Bug 1525 - Problem with my DVD player / recorder,it is inaccessible.
Summary: Problem with my DVD player / recorder,it is inaccessible.
Status: RESOLVED FIXED
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: 1
Hardware: x86_64 Linux
Priority: High critical
Target Milestone: ---
Assignee: Mageia Bug Squad
QA Contact:
URL:
Whiteboard:
Keywords: Triaged
: 1820 3064 (view as bug list)
Depends on:
Blocks:
 
Reported: 2011-06-02 18:04 CEST by David GEIGER
Modified: 2012-05-07 18:48 CEST (History)
22 users (show)

See Also:
Source RPM: mkinitrd
CVE:
Status comment:


Attachments
K3B error (33.54 KB, image/png)
2011-06-02 18:06 CEST, David GEIGER
Details
Nero Linux error (15.57 KB, image/png)
2011-06-02 18:07 CEST, David GEIGER
Details
Nero Linux error 2 (18.34 KB, image/png)
2011-06-02 18:07 CEST, David GEIGER
Details
Hardware cdrom (74.79 KB, image/png)
2011-06-02 18:08 CEST, David GEIGER
Details
cdrom mount (22.29 KB, image/png)
2011-06-02 18:09 CEST, David GEIGER
Details
Process "udevd" (66.57 KB, image/png)
2011-07-15 11:33 CEST, David GEIGER
Details
CPU frequency (32.30 KB, image/png)
2011-07-15 11:34 CEST, David GEIGER
Details
udev changelog up to version 173 (212.14 KB, text/plain)
2011-10-01 14:02 CEST, Samuel Verschelde
Details
dmesg (kernel-server-2.6.38.8-6mga1) (37.80 KB, application/octet-stream)
2011-10-12 07:04 CEST, David GEIGER
Details
file "dmesg from Mageia Beta 2" (44.98 KB, application/octet-stream)
2011-10-12 20:39 CEST, David GEIGER
Details
dmesg and syslog entries showing good drive on 1, bad on cauldron. (2.37 KB, text/plain)
2011-10-12 23:44 CEST, Dave Hodgins
Details
Patch showing changes I made to my initrd init script to get dvd working. (641 bytes, patch)
2011-10-13 00:34 CEST, Dave Hodgins
Details | Diff
unpackinitrd - bash script to unpack the initrd to a /tmp directory (237 bytes, text/plain)
2011-10-14 04:11 CEST, Dave Hodgins
Details
packinitrd - bash script to create initrd....test file (251 bytes, text/plain)
2011-10-14 04:17 CEST, Dave Hodgins
Details

Description David GEIGER 2011-06-02 18:04:22 CEST
Hello,

Description of problem:

-My reader / writer is inaccessible, since the installation of a Mageia 1 (official).
When I want to use K3B I get an error on the driver drive :

No optical drive found.
K3b did not find any optical device in your system.
Solution : Make sure HAL daemon is running, it is used by K3b for finding devices.

-While the reader is well recognized in the hardware.
-Nero Linux 4 have the same problem.
-What I find strange is that the tape device is dev / hdc, whereas usually it dev/sr0.
-Even if I insert a dvd or cd it nothing happens at all.

-I attached the screenshots.

Version-Release number of selected component (if applicable):

Mageia release 1 (Official) for x86_64 and Kernel-desktop-2.6.38.7-1.mga

My laptop is Acer Travelmate 7513 WSMI,and a burner TSSTcorp CD / DVDW TS-L632D


Again sorry for my bad English.

Goodbye to all the team Mageia.
Comment 1 David GEIGER 2011-06-02 18:06:07 CEST
Created attachment 504 [details]
K3B error
Comment 2 David GEIGER 2011-06-02 18:07:22 CEST
Created attachment 505 [details]
Nero Linux error
Comment 3 David GEIGER 2011-06-02 18:07:51 CEST
Created attachment 506 [details]
Nero Linux error 2
Comment 4 David GEIGER 2011-06-02 18:08:40 CEST
Created attachment 507 [details]
Hardware cdrom
Comment 5 David GEIGER 2011-06-02 18:09:39 CEST
Created attachment 508 [details]
cdrom mount
Comment 6 Ahmad Samir 2011-06-02 18:23:51 CEST
What's the output of:
ls -l /dev/cdrom
ls -l /dev/hdc
getfacl /dev/hdc

Keywords: (none) => NEEDINFO

Comment 7 David GEIGER 2011-06-02 18:29:51 CEST
 
[root@david ~]# ls -l /dev/cdrom
ls: impossible d'accéder à /dev/cdrom: Aucun fichier ou dossier de ce type


[root@david ~]# ls -l /dev/hdc
brw-rw-rw- 1 root disk 22, 0 juin   2 16:11 /dev/hdc


[root@david ~]# getfacl /dev/hdc
getfacl : suppression du premier « / » des noms de chemins absolus
# file: dev/hdc
# owner: root
# group: disk
user::rw-
group::rw-
other::rw-
Comment 8 Ahmad Samir 2011-06-02 18:42:45 CEST
OK, what's the output of:
ls -l /dev/*cd*
getfacl /dev/*cd*
Comment 9 David GEIGER 2011-06-02 19:09:02 CEST
[root@david ~]# ls -l /dev/*cd*
total 0
crw-rw---- 1 root cdrom 10, 62 juin   2 16:11 control


[root@david ~]# getfacl /dev/*cd*
getfacl : suppression du premier « / » des noms de chemins absolus
# file: dev/pktcdvd
# owner: root
# group: root
user::rwx
group::r-x
other::r-x
Ahmad Samir 2011-06-02 20:05:00 CEST

Keywords: NEEDINFO => Triaged
Source RPM: (none) => udev

Comment 10 David GEIGER 2011-06-03 21:07:25 CEST
Hello Team Mageia,


What happens next now ?
Will there be an update of udev?
Comment 11 Ahmad Samir 2011-06-04 00:17:05 CEST
Unfortunately there's no fix yet; the problem is identified, /dev/hdc on your system is an optical driver, but udev is making its group 'disk' whereas it should be 'cdrom', the fix would be for udev to assign the correct group.
Comment 12 Ahmad Samir 2011-06-04 00:23:17 CEST
A temporary/hackish workaround would be to chown the device node:
su
chown root:cdrom /dev/hdc
ln -s /dev/hdc /dev/dvd

see if this works.
Comment 13 David GEIGER 2011-06-04 15:02:12 CEST
Hello,
It does not work, still not available.



