reiser4 finished at Wed Apr 21 12:55:59 2004
> Closing fs...done
>
> 1 fixable corruptions were detected in the FileSystem. Run with --fix
> option to fix them.
> ===========================================================================
>===
>
>
> Thanks,
>
> Ray

Re: fsck.reiser4 can't fix 1 fixable corruption

[Carl-Daniel Hailfinger]

Vitaly Fertman wrote:
> On Wednesday 21 April 2004 13:59, Raymond A. Meijer wrote:
> 
>>On Wed 21 April 2004 11:55, Vitaly Fertman wrote:
>>
>>>would you try this patch, it should fix the problem,
>>>email me about the result please.
>>
>>No luck... :(
> 
> 
> ok, that patch avoids such problems if you have not fsck'ed yet.

The patch didn't apply in the first place, so it of course couldn't work
(see Raymonds earlier mail).


> I have put another internal reiser4progs snapshot with some more 
> fixes into the www.namesys.com/snapshots/2004.04.21-internal.testing/
> I have no problems with it, try it please.


Carl-Daniel
-- 
http://www.hailfinger.org/

Re: fsck.reiser4 can't fix 1 fixable corruption

[Raymond A. Meijer]

On Wed 21 April 2004 14:51, Vitaly Fertman wrote:

> > No luck... :(

> ok, that patch avoids such problems if you have not fsck'ed yet.

Aha! :)

> I have put another internal reiser4progs snapshot with some more
> fixes into the www.namesys.com/snapshots/2004.04.21-internal.testing/
> I have no problems with it, try it please.

Alright, downloading it now...I'll let you know what happens.


Thanks,

Ray

Re: fsck.reiser4 can't fix 1 fixable corruption

[Sander Sweers]

<citaat van="Raymond A. Meijer">
> On Wed 21 April 2004 14:51, Vitaly Fertman wrote:
>
>> > No luck... :(
>
>> ok, that patch avoids such problems if you have not fsck'ed yet.
>
> Aha! :)
>
>> I have put another internal reiser4progs snapshot with some more
>> fixes into the www.namesys.com/snapshots/2004.04.21-internal.testing/
>> I have no problems with it, try it please.
>
> Alright, downloading it now...I'll let you know what happens.
>
>
> Thanks,
>
> Ray
>

I'l try the new reiser4progs tonight when i'm home.

Sander

reiser4progs 0.5.4 build error (was Re: fsck.reiser4 can't fix 1 fixable corruption)

[Domenico Andreoli]

On Wed, Apr 21, 2004 at 03:51:56PM +0400, Vitaly Fertman wrote:
> 
> I have put another internal reiser4progs snapshot with some more 
> fixes into the www.namesys.com/snapshots/2004.04.21-internal.testing/
> I have no problems with it, try it please.
> 

bad news, you must have forgotten something. there is no obj40_repair.h
in reiser4progs tarball :/

 i386-linux-gcc -DHAVE_CONFIG_H -I. -I. -I../../.. -I../../../include -D_REENTRANT -D_FILE_OFFSET_BITS=64 -DENABLE_SYMLINKS -DENABLE_SPECIAL -DENABLE_R5_HASH -DENABLE_FNV1_HASH -DENABLE_RUPASOV_HASH -DENABLE_TEA_HASH -DENABLE_DEG_HASH -DENABLE_LARGE_KEYS -DENABLE_SHORT_KEYS -O0 -g -W -Wall -Wno-unused-parameter -Wredundant-decls -MT libobj40_static_la-obj40_repair.lo -MD -MP -MF .deps/libobj40_static_la-obj40_repair.Tpo -c obj40_repair.c -o libobj40_static_la-obj40_repair.o
obj40_repair.c:8:26: obj40_repair.h: No such file or directory
[all sort of errors]

cheers
domenico

-----[ Domenico Andreoli, aka cavok
 --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50

Re: fsck.reiser4 can't fix 1 fixable corruption

[Sander Sweers]

[-- Attachment #1: Type: text/plain, Size: 949 bytes --]

Sander Sweers wrote:

><citaat van="Raymond A. Meijer">
>  
>
>>On Wed 21 April 2004 14:51, Vitaly Fertman wrote:
>>
>>    
>>
>>>>No luck... :(
>>>>        
>>>>
>>>ok, that patch avoids such problems if you have not fsck'ed yet.
>>>      
>>>
>>Aha! :)
>>
>>    
>>
>>>I have put another internal reiser4progs snapshot with some more
>>>fixes into the www.namesys.com/snapshots/2004.04.21-internal.testing/
>>>I have no problems with it, try it please.
>>>      
>>>
>>Alright, downloading it now...I'll let you know what happens.
>>
>>
>>Thanks,
>>
>>Ray
>>
>>    
>>
>
>I'l try the new reiser4progs tonight when i'm home.
>
>Sander
>
>
>  
>
The new reiser4progs fixed the problem :)

I also attached 2 modified gentoo ebuilds for those who want them.
Is it ok if I post them to bugs.gentoo.org so other gentooist can enjoys 
the new versions?

