Re: Fix udev_volume_id for reiserfs v3.5

[Hannes Reinecke <hare@suse.de>]

[-- Attachment #1: Type: text/plain, Size: 1051 bytes --]

Kay Sievers wrote:
> On Tue, 2005-03-15 at 14:18 +0100, Hannes Reinecke wrote:
>>the attached patch fixes udev_volume_id for old reiserfs volumes.
> 
> The signature for that version can be at the offset of the later
> versions? We already check for that signature at a different offset.
> 
Appearently so.

>>How are we supposed to handle old/invalid labels?
>>The one I'm reading here is
>>eb 4d 72 17 06 4e 72 17 09 4e 72 17 53 4e 72 17
>>which is clearly bogus. I'll check with our fs-gurus but I think
>>we should filter them out in udev_volume_id.
>>Can't we test for isascii(rs->label[0]) ?
> 
> If that helps, sure. But what if the garbage starts at the second char?
> And what's with the UTF8 char sequences?
> 
Well, I just asked our resident reiserfs guru (thx mason :-).
Reiser v3.5 does not support labels nor uuids.
Corrected patch attached.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke			hare@suse.de
SuSE Linux AG				S390 & zSeries
Maxfeldstraße 5				+49 911 74053 688
90409 Nürnberg				http://www.suse.de

[-- Attachment #2: udev-053-bk-fix-reiserfs-v35.patch --]
[-- Type: text/plain, Size: 654 bytes --]

===== extras/volume_id/volume_id/reiserfs.c 1.7 vs edited =====
--- 1.7/extras/volume_id/volume_id/reiserfs.c	2005-03-09 16:58:00 +01:00
+++ edited/extras/volume_id/volume_id/reiserfs.c	2005-03-15 16:14:23 +01:00
@@ -80,6 +80,10 @@
 		return -1;
 
 	rs = (struct reiserfs_super_block *) buf;;
+	if (memcmp(rs->magic, "ReIsErFs", 8) == 0) {
+		strcpy(id->type_version, "3.5");
+		goto found;
+	}
 	if (memcmp(rs->magic, "ReIsEr2Fs", 9) == 0) {
 		strcpy(id->type_version, "3.6");
 		goto found_v3;
@@ -104,7 +108,7 @@
 
 	if (memcmp(rs->magic, "ReIsErFs", 8) == 0) {
 		strcpy(id->type_version, "3.5");
-		goto found_v3;
+		goto found;
 	}
 
 	return -1;