[root@david ~]# ls -l /dev/cdrom
ls: impossible d'accéder à /dev/cdrom: Aucun fichier ou dossier de ce type
[root@david ~]# ls -l /dev/hdc
brw-rw---- 1 root cdrom 22, 0 juin   4 14:43 /dev/hdc

[root@david ~]# getfacl /dev/hdc
getfacl : suppression du premier « / » des noms de chemins absolus
# file: dev/hdc
# owner: root
# group: cdrom
user::rw-
group::rw-
other::---



[root@david ~]# ls -l /dev/*cd*
total 0
crw-rw---- 1 root cdrom 10, 62 juin   4 14:43 control

[root@david ~]# getfacl /dev/*cd*
getfacl : suppression du premier « / » des noms de chemins absolus
# file: dev/pktcdvd
# owner: root
# group: root
user::rwx
group::r-x
other::r-x
Comment 14 Ahmad Samir 2011-06-04 23:07:43 CEST
Ah, yes, no ACL set.. try:
su
chown root:cdrom /dev/hdc
chmod 664 /dev/hdc
ln -s /dev/hdc /dev/dvd

note that this is a temporary workaround (and it won't survive a reboot...).
Comment 15 Bill Wilkinson 2011-06-06 15:19:14 CEST
I'm having the same problem, except my drive is mounted as hda, and the workaround (substituting hda for hdc in the instructions) isn't helping.  I had no problems with this under mdv2010.2, so that might help with tracking down the change.

CC: (none) => wrw105

Comment 16 Ahmad Samir 2011-06-16 18:21:27 CEST
*** Bug 1820 has been marked as a duplicate of this bug. ***

CC: (none) => gert-ereth

Ahmad Samir 2011-06-16 18:22:03 CEST

Priority: Normal => High
CC: (none) => pterjan, tmb
Severity: normal => major

Comment 17 Gert Ereth 2011-06-16 23:59:12 CEST
(In reply to comment #11)
> Unfortunately there's no fix yet; the problem is identified, /dev/hdc on your
> system is an optical driver, but udev is making its group 'disk' whereas it
> should be 'cdrom', the fix would be for udev to assign the correct group.

this does not happen with udev in Mandriva 2010.2 !
isn't it the same version in Mageia1 ???
Comment 18 Gert Ereth 2011-06-17 14:23:47 CEST
Mageia1 uses udev-166-5

OpenSuse 11.4 uses udev-166-5.1
with that Version the optical device works !
Comment 19 Ahmad Samir 2011-06-17 22:50:57 CEST
(IINM, it's not just udev, it's the whole stack, kernel/udev... etc).
Comment 20 Olivier Blin 2011-06-17 23:03:30 CEST
Is it an upgrade for a previous Mandriva release?
What is the result of this?
grep hdc /lib/udev/rules.d/*persistent*.rules

CC: (none) => mageia

Comment 21 Etienne Etienne 2011-06-25 15:34:44 CEST
Hello,
I have a similar problem on mageia1. But I can't find my dvd recorder in the "material explorer" and I can find it in the "local disk" in configuration center.

I have ubuntu in dual boot and no problem with it. And I didn't have any problem with mandriva 2010.2. It was not an upgrade from mandriva, I formated my disk / before install and keeped my /home.

I tried the tutorial from comments 12 and 14 but without success (my dvd recorder opened and closed alternatively without stopping).

CC: (none) => etienne15

Comment 22 Etienne Etienne 2011-06-25 15:36:15 CEST
I forgot to say that I installed mageia from a live cd i586 (and no x86-64).
Comment 23 Gert Ereth 2011-07-06 00:41:46 CEST
How and when will the problem be solved ?
Comment 24 Dave Hodgins 2011-07-06 03:36:34 CEST
From Comment 20
What is the result of this?
grep hdc /lib/udev/rules.d/*persistent*.rules

Also, from a google search, I found
http://ubuntuforums.org/showpost.php?p=10910206&postcount=13 which has
"What I believe the problem was is the SATA DVD RW Drive was plugged into one of my SATA 3.0 Ports and not the SATA 2.0 Port - when I changed it to the standard SATA Port it worked a charm"
This was for a different problem, but may be related.

CC: (none) => davidwhodgins

Comment 25 Gert Ereth 2011-07-06 10:23:03 CEST
(In reply to comment #24)
> From Comment 20
> What is the result of this?
> grep hdc /lib/udev/rules.d/*persistent*.rules

no (empty) output !
Comment 26 Etienne Etienne 2011-07-06 22:09:22 CEST
Idem
grep hdc /lib/udev/rules.d/*persistent*.rules
nothing is write and nothing happend
Comment 27 Stefan Lang 2011-07-08 13:33:10 CEST
Hello,
I'm having the same problem. This is a serious bug, is there any update available?
Thank you.

CC: (none) => grafilone

Comment 28 Thomas Backlund 2011-07-09 14:20:38 CEST
Can you try kernel-2.6.38.8-4.mga1 from Core Updates Testing media.

Does it make any difference ?
Comment 29 David GEIGER 2011-07-09 16:25:10 CEST
Hello,Thomas Backlund

I just tried the kernel-desktop-kernel-2.6.38.8-4.mga1.x86_64 and unfortunately it still does not work.Still the same problem.
I think the problem comes from the IDE DVD reader/writer.
Whereas with SATA2 drives it works ,and device dev/sr0
Device from IDE is dev/hdc.it's weird.

Thank you.
Comment 30 Bill Wilkinson 2011-07-09 18:06:36 CEST
I too have just tried 2.6.38.8, however the system hangs at "starting udev" on my Compaq Presario v6000 64 bit laptop....
Comment 31 Etienne Etienne 2011-07-09 20:59:54 CEST
Hello,
I added a second dvd reader/recorder (both are IDE).
When this one is alone and "master" drive, the problem is exactly the same as the first one (in the configuration center, it is not recognised in the "material explorer" but it is recognised in the "local drives"). And it is on /hdc.

But when the too drives are together, one "master" and one "slave", the slave one is recognised in the "material explorer" on /hdd (and both are recognised in "local drives".)

But anyone works !
It is a really unpleasant problem !

(I didn't try the new testing kernel because I don't want to have the same problem as Bill Wilkinson. Sorry)
Maurice Batey 2011-07-10 14:44:09 CEST

CC: (none) => maurice

Comment 32 David GEIGER 2011-07-15 11:31:38 CEST
Hello Mageia team,

Feedback with the desktop Kernel 2.6.38.8-4:

-Each time you start the PC, I have a process that uses a lot of CPU.
 And by the merest chance that the process "udevd" which is the cause.
 The frequency is 1600 Mhz CPU speed instead of 800Mhz,Then I am forced to "Kill 
 process" to resume normal operation of the processor.
 There I think the problem is getting worse.
-And of course the problem with my burner drive is still not resolved.

-Here is a screenshot of the process "udevd":
Comment 33 David GEIGER 2011-07-15 11:33:06 CEST
Created attachment 652 [details]
Process "udevd"
Comment 34 David GEIGER 2011-07-15 11:34:30 CEST
Created attachment 653 [details]
CPU frequency
Comment 35 Gert Ereth 2011-07-15 12:52:41 CEST
I have zhe same problem like David Geiger !

I made the update to Kernel 2.6.38.8-4.
Still the same problem with the optical device.
The process udevd takes 33% CPU.
Comment 36 Manuel Hiebel 2011-07-15 15:39:08 CEST
like bug 1954 ?
Comment 37 David GEIGER 2011-07-15 16:11:31 CEST
Hello Manuel Hiebel,

Yes it's like.