NOTE: the ebuilds have /--enable-stand-alone/ to have grub build reiser4 
support.

Thanks
Sander



[-- Attachment #2: libaal-0.5.1.ebuild --]
[-- Type: text/plain, Size: 759 bytes --]

# Copyright 1999-2004 Gentoo Technologies, Inc.
# Distributed under the terms of the GNU General Public License v2
# $Header: /home/cvsroot/gentoo-x86/sys-libs/libaal/libaal-0.5.0-r1ebuild,v 1.1 2004/02/06 14:12:24 vapier Exp $

DESCRIPTION="libaal library required by reiser4progs"
HOMEPAGE="http://www.namesys.com/v4/v4.html"
SRC_URI="http://thebsh.namesys.com/snapshots/2004.04.21-internal.testing.fixed/${P}.tar.gz"

LICENSE="GPL-2"
SLOT="0"
KEYWORDS="x86"

DEPEND=""

src_compile() {
	econf --libdir=/lib --enable-stand-alone || die "configure failed"
	emake || die "make failed"
}

src_install() {
	make DESTDIR=${D} install || die
	dodir /usr/lib
	mv ${D}/lib/libaal.{a,la} ${D}/usr/lib/
	dodoc AUTHORS BUGS CREDITS ChangeLog NEWS README THANKS TODO
}

[-- Attachment #3: reiser4progs-0.5.4.ebuild --]
[-- Type: text/plain, Size: 821 bytes --]

# Copyright 1999-2004 Gentoo Technologies, Inc.
# Distributed under the terms of the GNU General Public License v2
# $Header: /home/cvsroot/gentoo-x86/sys-fs/reiser4progs/reiser4progs-0.5.3.ebuild,v 1.3 2004/04/12 12:14:46 lv Exp $

DESCRIPTION="reiser4progs: mkfs, fsck, etc..."
HOMEPAGE="http://www.namesys.com/v4/v4.html"
SRC_URI="http://thebsh.namesys.com/snapshots/2004.04.21-internal.testing.fixed/${P}.tar.gz"

LICENSE="GPL-2"
SLOT="0"
KEYWORDS="~x86 ~ppc ~amd64"

DEPEND=">=sys-libs/libaal-0.5.1"

src_compile() {
	econf --sbindir=/sbin --libdir=/lib --enable-stand-alone || die "configure failed"
	emake || die "make failed"
}

src_install() {
	make DESTDIR=${D} install || die
	dodir /usr/lib
	mv ${D}/lib/lib{reiser4,repair}.{a,la} ${D}/usr/lib/
	dodoc AUTHORS BUGS CREDITS ChangeLog NEWS README THANKS TODO
}

Re: fsck.reiser4 can't fix 1 fixable corruption

[Domenico Andreoli]

On Wed, Apr 21, 2004 at 09:30:07PM +0200, Sander Sweers wrote:
> 
> NOTE: the ebuilds have /--enable-stand-alone/ to have grub build reiser4 
> support.

how hard is to have reiser4 support built in grub?

-----[ Domenico Andreoli, aka cavok
 --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50

Re: fsck.reiser4 can't fix 1 fixable corruption

[Sander Sweers]

Domenico Andreoli wrote:

>On Wed, Apr 21, 2004 at 09:30:07PM +0200, Sander Sweers wrote:
>  
>
>>NOTE: the ebuilds have /--enable-stand-alone/ to have grub build reiser4 
>>support.
>>    
>>
>
>how hard is to have reiser4 support built in grub?
>
>-----[ Domenico Andreoli, aka cavok
> --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
>   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50
>
>
>  
>
Building reiser4 support is easy, the patch applies cleanly on grub 0.94.
You have to add option --enable-stand-alone for reiser4progs and libaal 
before grub + patch will build it.*
*If you use gentoo see bug# 46410.

