[Bug 255413] New: Wrong device links created
https://bugzilla.novell.com/show_bug.cgi?id=255413 Summary: Wrong device links created Product: openSUSE 10.3 Version: Alpha 2 Platform: Other OS/Version: Other Status: NEW Severity: Normal Priority: P5 - None Component: Basesystem AssignedTo: bnc-team-screening@forge.provo.novell.com ReportedBy: skh@novell.com QAContact: qa@suse.de Update from 10.3 Alpha1 to 10.3 Alpha2 During update, I was not offered the installed system as one to be updated, I had to select "show all partitions". Update after that went fine (installed packages were shown and correctly updated etc.) until first boot. / was mounted fine, /local was not. Entries in /etc/fstab look like this: /dev/disk/by-id/ata-ExcelStor_Technology_J680_VNR21EG2076Y9A-part2 / ext3 acl,user_xattr 1 1 /dev/disk/by-id/ata-ExcelStor_Technology_J680_VNR21EG2076Y9A-part3 /local reiserfs defaults 1 2 /dev/disk/by-id/ata-ExcelStor_Technology_J680_VNR21EG2076Y9A-part1 swap swap defaults 0 0 but the links to the /dev/sda* files created look like this: /dev/disk/by-id/ata-ExcelStor_Techno_VNR21EG2076Y9A /dev/disk/by-id/ata-ExcelStor_Techno_VNR21EG2076Y9A-part1 /dev/disk/by-id/ata-ExcelStor_Techno_VNR21EG2076Y9A-part2 /dev/disk/by-id/ata-ExcelStor_Techno_VNR21EG2076Y9A-part3 /dev/disk/by-id/edd-int13_dev80 /dev/disk/by-id/edd-int13_dev80-part1 /dev/disk/by-id/edd-int13_dev80-part2 /dev/disk/by-id/edd-int13_dev80-part3 /dev/disk/by-id/scsi-SATA_ExcelStor_Techn_VNR21EG2076Y9A /dev/disk/by-id/scsi-SATA_ExcelStor_Techn_VNR21EG2076Y9A-part1 /dev/disk/by-id/scsi-SATA_ExcelStor_Techn_VNR21EG2076Y9A-part2 /dev/disk/by-id/scsi-SATA_ExcelStor_Techn_VNR21EG2076Y9A-part3 The /dev/disk/by-id/ata-ExcelStor_Techno_VNR21EG2076Y9A* links point to the correct /dev/sda* files and are mountable manually. I left the system (chomsky) in this state in case it is needed for debugging. YaST2 logs attached. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 ------- Comment #1 from skh@novell.com 2007-03-16 09:17 MST ------- Created an attachment (id=125045) --> (https://bugzilla.novell.com/attachment.cgi?id=125045&action=view) YaST2 logs -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 ------- Comment #2 from skh@novell.com 2007-03-16 09:32 MST ------- Error during mount is: mount: special device /dev/disk/by-id/ata-ExcelStor_Technology_J680_VNR21EG2076Y9A-part3 does not exist which during boot triggers root login, please repair manually etc. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 skh@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Severity|Normal |Critical -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 chrubis@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|bnc-team- |yast2-maintainers@suse.de |screening@forge.provo.novell| |.com | -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 aj@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|yast2-maintainers@suse.de |snwint@novell.com -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 snwint@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |fehr@novell.com, aj@novell.com, | |snwint@novell.com, kukuk@novell.com AssignedTo|snwint@novell.com |kasievers@novell.com Severity|Critical |Blocker ------- Comment #3 from snwint@novell.com 2007-03-19 05:43 MST ------- On a first look, I see no reason why udev can't create the same names with libata devices than it was before as ide device. E.g. device name is still 'Techno' (with 'o'), no need to cut it out. If the name is going to change every now and then when the kernel changes its driver model, we are no better off than with plain kernel names here. If this can't be addressed in udev, I strongly suggest to rethink the move to disk/by-id as default device names. Using disk/by-uuid would be the better choice, IMO. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 kasievers@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |ASSIGNED ------- Comment #4 from kasievers@novell.com 2007-03-19 09:49 MST ------- That happens for identifier longer than 16 char, because SCSI is per definition limited to that length. We may be able to ask scsi_id to read the identifier directly from disk (page 0x83) instead of the kernel. I'll change that with the next submit. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 ------- Comment #5 from snwint@novell.com 2007-03-19 10:15 MST ------- 'ExcelStor Techno' is not longer than 16. ;-) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 ------- Comment #6 from kasievers@novell.com 2007-03-19 10:20 MST ------- But the expected "ExcelStor_Technology" is longer, right? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 ------- Comment #7 from snwint@novell.com 2007-03-19 10:42 MST ------- Yes, agreed. :-) BTW, where will that 'J680' be coming from? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 kasievers@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |hare@novell.com ------- Comment #8 from kasievers@novell.com 2007-03-19 17:20 MST ------- That's the complete name of the device: $ sg_inq -p=0x83 /dev/sda ... 