Except with the 2.6.38.7-1 kernel it worked.
Comment 38 Etienne Etienne 2011-07-25 21:31:37 CEST
I have the same problems like David Geziger and Gert Ereth.
Comment 39 Gert Ereth 2011-08-11 16:36:31 CEST
Hi,
is there any solution for this problem (now or in future) ?
Comment 40 Dave Hodgins 2011-09-13 05:15:03 CEST
Can you test with the 2.6.38.8-5 from Core Updates Testing?
Comment 41 David GEIGER 2011-09-13 18:03:56 CEST
Hello,

Tested on Mageia release 1 (Official) for x86_64,and Sorry for me it does not work better.

And also I have another problem (probably related)Thomas Backlund deals with:

https://bugs.mageia.org/show_bug.cgi?id=1954#c1
Comment 42 Mark Dawson Butterworth 2011-09-18 21:57:26 CEST
I am also having the problem of an invisible DVD RW drive. The drive was working with Mandriva 2010.1 but with Mageia does not mount when a CD/DVD is inserted.
I have no /dev/cdrom or /dev/dvd.
I also have no /dev/hdX devices at all (which seems a bit odd) so I can't use the workaround suggested above.

System background:
/dev/sda, sdb, and sdc are all SATA handled by /dev/mapper and are all working
/dev/sdd is a HighPoint RocketRaid622 and is working
/dev/st0 is a SCSI (really) tape drive and is working

Any ideas?

CC: (none) => mark.dawsonbutterworth

Comment 43 Gert Ereth 2011-09-26 09:55:31 CEST
Hi
it doesn't work better with Kernel 2.6.38.8-5.

But I tested it with Kernel 2.6.38.7 and udev-168-2 (Mandriva 2011) and there it works very well !

The problem seems to be the udev-166-5 !!!
Comment 44 Samuel Verschelde 2011-10-01 11:56:25 CEST
Raising severity to critical given the number of users affected.

CC: (none) => stormi
Severity: major => critical

Comment 45 David GEIGER 2011-10-01 13:16:33 CEST
Hello,

So ,I can confirm that Gert Ereth said :

I have tested the package "udev-168-2" on the  repo MDV_2011_Main_Release and actually it works very well.
ftp://distrib-coffee.ipsl.jussieu.fr/pub/linux/MandrivaLinux/official/2011/x86_64/media/main/release/

ftp://distrib-coffee.ipsl.jussieu.fr/pub/linux/MandrivaLinux/official/2011/i586/media/main/release/

At first I tried with the "kernel-desktop-2.6.38.7-1" ,it work :

$ cat /etc/issue
Mageia release 1 (Official) for x86_64
Kernel 2.6.38.7-desktop-1.mga on a Dual-processor x86_64 / \l

$ rpm -qa | grep udev
lib64gudev1.0_0-168-2
lib64udev0-168-2
udev-168-2
libudev0-168-2
system-config-printer-udev-1.3.1-4.1.mga1
libgudev1.0_0-168-2


And at second time I tried with the "kernel-server-2.6.38.8-6",it work too :

$ cat /etc/issue
Mageia release 1 (Official) for x86_64
Kernel 2.6.38.8-server-6.mga on a Dual-processor x86_64 / \l

$ rpm -qa | grep udev
lib64gudev1.0_0-168-2
lib64udev0-168-2
udev-168-2
libudev0-168-2
system-config-printer-udev-1.3.1-4.1.mga1
libgudev1.0_0-168-2

 The output of :

# LC_ALL=C ls -l /dev/cdrom
ls: cannot access /dev/cdrom: No such file or directory


# LC_ALL=C ls -l /dev/hdc
brw-rw----+ 1 root disk 22, 0 Oct  1 12:43 /dev/hdc


# LC_ALL=C getfacl /dev/hdc
getfacl: Removing leading '/' from absolute path names
# file: dev/hdc
# owner: root
# group: disk
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---


# LC_ALL=C getfacl /dev/*cd*
getfacl: Removing leading '/' from absolute path names
# file: dev/cdrom1
# owner: root
# group: disk
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---

# file: dev/cdrw1
# owner: root
# group: disk
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---

# file: dev/pktcdvd
# owner: root
# group: root
user::rwx
group::r-x
other::r-x

We can therefore really deduce a problem of "udev-166-5" of Mageia 1 deposits.
Can someone try to package "udev-168-2" MDV 2011?

It would be really great to finally fix this bug. Thank you very much.
Comment 46 Samuel Verschelde 2011-10-01 14:02:09 CEST
Created attachment 873 [details]
udev changelog up to version 173

udev changelog up to version 173, in case you can find a relevant entry in it for your problem
Comment 47 David GEIGER 2011-10-01 15:19:36 CEST
I continued my research on "udev" and I tested the version of deposits Cauldron.

$ rpm -qa | grep udev
udev-173-3.mga2
lib64gudev1.0_0-173-3.mga2
lib64udev0-173-3.mga2
libgudev1.0_0-173-3.mga2
system-config-printer-udev-1.3.1-4.1.mga1
libudev0-173-3.mga2

And two outbuildings proposed during the installation:

$ rpm -qa | grep gobject-introspection
gobject-introspection-1.30.0-1.mga2
$ rpm -qa | grep lib64glib-gir2.0
lib64glib-gir2.0-1.30.0-1.mga2


By cons, there I can confirm that it doesn't work. There is now again the problem with my DVD player.


The output of :

# LC_ALL=C ls -l /dev/cdrom
ls: cannot access /dev/cdrom: No such file or directory

# LC_ALL=C ls -l /dev/hdc
brw-rw---- 1 root disk 22, 0 Oct  1 14:44 /dev/hdc

# LC_ALL=C getfacl /dev/hdc
getfacl: Removing leading '/' from absolute path names
# file: dev/hdc
# owner: root
# group: disk
user::rw-
group::rw-
other::---

# LC_ALL=C ls -l /dev/*cd*
total 0
crw-rw---- 1 root cdrom 10, 62 Oct  1 14:44 control

# LC_ALL=C getfacl /dev/*cd*
getfacl: Removing leading '/' from absolute path names
# file: dev/pktcdvd
# owner: root
# group: root
user::rwx
group::r-x
other::r-x
Comment 48 Samuel Verschelde 2011-10-02 18:38:50 CEST
Adding more udev committers in CC. 

The problem has being narrowed down to:
- udev-166-5 from Mageia 1 : doesn't work
- udev-166-5.1 from openSuse : works
- udev-168-2 from Mandriva 2011 : works
- udev-173 from Mageia cauldron : doesn't work !

So it's highly possible that one of our patches is causing this.

CC: (none) => anssi.hannula, dmorganec, eugeni, mageia, thierry.vignaud

Comment 49 Thomas Backlund 2011-10-02 19:08:35 CEST
I'm on it...
Comment 50 Thomas Backlund 2011-10-02 19:30:10 CEST
There is now a 	udev-166-5.1.mga1 building and should soon show up in core/updates_testing

Please test it out and report back if it works
Comment 51 Bill Wilkinson 2011-10-02 21:34:52 CEST
Thomas,

So far:
CD ROM is now found.
Networking problems appear to be fixed (i.e. device ID shows up for wireless on network configuration).
Easytether now works (had been connecting but not passing data)

I'm in the process of copying a CD, reading has gone fine, but k3b has been calibrating for the write for >10 minutes.
Comment 52 David GEIGER 2011-10-03 20:17:43 CEST
Tested on Mageia release 1 (Official) for x86_64.