**I have not tested it yet but am looking forward to it :)

Sander*
*

Re: fsck.reiser4 can't fix 1 fixable corruption

[Domenico Andreoli]

On Wed, Apr 21, 2004 at 10:13:55PM +0200, Sander Sweers wrote:
>
> Building reiser4 support is easy, the patch applies cleanly on grub 0.94.
> You have to add option --enable-stand-alone for reiser4progs and libaal 
> before grub + patch will build it.*

what is this --enable-stand-alone argument? help messages says "enable
stand alone library" but i don't know which library it is talking about.

thanks
domenico

-----[ Domenico Andreoli, aka cavok
 --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50

Re: fsck.reiser4 can't fix 1 fixable corruption

[Vitaly Fertman]

On Thursday 22 April 2004 00:53, Domenico Andreoli wrote:
> On Wed, Apr 21, 2004 at 10:13:55PM +0200, Sander Sweers wrote:
> > Building reiser4 support is easy, the patch applies cleanly on grub 0.94.
> > You have to add option --enable-stand-alone for reiser4progs and libaal
> > before grub + patch will build it.*
>
> what is this --enable-stand-alone argument? help messages says "enable
> stand alone library" but i don't know which library it is talking about.

this is probably not very clear phasing, it configures the current library 
(libreiser4 for the reiser4progs package and libaal for the libaal package) 
in the stand alone mode.

-- 
Thanks,
Vitaly Fertman

Re: fsck.reiser4 can't fix 1 fixable corruption

[Domenico Andreoli]

On Thu, Apr 22, 2004 at 01:13:28AM +0400, Vitaly Fertman wrote:
> On Thursday 22 April 2004 00:53, Domenico Andreoli wrote:
> > On Wed, Apr 21, 2004 at 10:13:55PM +0200, Sander Sweers wrote:
> > > Building reiser4 support is easy, the patch applies cleanly on grub 0.94.
> > > You have to add option --enable-stand-alone for reiser4progs and libaal
> > > before grub + patch will build it.*
> >
> > what is this --enable-stand-alone argument? help messages says "enable
> > stand alone library" but i don't know which library it is talking about.
> 
> this is probably not very clear phasing, it configures the current library 
> (libreiser4 for the reiser4progs package and libaal for the libaal package) 
> in the stand alone mode.

sorry, it's still not clear to me what configuring a library for stand
alone mode means. do exist not stand alone libraries?

what does really change in the libraries? is something related to
reiser4 or to libtool & co?

is it legal for debian to distribute libaal and libreiser4 packages
compiled with this option?

thank you for the answers you will post :)

cheers
domenico

-----[ Domenico Andreoli, aka cavok
 --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50

Re: fsck.reiser4 can't fix 1 fixable corruption

[Vitaly Fertman]

On Thursday 22 April 2004 01:58, Domenico Andreoli wrote:
> On Thu, Apr 22, 2004 at 01:13:28AM +0400, Vitaly Fertman wrote:
> > On Thursday 22 April 2004 00:53, Domenico Andreoli wrote:
> > > On Wed, Apr 21, 2004 at 10:13:55PM +0200, Sander Sweers wrote:
> > > > Building reiser4 support is easy, the patch applies cleanly on grub
> > > > 0.94. You have to add option --enable-stand-alone for reiser4progs
> > > > and libaal before grub + patch will build it.*
> > >
> > > what is this --enable-stand-alone argument? help messages says "enable
> > > stand alone library" but i don't know which library it is talking
> > > about.
> >
> > this is probably not very clear phasing, it configures the current
> > library (libreiser4 for the reiser4progs package and libaal for the
> > libaal package) in the stand alone mode.
>
> sorry, it's still not clear to me what configuring a library for stand
> alone mode means. do exist not stand alone libraries?

with --enable-stand-alone you get another 'stand alone' library that 
includes the minimum of the code needed for programs like grub --
no librepair, no tree balancing code, no modification code, etc. The
code is minimised to be used with grub stage1_5. Libaal in stand-alone
includes simple memory menagement methods (malloc, free) to avoid
linking to libc -- as it links to no other libraries it is called stand-alone.

> what does really change in the libraries? is something related to
> reiser4 or to libtool & co?
>
> is it legal for debian to distribute libaal and libreiser4 packages
> compiled with this option?

yes. Although new changes are tested not in the stand-alone mode
first, and the stand-alone support may be delayed sometimes.

> thank you for the answers you will post :)

-- 
Thanks,
Vitaly Fertman

Re: fsck.reiser4 can't fix 1 fixable corruption

[Domenico Andreoli]

On Thu, Apr 22, 2004 at 02:46:21PM +0400, Vitaly Fertman wrote:
> with --enable-stand-alone you get another 'stand alone' library that 
> includes the minimum of the code needed for programs like grub --
> no librepair, no tree balancing code, no modification code, etc. The
> code is minimised to be used with grub stage1_5. Libaal in stand-alone
> includes simple memory menagement methods (malloc, free) to avoid
> linking to libc -- as it links to no other libraries it is called stand-alone.

thanks, now it's much more clear :)