00 00 83 00 60 02 00 00 14 20 20 20 20 20 20 56 4e ...`.... VN 10 52 32 31 45 47 32 30 37 36 59 39 41 02 01 00 44 R21EG2076Y9A...D 20 41 54 41 20 20 20 20 20 45 78 63 65 6c 53 74 6f ATA ExcelSto 30 72 20 54 65 63 68 6e 6f 6c 6f 67 79 20 4a 36 38 r Technology J68 40 30 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 0 ... But scsi_id gives me: $ scsi_id -p 0x83 -x -g -u -s /block/sda ID_VENDOR=ATA ID_MODEL=ExcelStor_Techno ID_REVISION=V32O ID_SERIAL=1ATA_ExcelStor_Technology_J680_VNR21EG2076Y9A ID_SERIAL_SHORT=ATA_ExcelStor_Technology_J680_VNR21EG2076Y9A ... I have no idea, how to extract the full model string: "ExcelStor_Technology_J680" from the (now) scsi disk, it is truncated by the scsi-stack to the usual 16 chars. The vpd page contains it, but scsi_id finds a T10 string which already contains the prepended vendor/product and serial. Hannes, any idea what to do? Wasn't there something like "ata pass-through" for libata, so that we could try to use ata_id on the (now) scsi disk, if we see "ATA" as vendor? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 kasievers@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |teheo@novell.com ------- Comment #9 from kasievers@novell.com 2007-03-20 04:10 MST ------- Tejun, we need the original strings from the ata-disk to be able to create the same identifiers for the ide to libata driver-conversion. The long ide strings get truncated by the scsi-stack. Can we make ioctl(HDIO_GET_IDENTITY) working on libata disks? It currently says: $ /lib/udev/ata_id /dev/sda ... ioctl(3, 0x30d, 0x7fff238d8080) = -1 ENOTTY (Inappropriate ioctl for device) ... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 teheo@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|teheo@novell.com | ------- Comment #10 from teheo@novell.com 2007-03-20 11:03 MST ------- Okay, patch committed to SL102_BRANCH and SLES10_SP1_BRANCH. JFYI, HEAD already contained the change, so /lib/udev/ata_id /dev/sda should have worked in SL103. You can do 'hdparm -i /dev/sda' to see whether the ioctl works. # uname -a Linux t1.dyndns.org 2.6.21-rc4-work #408 PREEMPT Wed Mar 21 01:47:12 KST 2007 i686 GNU/Linux # hdparm -i /dev/sda /dev/sda: Model=Maxtor 6B080M0 , FwRev=BANC1980, SerialNo=B202XD1H Config={ Fixed } RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=4 .... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 ------- Comment #11 from aj@novell.com 2007-03-20 13:00 MST ------- So, it will work with 2.6.21-rc4. 10.3 alpha2 had 2.6.20 in it - but if someone submits a new kernel to FACTORY, we will have 2.6.21. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 ------- Comment #12 from skh@novell.com 2007-03-22 07:40 MST ------- I tested with /mounts/dist/install/stable-x86/suse/i586/kernel-default-2.6.21_rc4_git5-2.i586.rpm as prompted by kay, which didn't help. On boot, grub shows: root (/dev/disk/by-id/ata-ExcelStor_Technology_J680_VNR21EG2076Y9A-part2) Error 23: Error while parsing number Press any key to continue... after which I'm in /bin/sh without any filesystem. Booting with root=/dev/sda2 works. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 ------- Comment #13 from kasievers@novell.com 2007-03-22 07:49 MST ------- Created an attachment (id=125958) --> (https://bugzilla.novell.com/attachment.cgi?id=125958&action=view) ata compat rules 2.6.21 only allows us to query the "real" ide strings. But you will need need a new: /etc/udev/rules.d/60-persistent-storage.rules to make use of that. Attached file works for me, but I don't have a device with such a long name here to test. :) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 ------- Comment #14 from skh@novell.com 2007-03-22 07:56 MST ------- OK, comment #12 was just the result of a completely broken grub configuration. The links now look like this: lrwxrwxrwx 1 root root 9 2007-03-22 14:48 /dev/disk/by-id/ata-ExcelStor_Techno_ATA_ExcelStor_Technology_J680_VNR21EG2076Y9A -> ../../sda lrwxrwxrwx 1 root root 10 2007-03-22 14:48 /dev/disk/by-id/ata-ExcelStor_Techno_ATA_ExcelStor_Technology_J680_VNR21EG2076Y9A-part1 -> ../../sda1 lrwxrwxrwx 1 root root 10 2007-03-22 14:48 /dev/disk/by-id/ata-ExcelStor_Techno_ATA_ExcelStor_Technology_J680_VNR21EG2076Y9A-part2 -> ../../sda2 lrwxrwxrwx 1 root root 10 2007-03-22 14:48 /dev/disk/by-id/ata-ExcelStor_Techno_ATA_ExcelStor_Technology_J680_VNR21EG2076Y9A-part3 -> ../../sda3 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=255413 kasievers@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |RESOLVED Resolution| |FIXED ------- Comment #15 from kasievers@novell.com 2007-03-22 10:20 MST ------- We call ata_id now for scsi devices with vendor=ATA. This needs kernel 2.6.21. New rules submitted. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
participants (1)
-
bugzilla_noreply@novell.com