Excellent for me it work correctly too with "udev-166-5.1.mga1"

My CD ROM is now found too.

Thank you very much Thomas.



The output of :

# LC_ALL=C ls -l /dev/cdrom
ls: cannot access /dev/cdrom: No such file or directory

# LC_ALL=C ls -l /dev/hdc
brw-rw----+ 1 root disk 22, 0 Oct  3 18:18 /dev/hdc

# LC_ALL=C getfacl /dev/hdc
getfacl: Removing leading '/' from absolute path names
# file: dev/hdc
# owner: root
# group: disk
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---

# LC_ALL=C ls -l /dev/*cd*
lrwxrwxrwx 1 root root  3 Oct  3 18:18 /dev/cdrom1 -> hdc
lrwxrwxrwx 1 root root  3 Oct  3 18:18 /dev/cdrw1 -> hdc

/dev/pktcdvd:
total 0
crw-rw---- 1 root cdrom 10, 62 Oct  3 18:10 control


# LC_ALL=C getfacl /dev/*cd*
getfacl: Removing leading '/' from absolute path names
# file: dev/cdrom1
# owner: root
# group: disk
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---

# file: dev/cdrw1
# owner: root
# group: disk
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---

# file: dev/pktcdvd
# owner: root
# group: root
user::rwx
group::r-x
other::r-x
Comment 53 Gert Ereth 2011-10-04 00:12:00 CEST
I tested with udev-166-5.1.mga1.

Result:
the optical device is still connected to /dev/hdc
the device /dev/sr0 has not been created

the cdrom and dvdrom device are created, but linked to /dev/hdc

K3b starts without error.

Inserted cds or dvds were not shown in "Geräteüberwachung" (I don't know the english word)

it looks like that:

ls -l /dev/sr*
ls: Zugriff auf /dev/sr* nicht möglich: Datei oder Verzeichnis nicht gefunden

ls -l /dev/hd*
brw-rw----+ 1 root disk 22, 0 Okt  3 23:27 /dev/hdc

ls -l /dev/cd*
lrwxrwxrwx 1 root root 3 Okt  3 23:27 /dev/cdrom1 -> hdc
lrwxrwxrwx 1 root root 3 Okt  3 23:27 /dev/cdrw1 -> hdc

ls -l /dev/dvd*
lrwxrwxrwx 1 root root 3 Okt  3 23:27 /dev/dvd1 -> hdc
lrwxrwxrwx 1 root root 3 Okt  3 23:27 /dev/dvdrw1 -> hdc

but it should look like that:

ls -l /dev/sr*
brw-rw----+ 1 root cdrom 11, 0 Okt  3 23:00 /dev/sr0

ls -l /dev/hd*
ls: Zugriff auf /dev/hd* nicht möglich: Datei oder Verzeichnis nicht gefunden

ls -l /dev/cd*
lrwxrwxrwx 1 root root 3 Okt  3 23:00 /dev/cdrom -> sr0
lrwxrwxrwx 1 root root 3 Okt  3 23:00 /dev/cdrw -> sr0

ls -l /dev/dvd*
lrwxrwxrwx 1 root root 3 Okt  3 23:00 /dev/dvd -> sr0
lrwxrwxrwx 1 root root 3 Okt  3 23:00 /dev/dvdrw -> sr0

sorry, but the problem seems not to be fixed !
Comment 54 Dave Hodgins 2011-10-04 04:55:06 CEST
Gert, try deleting all of the non-comment lines from
/etc/udev/rules.d/70-persistent-cd.rules
and then reboot.
Comment 55 Gert Ereth 2011-10-04 10:19:36 CEST
(In reply to comment #54)
> Gert, try deleting all of the non-comment lines from
> /etc/udev/rules.d/70-persistent-cd.rules
> and then reboot.

I did it, but the result was the same as before !

No device sr0 and so on.

What a pity!
Thierry Vignaud 2011-10-04 17:41:46 CEST

Component: Installer => RPM Packages

Comment 56 David GEIGER 2011-10-09 21:12:45 CEST
(In reply to comment #53)
> I tested with udev-166-5.1.mga1.
> 
> Result:
> the optical device is still connected to /dev/hdc
> the device /dev/sr0 has not been created

It's true it is pity that the device /dev/sr0 has not been created.

Is it important or not? and is this normal?
Comment 57 David GEIGER 2011-10-09 21:26:01 CEST
Also, I meet a new problem with "udev-166-5.1.mga1":

I use "nerolinux-4.0.0.0-1.x86_64" which has been not working properly,when I start Nero It takes a long time to open, and the window is white.Impossible to do anything.I tried Nero from the command line but no results or error.

I can confirm that it is a conflict with "udev-166-5.1.mga1" because with "udev-166-5.mga1" it works well.

So what can we do?
Comment 58 Thomas Backlund 2011-10-09 22:10:12 CEST
What kind of hardvare are you using ?
(cpu/mb/chipset/ram/cdrom/...)

I'll try to grab some older hw from work to see if I can reproduce it locally...
(but I wont have time to do much until next weekend)
Comment 59 Gert Ereth 2011-10-10 10:04:53 CEST
On my Notebook the optical device is connected as "secondary master" (IDE) and
it does not work (no sr0, but /dev/hdc).

On a PC I had this configuration:
 optical device as "secondary master" and it did not work (no sr0, but /dev/hdc).

 Than I inserted a additional Harddisk. I connectes the Hardisk as "secondary
 master" and the optical device as "secondary slave".
 Now there it works very well. (sr0 is created).

So, perhaps the problem occurs only if the optical device is conected to in IDE-Master ?!
Comment 60 David GEIGER 2011-10-10 18:35:22 CEST
(In reply to comment #58)
> What kind of hardvare are you using ?
> (cpu/mb/chipset/ram/cdrom/...)