> > what does really change in the libraries? is something related to
> > reiser4 or to libtool & co?
> >
> > is it legal for debian to distribute libaal and libreiser4 packages
> > compiled with this option?
> 
> yes. Although new changes are tested not in the stand-alone mode
> first, and the stand-alone support may be delayed sometimes.

this is not a problem, more people is able to use it, more will test
it for bugs :) this is the main reason i'm doing this packaging stuff ;)

ok, i'm trying to build packages (always targeted to debian experimental)
with these stand alone libraries. i'm also reading in configure help
message about "--enable-memory-manager". is anything a distribution
could need [to test]?

i admit, i didn't even look in namesys documentation, so feel free to
point me to docs...

thanks again
domenico

-----[ Domenico Andreoli, aka cavok
 --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50

Re: fsck.reiser4 can't fix 1 fixable corruption

[Domenico Andreoli]

On Thu, Apr 22, 2004 at 02:46:21PM +0400, Vitaly Fertman wrote:
> > is it legal for debian to distribute libaal and libreiser4 packages
> > compiled with this option?
> 
> yes. Although new changes are tested not in the stand-alone mode
> first, and the stand-alone support may be delayed sometimes.

i just uploaded new libaal 0.5.1-2 and reiser4progs 0.5.4-2 packages
to debian experimental, they now provide stand alone libraries.

they are going to be available on debian archives starting from tonight
and immediately at http://incoming.debian.org and my own homepage
http://people.debian.org/~cavok/ .

cheers
domenico

-----[ Domenico Andreoli, aka cavok
 --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50

Re: fsck.reiser4 can't fix 1 fixable corruption

[Hans Reiser]

Domenico Andreoli wrote:

>On Thu, Apr 22, 2004 at 02:46:21PM +0400, Vitaly Fertman wrote:
>  
>
>>>is it legal for debian to distribute libaal and libreiser4 packages
>>>compiled with this option?
>>>      
>>>
>>yes. Although new changes are tested not in the stand-alone mode
>>first, and the stand-alone support may be delayed sometimes.
>>    
>>
>
>i just uploaded new libaal 0.5.1-2 and reiser4progs 0.5.4-2 packages
>to debian experimental, they now provide stand alone libraries.
>
>they are going to be available on debian archives starting from tonight
>and immediately at http://incoming.debian.org and my own homepage
>http://people.debian.org/~cavok/ .
>
>cheers
>domenico
>
>-----[ Domenico Andreoli, aka cavok
> --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
>   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50
>
>
>  
>
I really appreciate your work in getting us well integrated into 
Debian.  It makes a difference.

Re: fsck.reiser4 can't fix 1 fixable corruption

[Hans Reiser]

Domenico Andreoli wrote:

>On Wed, Apr 21, 2004 at 09:30:07PM +0200, Sander Sweers wrote:
>  
>
>>NOTE: the ebuilds have /--enable-stand-alone/ to have grub build reiser4 
>>support.
>>    
>>
>
>how hard is to have reiser4 support built in grub?
>
>-----[ Domenico Andreoli, aka cavok
> --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
>   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50
>
>
>  
>
Yury told me it was done....  Yury....

Re: fsck.reiser4 can't fix 1 fixable corruption

[Domenico Andreoli]

On Thu, Apr 22, 2004 at 11:30:52AM -0700, Hans Reiser wrote:
> Domenico Andreoli wrote:
> 
> >how hard is to have reiser4 support built in grub?
>
> Yury told me it was done....  Yury....

FYI, i had to enable also libaal memomry manager (configure option
--enable-memory-manager) to build the right stand alone library for grub
reiser4 stage1.5, without it linker complains on missing malloc and free.

updated libaal and reiser4progs packages together with a reiser4 enabled
grub package are available at http://people.debian.org/~cavok/reiser4/.

hmmm... btw, i've never been able to use a stage1.5, any hint?

cheers
domenico

-----[ Domenico Andreoli, aka cavok
 --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50

Re: fsck.reiser4 can't fix 1 fixable corruption

[Yury Umanets]

On Thu, 2004-04-22 at 21:30, Hans Reiser wrote:
> Domenico Andreoli wrote:
> 
> >On Wed, Apr 21, 2004 at 09:30:07PM +0200, Sander Sweers wrote:
> >  
> >
> >>NOTE: the ebuilds have /--enable-stand-alone/ to have grub build reiser4 
> >>support.
> >>    
> >>
> >
> >how hard is to have reiser4 support built in grub?
> >
> >-----[ Domenico Andreoli, aka cavok
> > --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
> >   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50
> >
> >
> >  
> >
> Yury told me it was done....  Yury....
Hello!

Yes, it was done. But probably it is broken now due to lots of last
changes like new plugins, plugin inheritance, interface changes etc.