-ACER TravelMate 7513 Wsmi Laptop

<<CD/DVD burners:>>
Identification
Disk identifier: âTSSTcorpCD/DVDW TS-L632D
Media class: âcdrom
Bus identification
Target id number: â0
Connection
Bus: âide
Channel: âsecondary
Target id number: â0
Device
Old device file: â/dev/hdc
Misc
Drive capacity: âburner DVD 
Disk controller: â0

<<Processors:>>
Identification
Processor ID: â1
Vendor: âAuthenticAMD
Model name: âAMD Turion(tm) 64 X2 Mobile Technology TL-52
Cpuid family: â15
Model: â72
Model stepping: â2
Cpuid level: â1
Performances
Frequency (MHz): â800.000
Cache size: â512 KB
Bogomips: â1607.35
Cores
Cores: â2
Core ID: â0
Physical ID: â0
ACPI ID: â0

Identification
Processor ID: â2
Vendor: âAuthenticAMD
Model name: âAMD Turion(tm) 64 X2 Mobile Technology TL-52
Cpuid family: â15
Model: â72
Model stepping: â2
Cpuid level: â1
Performances
Frequency (MHz): â800.000
Cache size: â512 KB
Bogomips: â1607.35
Cores
Cores: â2
Core ID: â1
Physical ID: â0
ACPI ID: â1

<<Ram :>>
2 X 2Go DDR2 SO-DIMM

<<VideoCard :>>
Identification
Vendor: ânVidia Corporation
Description: âG72M [Quadro NVS 110M/GeForce Go 7300]
Media class: âVGA compatible controller
Connection
Bus: âPCI Express
PCI domain: â0
Bus PCI #: â3
PCI device #: â0
PCI function #: â0
PCI revision: â0xa1
Vendor ID: â0x10de
Device ID: â0x01d7
Sub vendor ID: â0x1025
Sub device ID: â0x0112
Misc
Module: âCard:NVIDIA GeForce 6100 to GeForce 360

<<Hard Disk :>>
Disk identifier: âWestern Digital Corp. WD3200BEKT-0
Media class: âhd
Bus identification
Target id number: â0
Connection
Bus: âSCSI
Channel: â0
Logical unit number: â0
Target id number: â0
Device
Old device file: â/dev/sda
Partitions
Primary partitions: â2 (sda1, sda2)
Extended partitions: â6 (sda5, sda6, sda7, sda8, sda9, sda10)
Misc
Geometry: â38913/255/63 (CHS)
Disk controller: â0
Module: âsata_nv
Comment 61 Dave Hodgins 2011-10-10 21:26:59 CEST
Are the ide cables 40 wire or 80 wire?

I'm wondering if this is related to
https://groups.google.com/group/alt.os.linux.mandriva/msg/e1c621e74fdc8e2f?hl=en
Comment 62 Gert Ereth 2011-10-11 08:59:14 CEST
(In reply to comment #61)
> Are the ide cables 40 wire or 80 wire?
> 
40 wire

> I'm wondering if this is related to
so do I !

@David Geiger,
youe have one Hardisk and one DVD-Burner ?
if the DVD.Burner is connected as "Secondary Master" then would you please try to connect it as "Primary Slave", if it is possible (don't forget the jumper) ?
Does it change something ?
Comment 63 David GEIGER 2011-10-11 18:34:29 CEST
(In reply to comment #62)
 
> @David Geiger,
> youe have one Hardisk and one DVD-Burner ?
>
Yes I have just one Hardisk and one DVD-Burner.


> if the DVD.Burner is connected as "Secondary Master" then would you please try
> to connect it as "Primary Slave", if it is possible (don't forget the jumper) ?
> Does it change something ?
>
No, it's no possible because this is a PC LAPTOP. I can't change this.
Well, not to my knowledge??
Comment 64 Dave Hodgins 2011-10-11 18:40:45 CEST
(In reply to comment #63)
> (In reply to comment #62)

> No, it's no possible because this is a PC LAPTOP. I can't change this.
> Well, not to my knowledge??

Can you try the fix from comment 61?

Add a line to /etc/modprobe.conf with ... 
 options pata_sis libata.force=2:udma33

Then reboot.
Comment 65 David GEIGER 2011-10-11 19:43:23 CEST
(In reply to comment #64)
> 
> Can you try the fix from comment 61?
> 
> Add a line to /etc/modprobe.conf with ... 
>  options pata_sis libata.force=2:udma33
> 
> Then reboot.

So,tested with the line to /etc/modprobe.conf with ... 
>  options pata_sis libata.force=2:udma33

It doesn't work ,the device is always dev/hdc.

options snd-ac97-codec power_save=1
install scsi_hostadapter /sbin/modprobe sata_nv; /bin/true
remove tifm_7xx1 /sbin/modprobe -r mmc_block; /sbin/modprobe -r tifm_sd; /sbin/modprobe --first-time -r --ignore-remove tifm_7xx1
install tifm_7xx1 /sbin/modprobe --first-time --ignore-install tifm_7xx1 && { /sbin/modprobe mmc_block; /sbin/modprobe tifm_sd; /bin/true; }
install usb-interface /sbin/modprobe ehci_hcd; /sbin/modprobe ohci_hcd; /bin/true
install ide-controller /sbin/modprobe amd74xx; /bin/true
options pata_sis libata.force=2:udma33
Comment 66 Dave Hodgins 2011-10-11 22:43:05 CEST
Try booting with the kernel parameter libata.dma=0 which will force all i/o
to pio mode.  I expect the system will be slower, but that will confirm for
sure if it is a speed problem.

The attach /var/log/dmesg.
Comment 67 David GEIGER 2011-10-11 23:02:25 CEST
(In reply to comment #66)
> Try booting with the kernel parameter libata.dma=0 which will force all i/o
> to pio mode.  I expect the system will be slower, but that will confirm for
> sure if it is a speed problem.
> 
> The attach /var/log/dmesg.

Sorry,but I do not know how to do this? Can you give me more details ,please?
Comment 68 Dave Hodgins 2011-10-12 03:20:12 CEST
At boot up, in grub, press the e key, to go into edit mode.
Press the e key again, to edit the line with the kernel parameters.
Press the end key, to get the cursor to the end of the line.
Add a space then libata.dma=0
Press enter, to end the editing of the line.
Press the b key to boot.
Comment 69 David GEIGER 2011-10-12 07:04:15 CEST
Created attachment 942 [details]
dmesg (kernel-server-2.6.38.8-6mga1)

So,tested with the kernel parameter libata.dma=0 at boot.

Indeed the system has become slower.
Comment 70 Dave Hodgins 2011-10-12 10:43:22 CEST
Thanks for the test.

While I strongly recommend replacing the 40 conductor cable with an 80 conductor
cable, that does not appear to be the cause of this problem.

I'm out of ideas for now, on what else to try, other then changing
the jumpers on the drive from master to slave.
Comment 71 David GEIGER 2011-10-12 20:39:41 CEST
Created attachment 945 [details]
file "dmesg from Mageia Beta 2"

So I pushed a little more research :

I installed "Mageia_Beta_2_x86_64_DVD" for comparison. So I can confirm that the problem is occurring from a "Mageia 1".

With Mageia Beta 2 ,my DVD-burner is well created on device "Sr0".
And everything works fine. NeroLinux work fine too.