We should ask Vitaly what is the status of it now...

-- 
umka

Re: fsck.reiser4 can't fix 1 fixable corruption

[Vitaly Fertman]

On Friday 23 April 2004 11:04, Yury Umanets wrote:
> On Thu, 2004-04-22 at 21:30, Hans Reiser wrote:
> > Domenico Andreoli wrote:
> > >On Wed, Apr 21, 2004 at 09:30:07PM +0200, Sander Sweers wrote:
> > >>NOTE: the ebuilds have /--enable-stand-alone/ to have grub build
> > >> reiser4 support.
> > >
> > >how hard is to have reiser4 support built in grub?
> > >
> > >-----[ Domenico Andreoli, aka cavok
> > > --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
> > >   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50
> >
> > Yury told me it was done....  Yury....
>
> Hello!
>
> Yes, it was done. But probably it is broken now due to lots of last
> changes like new plugins, plugin inheritance, interface changes etc.
>
> We should ask Vitaly what is the status of it now...

It seems Hans asks why reiser4 support is not included into the 
grub yet. The patch for the grub is in the reiser4progs package, 
but IIRC will not be included into the grub untill reiser4 ships. 
Yura, did you get this answer?

-- 
Thanks,
Vitaly Fertman

Re: fsck.reiser4 can't fix 1 fixable corruption

[Vitaly Fertman]

On Thursday 22 April 2004 15:44, Domenico Andreoli wrote:
> On Thu, Apr 22, 2004 at 02:46:21PM +0400, Vitaly Fertman wrote:
> > with --enable-stand-alone you get another 'stand alone' library that
> > includes the minimum of the code needed for programs like grub --
> > no librepair, no tree balancing code, no modification code, etc. The
> > code is minimised to be used with grub stage1_5. Libaal in stand-alone
> > includes simple memory menagement methods (malloc, free) to avoid
> > linking to libc -- as it links to no other libraries it is called
> > stand-alone.
>
> thanks, now it's much more clear :)
>
> > > what does really change in the libraries? is something related to
> > > reiser4 or to libtool & co?
> > >
> > > is it legal for debian to distribute libaal and libreiser4 packages
> > > compiled with this option?
> >
> > yes. Although new changes are tested not in the stand-alone mode
> > first, and the stand-alone support may be delayed sometimes.
>
> this is not a problem, more people is able to use it, more will test
> it for bugs :) this is the main reason i'm doing this packaging stuff ;)
>
> ok, i'm trying to build packages (always targeted to debian experimental)
> with these stand alone libraries. i'm also reading in configure help
> message about "--enable-memory-manager". is anything a distribution
> could need [to test]?

yes, I forgot to tell the memory manager I mentioned above is enabled
with --enable-stand-alone and with --enable-memory-manager. Without
--enable-stand-alone you use libc methods, with --enable-stand-alone
only you can use you own memory methods implementation.

-- 
Thanks,
Vitaly Fertman

Re: fsck.reiser4 can't fix 1 fixable corruption

[Domenico Andreoli]

hi,

On Fri, Apr 23, 2004 at 01:03:50PM +0400, Vitaly Fertman wrote:
> > Yes, it was done. But probably it is broken now due to lots of last
> > changes like new plugins, plugin inheritance, interface changes etc.

i was unable to make it work, but this is probably a problem i have
with stage1_5.

i usually use this command in grub shell to install stuff:

grub> install (hd0,6)/boot/stage1 d (hd0) (hd0,6)/boot/stage2 p (hd0,6)/etc/grub.conf

probably with reiser4_stage1_5 it should be:

grub> install (hd0,6)/boot/stage1 d (hd0) (hd0,6)/boot/reiser4_stage1_5 p (hd0,6)/boot/stage2 (hd0,6)/etc/grub.conf

but i get a "Error 16: Inconsistent filesystem structure"


> It seems Hans asks why reiser4 support is not included into the 
> grub yet. The patch for the grub is in the reiser4progs package, 
> but IIRC will not be included into the grub untill reiser4 ships. 
> Yura, did you get this answer?

the most recent patch for grub i found is in snapshot 2004.03.26 and is
a separate diff. is it really supposed to be in reiser4progs tarball? i
didn't find it in latest 2004.04.21 :/

thanks
domenico

-----[ Domenico Andreoli, aka cavok
 --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50

Re: fsck.reiser4 can't fix 1 fixable corruption

[Yury Umanets]

On Fri, 2004-04-23 at 12:03, Vitaly Fertman wrote:
> On Friday 23 April 2004 11:04, Yury Umanets wrote:
> > On Thu, 2004-04-22 at 21:30, Hans Reiser wrote:
> > > Domenico Andreoli wrote:
> > > >On Wed, Apr 21, 2004 at 09:30:07PM +0200, Sander Sweers wrote:
> > > >>NOTE: the ebuilds have /--enable-stand-alone/ to have grub build
> > > >> reiser4 support.
> > > >
> > > >how hard is to have reiser4 support built in grub?
> > > >
> > > >-----[ Domenico Andreoli, aka cavok
> > > > --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
> > > >   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50
> > >
> > > Yury told me it was done....  Yury....
> >
> > Hello!
> >
> > Yes, it was done. But probably it is broken now due to lots of last
> > changes like new plugins, plugin inheritance, interface changes etc.
> >
> > We should ask Vitaly what is the status of it now...
> 
> It seems Hans asks why reiser4 support is not included into the 
> grub yet. The patch for the grub is in the reiser4progs package, 
> but IIRC will not be included into the grub untill reiser4 ships. 
> Yura, did you get this answer?
No, I have not asked about. We just discussed they we have implemented
grub support and grub maintainer was agree with me, that library is good
on this. 

That times reiser4progs interfaces was under heavy change and there was
no point to ask for inclusion. I guess, that now it is good time.
Reiser4 is almost shipped, interfaces are stable...

-- 
umka

Re: fsck.reiser4 can't fix 1 fixable corruption

[Vitaly Fertman]

On Friday 23 April 2004 14:27, Domenico Andreoli wrote:
> hi,
>
> On Fri, Apr 23, 2004 at 01:03:50PM +0400, Vitaly Fertman wrote:
> > > Yes, it was done. But probably it is broken now due to lots of last
> > > changes like new plugins, plugin inheritance, interface changes etc.
>
> i was unable to make it work, but this is probably a problem i have
> with stage1_5.
>
> i usually use this command in grub shell to install stuff:
>
> grub> install (hd0,6)/boot/stage1 d (hd0) (hd0,6)/boot/stage2 p
> (hd0,6)/etc/grub.conf
>
> probably with reiser4_stage1_5 it should be:
>
> grub> install (hd0,6)/boot/stage1 d (hd0) (hd0,6)/boot/reiser4_stage1_5 p
> (hd0,6)/boot/stage2 (hd0,6)/etc/grub.conf
>
> but i get a "Error 16: Inconsistent filesystem structure"

Is root set correctly?

does it work with the stage2 ?

check info grub please for things like:
*Caution3:* You must specify the option `--stage2' in the grub
     shell, if you cannot unmount the filesystem where your stage2 file
     resides. The argument should be the file name in your operating
     system.

did you try 'setup' command?

could also be a problem in progs.

> > It seems Hans asks why reiser4 support is not included into the
> > grub yet. The patch for the grub is in the reiser4progs package,
> > but IIRC will not be included into the grub untill reiser4 ships.
> > Yura, did you get this answer?
>
> the most recent patch for grub i found is in snapshot 2004.03.26 and is
> a separate diff. is it really supposed to be in reiser4progs tarball? i
> didn't find it in latest 2004.04.21 :/

I mean reiser4progs bk repository. As there are still many changes that 
break the grub support, it is released separately with the progs package
where it works. And later it is supposed to be included into the grub.

-- 
Thanks,
Vitaly Fertman

Re: fsck.reiser4 can't fix 1 fixable corruption

[Hans Reiser]

Vitaly Fertman wrote:

>On Thursday 22 April 2004 15:44, Domenico Andreoli wrote:
>  
>
>>On Thu, Apr 22, 2004 at 02:46:21PM +0400, Vitaly Fertman wrote:
>>    
>>
>>>with --enable-stand-alone you get another 'stand alone' library that
>>>includes the minimum of the code needed for programs like grub --
>>>no librepair, no tree balancing code, no modification code, etc. The
>>>code is minimised to be used with grub stage1_5. Libaal in stand-alone
>>>includes simple memory menagement methods (malloc, free) to avoid
>>>linking to libc -- as it links to no other libraries it is called
>>>stand-alone.
>>>      
>>>
>>thanks, now it's much more clear :)
>>    
>>
how about calling it --minimal-footprint and fixing the documentation to 
include some of the words above?

Please work on reviewing your documentation and options naming for 
greater clarity for a day.