The output of :

$ cat /etc/issue
Mageia release 1 (Cauldron) for x86_64
Kernel 2.6.38.4-desktop-1.mga on a Dual-processor x86_64 / \l

$ rpm -qa | grep udev
udev-166-4.mga1
lib64udev0-166-4.mga1
lib64gudev1.0_0-166-4.mga1
system-config-printer-udev-1.3.1-3.mga1

<<CD/DVD burners>>
Identification
Vendor: âTSSTcorp
Description: âCD/DVDW TS-L632D
Disk identifier: âTSSTcorp CD/DVDW TS-L632D
Media class: âcdrom
Bus identification
Target id number: â0
Connection
Bus: âSCSI
Channel: â0
Logical unit number: â0
Target id number: â0
Device
Old device file: â/dev/sr0
Misc
Drive capacity: âburner DVD 
Disk controller: â3
Module: âpata_amd



# LC_ALL=C ls -l /dev/cdrom
lrwxrwxrwx 1 root root 3 Oct 12 19:01 /dev/cdrom -> sr0

# LC_ALL=C ls -l /dev/hdc
ls: cannot access /dev/hdc: No such file or directory

# LC_ALL=C getfacl /dev/hdc
getfacl: /dev/hdc: No such file or directory

# LC_ALL=C ls -l /dev/*cd*
lrwxrwxrwx 1 root root  3 Oct 12 19:04 /dev/cdrom -> sr0
lrwxrwxrwx 1 root root  3 Oct 12 19:04 /dev/cdrw -> sr0
lrwxrwxrwx 1 root root  3 Oct 12 19:04 /dev/scd0 -> sr0

/dev/pktcdvd:
total 0
crw-rw---- 1 root cdrom 10, 62 Oct 12 18:57 control

# LC_ALL=C getfacl /dev/*cd*
getfacl: Removing leading '/' from absolute path names
# file: dev/cdrom
# owner: root
# group: cdrom
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---

# file: dev/cdrw
# owner: root
# group: cdrom
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---

# file: dev/pktcdvd
# owner: root
# group: root
user::rwx
group::r-x
other::r-x

# file: dev/scd0
# owner: root
# group: cdrom
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---

I think probably a good trail to follow. What do you think?
Comment 72 David GEIGER 2011-10-12 20:47:48 CEST
Still something that concerns me , Comparing these two results:

$ rpm -qa | grep udev
udev-166-4.mga1
lib64udev0-166-4.mga1
lib64gudev1.0_0-166-4.mga1
system-config-printer-udev-1.3.1-3.mga1

And  

$ rpm -qa | grep udev
lib64gudev1.0_0-166-5.1.mga1
libudev0-166-5.1.mga1
lib64udev0-166-5.1.mga1
system-config-printer-udev-1.3.1-4.1.mga1
udev-166-5.1.mga1
libgudev1.0_0-166-5.1.mga1

Is it normal for the two 32-bit lib on a 64bit system?

-libgudev1.0_0-166-5.1.mga1
-libudev0-166-5.1.mga1
Comment 73 Dave Hodgins 2011-10-12 23:44:17 CEST
Created attachment 947 [details]
dmesg and syslog entries showing good drive on 1, bad on cauldron.

I just tried cauldron with server 3.1.0rc9-1.mga2, and have the problem myself
there.  My dvd is working in Mageia 1, with all updates including testing
installed, but is no longer working in Cauldron.  I tried switching it
from master to slave, with no affect.  I have 80 wire ide cables.
Comment 74 Dave Hodgins 2011-10-13 00:34:04 CEST
Created attachment 948 [details]
Patch showing changes I made to my initrd init script to get dvd working.

After looking at the dmesg output from Mageia 1 where the dvd works, and
Cauldron, where it doesn't, it looked likely to be a problem with the order
the modules were loaded.

I modified the init script in the initrd file, to load pata_via before pata_acpi, and I have my dvd drive working again, in Cauldron.

I expect the problem others are having in Mageia 1 are being caused by a similar
prder of modules loading problem.
Dave Hodgins 2011-10-13 00:35:09 CEST

Source RPM: udev => mkinitrd

Comment 75 Dave Hodgins 2011-10-13 00:59:38 CEST
(In reply to comment #71)

Looks to me like the same problem I had in cauldron.

In beta 2, the dvd drive was handled by pata_amd.

In Mageia 1, the dvd drive was taken by pata_acpi, and pata_amd was
either not loaded, or didn't do anything.

Can you try modifying the init script in the initrd file to load
pata_amd before pata_acpi, to confirm it fixes the probelm?
Comment 76 David GEIGER 2011-10-13 14:44:31 CEST
Where is the file initrd? and how to edit it?

I found this but I can't edit it with kwrite in ROOT :

<< /boot/initrd-2.6.38.8-server-6.mga.img >>

echo "Loading usbhid module"
modprobe -q usbhid
echo "Loading ehci-hcd module"
modprobe -q ehci-hcd
echo "Loading ohci-hcd module"
modprobe -q ohci-hcd
echo "Loading uhci-hcd module"
modprobe -q uhci-hcd
mount -t usbfs /proc/bus/usb /proc/bus/usb
echo "Loading ext4 module"
modprobe -q ext4
echo "Loading crc-t10dif module"
modprobe -q crc-t10dif
echo "Loading scsi_mod module"
modprobe -q scsi_mod
echo "Loading sd_mod module"
modprobe -q sd_mod
echo "Loading libata module"
modprobe -q libata
echo "Loading sata_nv module"
modprobe -q sata_nv
echo "Loading pata_acpi module"
modprobe -q pata_acpi
echo "Loading ide-core module"
modprobe -q ide-core
echo "Loading ide-disk module"
modprobe -q ide-disk
echo "Loading ide-pci-generic module"
modprobe -q ide-pci-generic
echo "Loading ata_generic module"
modprobe -q ata_generic
echo "Loading amd74xx module"
modprobe -q amd74xx
echo waiting for device sda8 to appear (timeout 1min)


And comparative this with Mandriva 2010.2 :

/media/94bdf042-3719-11dd-93c1-dfc6c46ebb97/boot/initrd-2.6.33.7-server-2mnb.img

echo "Loading usbhid module"
modprobe -q usbhid
echo "Loading ehci-hcd module"
modprobe -q ehci-hcd
echo "Loading ohci-hcd module"
modprobe -q ohci-hcd
echo "Loading uhci-hcd module"
modprobe -q uhci-hcd
mount -t usbfs /proc/bus/usb /proc/bus/usb
echo "Loading ext3 module"
modprobe -q ext3
echo "Loading crc-t10dif module"
modprobe -q crc-t10dif
echo "Loading scsi_mod module"
modprobe -q scsi_mod
echo "Loading sd_mod module"
modprobe -q sd_mod
echo "Loading libata module"
modprobe -q libata
echo "Loading sata_nv module"
modprobe -q sata_nv
echo "Loading pata_acpi module"
modprobe -q pata_acpi
echo "Loading ide-core module"
modprobe -q ide-core
echo "Loading ide-disk module"
modprobe -q ide-disk
echo "Loading ide-pci-generic module"
modprobe -q ide-pci-generic
echo "Loading ata_generic module"
modprobe -q ata_generic
echo "Loading pata_amd module"
modprobe -q pata_amd
echo waiting for device sda5 to appear (timeout 1min)
Comment 77 David GEIGER 2011-10-13 19:11:50 CEST
Here is the file initrd from Mageia Beta 2 :


echo "Loading usbhid module"
modprobe -q usbhid
echo "Loading ehci-hcd module"
modprobe -q ehci-hcd
echo "Loading ohci-hcd module"
modprobe -q ohci-hcd
echo "Loading uhci-hcd module"
modprobe -q uhci-hcd
mount -t usbfs /proc/bus/usb /proc/bus/usb
echo "Loading ext4 module"
modprobe -q ext4
echo "Loading crc-t10dif module"
modprobe -q crc-t10dif
echo "Loading scsi_mod module"
modprobe -q scsi_mod
echo "Loading sd_mod module"
modprobe -q sd_mod
echo "Loading libata module"
modprobe -q libata
echo "Loading sata_nv module"
modprobe -q sata_nv
echo "Loading pata_acpi module"
modprobe -q pata_acpi
echo "Loading ide-core module"
modprobe -q ide-core
echo "Loading ide-disk module"
modprobe -q ide-disk
echo "Loading ide-pci-generic module"
modprobe -q ide-pci-generic
echo "Loading ata_generic module"
modprobe -q ata_generic
echo "Loading pata_amd module"
modprobe -q pata_amd
echo waiting for device sda8 to appear (timeout 1min)