>>    
>>
>>>>what does really change in the libraries? is something related to
>>>>reiser4 or to libtool & co?
>>>>
>>>>is it legal for debian to distribute libaal and libreiser4 packages
>>>>compiled with this option?
>>>>        
>>>>
>>>yes. Although new changes are tested not in the stand-alone mode
>>>first, and the stand-alone support may be delayed sometimes.
>>>      
>>>
>>this is not a problem, more people is able to use it, more will test
>>it for bugs :) this is the main reason i'm doing this packaging stuff ;)
>>
>>ok, i'm trying to build packages (always targeted to debian experimental)
>>with these stand alone libraries. i'm also reading in configure help
>>message about "--enable-memory-manager". is anything a distribution
>>could need [to test]?
>>    
>>
>
>yes, I forgot to tell the memory manager I mentioned above is enabled
>with --enable-stand-alone and with --enable-memory-manager. Without
>--enable-stand-alone you use libc methods, with --enable-stand-alone
>only you can use you own memory methods implementation.
>
>  
>

Re: fsck.reiser4 can't fix 1 fixable corruption

[Vitaly Fertman]

On Saturday 24 April 2004 00:19, Hans Reiser wrote:
> Vitaly Fertman wrote:
> >On Thursday 22 April 2004 15:44, Domenico Andreoli wrote:
> >>On Thu, Apr 22, 2004 at 02:46:21PM +0400, Vitaly Fertman wrote:
> >>>with --enable-stand-alone you get another 'stand alone' library that
> >>>includes the minimum of the code needed for programs like grub --
> >>>no librepair, no tree balancing code, no modification code, etc. The
> >>>code is minimised to be used with grub stage1_5. Libaal in stand-alone
> >>>includes simple memory menagement methods (malloc, free) to avoid
> >>>linking to libc -- as it links to no other libraries it is called
> >>>stand-alone.
> >>
> >>thanks, now it's much more clear :)
>
> how about calling it --minimal-footprint and fixing the documentation to
> include some of the words above?

It have to be --enable-minimal-footprint as you get this minimal-footprint 
library in addition to the base library.

> Please work on reviewing your documentation and options naming for
> greater clarity for a day.

yes, I am going to review all documentation just after finishing with 
the plugin inheritance.

-- 
Thanks,
Vitaly Fertman

Re: fsck.reiser4 can't fix 1 fixable corruption

[Hans Reiser]

Vitaly Fertman wrote:

>On Saturday 24 April 2004 00:19, Hans Reiser wrote:
>  
>
>>Vitaly Fertman wrote:
>>    
>>
>>>On Thursday 22 April 2004 15:44, Domenico Andreoli wrote:
>>>      
>>>
>>>>On Thu, Apr 22, 2004 at 02:46:21PM +0400, Vitaly Fertman wrote:
>>>>        
>>>>
>>>>>with --enable-stand-alone you get another 'stand alone' library that
>>>>>includes the minimum of the code needed for programs like grub --
>>>>>no librepair, no tree balancing code, no modification code, etc. The
>>>>>code is minimised to be used with grub stage1_5. Libaal in stand-alone
>>>>>includes simple memory menagement methods (malloc, free) to avoid
>>>>>linking to libc -- as it links to no other libraries it is called
>>>>>stand-alone.
>>>>>          
>>>>>
>>>>thanks, now it's much more clear :)
>>>>        
>>>>
>>how about calling it --minimal-footprint and fixing the documentation to
>>include some of the words above?
>>    
>>
>
>It have to be --enable-minimal-footprint as you get this minimal-footprint 
>library in addition to the base library.
>  
>
I don't understand what you said.

>  
>
>>Please work on reviewing your documentation and options naming for
>>greater clarity for a day.
>>    
>>
>
>yes, I am going to review all documentation just after finishing with 
>the plugin inheritance.
>
>  
>

Re: fsck.reiser4 can't fix 1 fixable corruption

[Hans Reiser]

Domenico Andreoli wrote:

>On Thu, Apr 22, 2004 at 11:30:52AM -0700, Hans Reiser wrote:
>  
>
>>Domenico Andreoli wrote:
>>
>>    
>>
>>>how hard is to have reiser4 support built in grub?
>>>      
>>>
>>Yury told me it was done....  Yury....
>>    
>>
>
>FYI, i had to enable also libaal memomry manager (configure option
>--enable-memory-manager) to build the right stand alone library for grub
>reiser4 stage1.5, without it linker complains on missing malloc and free.
>
>updated libaal and reiser4progs packages together with a reiser4 enabled
>grub package are available at http://people.debian.org/~cavok/reiser4/.
>
>hmmm... btw, i've never been able to use a stage1.5, any hint?
>
>cheers
>domenico
>
>-----[ Domenico Andreoli, aka cavok
> --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
>   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50
>
>
>  
>
Vitaly, please install your system using grub, write down every step as 
you do it, and create a web page.  Hint: try to eliminate the need for 
reading your instructions.  I would like to see your instructions 
consist of:

1: install grub by doing XXX

2: install reiser4 by doing YYY

Use grub normally and it works.

I haven't followed the details of your conversation with Domenico, but 
it looks way more complicated for him than it should be.  Please fix.

Hans

Re: fsck.reiser4 can't fix 1 fixable corruption

[Domenico Andreoli]

On Sun, Apr 25, 2004 at 09:31:58AM -0700, Hans Reiser wrote:
> Vitaly, please install your system using grub, write down every step as 
> you do it, and create a web page.  Hint: try to eliminate the need for 
> reading your instructions.  I would like to see your instructions 
> consist of:
> 
> 1: install grub by doing XXX
> 
> 2: install reiser4 by doing YYY
> 
> Use grub normally and it works.
> 
> I haven't followed the details of your conversation with Domenico, but 
> it looks way more complicated for him than it should be.  Please fix.

i'd be happy also with these steps:

  1. install reiser4 (with stand alone libraries)

  2. build a patched grub

  3. install grub

  4. use grub

my problem is that if something doesn't work i'm not able to say where
is it :/

thanks
domenico

ps: surely Vitaly is hating me... :(

-----[ Domenico Andreoli, aka cavok
 --[ http://people.debian.org/~cavok/gpgkey.asc
   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50

Re: fsck.reiser4 can't fix 1 fixable corruption

[Vitaly Fertman]

On Sunday 25 April 2004 20:31, Hans Reiser wrote:
> Domenico Andreoli wrote:
> >On Thu, Apr 22, 2004 at 11:30:52AM -0700, Hans Reiser wrote:
> >>Domenico Andreoli wrote:
> >>>how hard is to have reiser4 support built in grub?
> >>
> >>Yury told me it was done....  Yury....
> >
> >FYI, i had to enable also libaal memomry manager (configure option
> >--enable-memory-manager) to build the right stand alone library for grub
> >reiser4 stage1.5, without it linker complains on missing malloc and free.
> >
> >updated libaal and reiser4progs packages together with a reiser4 enabled
> >grub package are available at http://people.debian.org/~cavok/reiser4/.
> >
> >hmmm... btw, i've never been able to use a stage1.5, any hint?
> >
> >cheers
> >domenico
> >
> >-----[ Domenico Andreoli, aka cavok
> > --[ http://filibusta.crema.unimi.it/~cavok/gpgkey.asc
> >   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50
>
> Vitaly, please install your system using grub, write down every step as
> you do it, and create a web page.  Hint: try to eliminate the need for
> reading your instructions.  I would like to see your instructions
> consist of:
>
> 1: install grub by doing XXX
>
> 2: install reiser4 by doing YYY
>
> Use grub normally and it works.

ok, I will.

> I haven't followed the details of your conversation with Domenico, but
> it looks way more complicated for him than it should be.  Please fix.

Probably this is complicated due to some bug, probably in progs. 
Although if I understand correctly grub works for Domenico without 
stage1_5, right?

-- 
Thanks,
Vitaly Fertman

Re: fsck.reiser4 can't fix 1 fixable corruption

[Hans Reiser]

Vitaly, please report on the status of your solution in this matter.

Hans

Domenico Andreoli wrote:

>On Sun, Apr 25, 2004 at 09:31:58AM -0700, Hans Reiser wrote:
>  
>
>>Vitaly, please install your system using grub, write down every step as 
>>you do it, and create a web page.  Hint: try to eliminate the need for 
>>reading your instructions.  I would like to see your instructions 
>>consist of:
>>
>>1: install grub by doing XXX
>>
>>2: install reiser4 by doing YYY
>>
>>Use grub normally and it works.
>>
>>I haven't followed the details of your conversation with Domenico, but 
>>it looks way more complicated for him than it should be.  Please fix.
>>    
>>
>
>i'd be happy also with these steps:
>
>  1. install reiser4 (with stand alone libraries)
>
>  2. build a patched grub
>
>  3. install grub
>
>  4. use grub
>
>my problem is that if something doesn't work i'm not able to say where
>is it :/
>
>thanks
>domenico
>
>ps: surely Vitaly is hating me... :(
>
>-----[ Domenico Andreoli, aka cavok
> --[ http://people.debian.org/~cavok/gpgkey.asc
>   ---[ 3A0F 2F80 F79C 678A 8936  4FEE 0677 9033 A20E BC50
>
>
>  
>

Re: fsck.reiser4 can't fix 1 fixable corruption

[Vitaly Fertman]

On Friday 30 April 2004 07:16, Hans Reiser wrote:
> Vitaly, please report on the status of your solution in this matter.

currently stand-alone libreiser4 is not compilable.
I am not finished yet with the repair code with the plugin
inheritance, this is what  am working on now. Then I will 
work on the stand-alone libreiser4 and grub and then get 
a solution.

-- 
Thanks,
Vitaly Fertman