We can deduce that the module "pata_amd" is missing on Mageia 1 in the kernel 2.6.38.8 .
Comment 78 Dave Hodgins 2011-10-14 04:11:21 CEST
Created attachment 958 [details]
unpackinitrd - bash script to unpack the initrd to a /tmp directory

Run the attached script, then
cd initrddirname="/tmp/initrd.$(uname -r)
then edit the init file, and add/move the lines
echo "Loading pata_amd module"
modprobe -q pata_amd
to be before the lines for pata_acpi.
Comment 79 Dave Hodgins 2011-10-14 04:17:18 CEST
Created attachment 959 [details]
packinitrd - bash script to create initrd....test file

After making the changes in the prior commont, run the attached script
to create the initrd image in /boot.  Note that it has the .test on
the end.

Then edit /boot/grub/menu.lst, and copy the boot stanza for the running
kernel.

In the copy, add the .test to the end of the initrd file name, and
the title.

Then reboot, and select the test stanza.
Comment 80 David GEIGER 2011-10-14 19:11:16 CEST
So, I have run the script unpackinitrd - bash ,then edit the "init" file with kwrite in ROOT.
I have add (no move because the lines doesn't exist) the lines:
echo "Loading pata_amd module"
modprobe -q pata_amd
before the lines for pata_acpi.

Then I have run the script packinitrd - bash to create initrd....test file.
Then I have renamed the file "initrd-2.6.38.8-server-6.mga.img" to "initrd-2.6.38.8-server-6.mga.img.old" in the /boot/
to use the new file "initrd-2.6.38.8-server-6.mga.img.test" by renaming it "initrd-2.6.38.8-server-6.mga.img"

Now I have reboot the PC and there still nothing, the DVD-Burner is still 
/dev/hdc.

By cons, there is something that concerns me:

In the new files "initrd-2.6.38.8-server-6.mga.img.test" top of the list there is this: (no lines echo "Loading pata_amd module"........)

echo "Loading usbhid module"
modprobe -q usbhid
echo "Loading ehci-hcd module"
modprobe -q ehci-hcd
echo "Loading ohci-hcd module"
modprobe -q ohci-hcd
echo "Loading uhci-hcd module"
modprobe -q uhci-hcd
mount -t usbfs /proc/bus/usb /proc/bus/usb
echo "Loading ext4 module"
modprobe -q ext4
echo "Loading crc-t10dif module"
modprobe -q crc-t10dif
echo "Loading scsi_mod module"
modprobe -q scsi_mod
echo "Loading sd_mod module"
modprobe -q sd_mod
echo "Loading libata module"
modprobe -q libata
echo "Loading sata_nv module"
modprobe -q sata_nv
echo "Loading pata_acpi module"
modprobe -q pata_acpi
echo "Loading ide-core module"
modprobe -q ide-core
echo "Loading ide-disk module"
modprobe -q ide-disk
echo "Loading ide-pci-generic module"
modprobe -q ide-pci-generic
echo "Loading ata_generic module"
modprobe -q ata_generic
echo "Loading amd74xx module"
modprobe -q amd74xx
echo waiting for device sda8 to appear (timeout 1min)

And at the end of the file there is a second identical list:(but there is the line echo "Loading pata_amd module".........)

echo "Loading usbhid module"
modprobe -q usbhid
echo "Loading ehci-hcd module"
modprobe -q ehci-hcd
echo "Loading ohci-hcd module"
modprobe -q ohci-hcd
echo "Loading uhci-hcd module"
modprobe -q uhci-hcd
mount -t usbfs /proc/bus/usb /proc/bus/usb
echo "Loading ext4 module"
modprobe -q ext4
echo "Loading crc-t10dif module"
modprobe -q crc-t10dif
echo "Loading scsi_mod module"
modprobe -q scsi_mod
echo "Loading sd_mod module"
modprobe -q sd_mod
echo "Loading libata module"
modprobe -q libata
echo "Loading sata_nv module"
modprobe -q sata_nv
echo "Loading pata_amd module"
modprobe -q pata_amd
echo "Loading pata_acpi module"
modprobe -q pata_acpi
echo "Loading ide-core module"
modprobe -q ide-core
echo "Loading ide-disk module"
modprobe -q ide-disk
echo "Loading ide-pci-generic module"
modprobe -q ide-pci-generic
echo "Loading ata_generic module"
modprobe -q ata_generic
echo "Loading amd74xx module"
modprobe -q amd74xx
echo waiting for device sda8 to appear (timeout 1min)

It is rather strange is not it?
Comment 81 Dave Hodgins 2011-10-15 03:51:50 CEST
(In reply to comment #80)
> So, I have run the script unpackinitrd - bash ,then edit the "init" file with
> kwrite in ROOT.
> I have add (no move because the lines doesn't exist) the lines:
> echo "Loading pata_amd module"
> modprobe -q pata_amd
> before the lines for pata_acpi.

Since the pata_amd module isn't there, there's going to be a bit more to
it, as the module isn't in the initrd directories, or the modules.dep.

Try running mkinitrd as in ...
mkinitrd -f --preload pata_amd /boot/initrd-2.6.38.8-desktop-6.1.mga.img 2.6.38.8-desktop-6.1.mga

Then run lsinitrd /boot/initrd-2.6.38.8-desktop-6.1.mga.img to confirm the
pata_amd module is going to be loaded before the pata_acpi module.  If it
isn't then redo the unpackinitrd/edit/packinitrd.
Comment 82 David GEIGER 2011-10-15 08:01:21 CEST
Excellent, now it work fine. The device for the DVD/burner is dev/sr0
No error message with K3B and Nero Linux 4. It work perfectly.

--------------------------------------------------------------------------
# mkinitrd -f --preload pata_amd /boot/initrd-2.6.38.8-server-6.mga.img 2.6.38.8-server-6.mga
# lsinitrd /boot/initrd-2.6.38.8-server-6.mga.img
echo Creating character device nodes.
mkchardevs
echo "Loading pata_amd module"
modprobe -q pata_amd
echo "Loading usbhid module"
modprobe -q usbhid
echo "Loading ehci-hcd module"
modprobe -q ehci-hcd
echo "Loading ohci-hcd module"
modprobe -q ohci-hcd
echo "Loading uhci-hcd module"
modprobe -q uhci-hcd
mount -t usbfs /proc/bus/usb /proc/bus/usb
echo "Loading ext4 module"
modprobe -q ext4
echo "Loading crc-t10dif module"
modprobe -q crc-t10dif
echo "Loading scsi_mod module"
modprobe -q scsi_mod
echo "Loading sd_mod module"
modprobe -q sd_mod
echo "Loading libata module"
modprobe -q libata
echo "Loading sata_nv module"
modprobe -q sata_nv
echo "Loading pata_acpi module"
modprobe -q pata_acpi
echo "Loading ide-core module"
modprobe -q ide-core
echo "Loading ide-disk module"
modprobe -q ide-disk
echo "Loading ide-pci-generic module"
modprobe -q ide-pci-generic
echo "Loading ata_generic module"
modprobe -q ata_generic
echo "Loading amd74xx module"
modprobe -q amd74xx
echo waiting for device sda8 to appear (timeout 1min)


-------------------------------------------------------------------------------
Identification
Fabricant: âTSSTcorp
Description: âCD/DVDW TS-L632D
Identifiant du disque: âTSSTcorp CD/DVDW TS-L632D
Classe de matériel: âcdrom
Identification de Bus
Numéro d'id cible: â0
Périphérique
Ancien nom de périphérique: â/dev/sr0
Connexion
Bus: âSCSI
Canal: â0
Numéro d'unité logique: â0
Numéro d'id cible: â0
Divers
Spécificités du lecteur: âgraveur DVD 
Contrôleur disque: â1
Module: âpata_amd
-------------------------------------------------------------------------------

And then I use the package 'udev-166-5 "and not" udev-166-5.1 "
I think the update of udev in the repo Core_Updates_Testing can be removed.
there is no need. What do you think?

I think the problem is found, thank you Dave.
Comment 83 Gert Ereth 2011-10-15 17:12:13 CEST
(In reply to comment #81)
> Since the pata_amd module isn't there, there's going to be a bit more to
> it, as the module isn't in the initrd directories, or the modules.dep.
> 
> Try running mkinitrd as in ...
> mkinitrd -f --preload pata_amd /boot/initrd-2.6.38.8-desktop-6.1.mga.img
> 2.6.38.8-desktop-6.1.mga
> 
I can confirm, this fixed the problem !!!

Now I have /dev/sr0 and everything works fine !

Thank you Dave !!!

can you please (shortly) explain what the module "pata_amd" does ?!
(just to understand what I (we) have done)
Comment 84 Thomas Backlund 2011-10-15 17:31:07 CEST
pata_amd is the ata based driver for controlling the disc controller.

by using --preload you force it to init and take control before the ide based amd74xx driver can take control of the same controller.

can both of you that had this problem provide the output of:

lspcidrake -v | grep STORAGE_IDE 

so we can fix ldetect-lst to prioritize pata_amd for your systems.

I'll nuke the udev package from testing
Comment 85 David GEIGER 2011-10-15 17:37:06 CEST
The output of :

$ lspcidrake -v | grep STORAGE_IDE
sata_nv         : nVidia Corporation|MCP51 Serial ATA Controller [STORAGE_IDE] (vendor:10de device:0266 subv:1025 subd:0112) (rev: f1)
amd74xx         : nVidia Corporation|MCP51 IDE [STORAGE_IDE] (vendor:10de device:0265 subv:1025 subd:0112) (rev: f1)
Comment 86 Dave Hodgins 2011-10-15 19:31:32 CEST
$ lspcidrake -v | grep STORAGE_IDE
pata_via        : VIA Technologies, Inc.|VT82C586A/B/VT82C686/A/B/VT823x/A/C PIPC Bus Master IDE [STORAGE_IDE] (vendor:1106 device:0571 subv:1019 subd:1899) (rev: 06)

Note the differences between to two cases.

First case, affects Mageia 1, where pata_amd is not being loaded at all.

Second case affects Cauldron only, where pata_via is being loaded after
pata_acpi, but needs to be loaded before it.
Comment 87 Gert Ereth 2011-10-16 13:51:17 CEST
$ lspcidrake -v | grep STORAGE_IDE
sata_nv         : nVidia Corporation|MCP51 Serial ATA Controller [STORAGE_IDE] (vendor:10de device:0266 subv:17c0 subd:4076) (rev: f1)
amd74xx         : nVidia Corporation|MCP51 IDE [STORAGE_IDE] (vendor:10de device:0265 subv:17c0 subd:4076) (rev: f1)
Comment 88 Bill Wilkinson 2011-10-16 20:10:47 CEST
To the Mageia team:

Is there going to be a package of init scripts to test or should I start getting under the hood?

Thanks!
Comment 89 Dave Hodgins 2011-10-17 00:12:38 CEST
(In reply to comment #88)
 
> Is there going to be a package of init scripts to test or should I start
> getting under the hood?

Now that we know what the cause of the problem is, I expect there will be
an update fairly soon.

Please post the output of
lspcidrake -v | grep STORAGE_IDE
as the info may help for the update, or we use it to provide a work around
for you.
Comment 90 Bill Wilkinson 2011-10-17 04:28:15 CEST
Dave,

Here you go-Thanks to you and Thomas for all your help!

sata_nv         : nVidia Corporation|MCP51 Serial ATA Controller [STORAGE_IDE] (vendor:10de device:0266 subv:103c subd:30b7) (rev: f1)
amd74xx         : nVidia Corporation|MCP51 IDE [STORAGE_IDE] (vendor:10de device:0265 subv:103c subd:30b7) (rev: f1)
Comment 91 Thomas Backlund 2011-10-17 06:31:10 CEST
For those of you that using pata_amd helps, look in /etc/modprobe.conf and change 

install ide-controller /sbin/modprobe amd74xx; /bin/true

to:

install ide-controller /sbin/modprobe pata_amd; /bin/true
Comment 92 Thomas Backlund 2011-10-17 07:16:22 CEST
Ok, now this gets tricky...

The reason for amd74xx being used is that we enforce it due to an earlier bugreport:
https://bugs.mageia.org/show_bug.cgi?id=933

I need to think some more about this one.
Comment 93 Thomas Backlund 2011-10-17 08:01:42 CEST
Ok, I have now submitted ldetect-lst-0.1.291-9.1.mga1 to core/updates_testing

Can you confirm that after installing it the output of: 

lspcidrake -v | grep STORAGE_IDE

will list pata_amd, not amd74xx
Comment 94 Bill Wilkinson 2011-10-17 14:42:46 CEST
I can confirm pata_amd and not amd74xx is working.

A test burn is now giving me "cdrecord does not have permission to open the device"

I've gone into manage users and explicitly enabled the cdwriter group to my ID, but still no permission.
Comment 95 David GEIGER 2011-10-17 17:47:08 CEST
Tested the package "ldetect-lst-0.1.291-9.1.mga1" to core/updates_testing
and now the output is:

$ lspcidrake -v | grep STORAGE_IDE
sata_nv         : nVidia Corporation|MCP51 Serial ATA Controller [STORAGE_IDE] (vendor:10de device:0266 subv:1025 subd:0112) (rev: f1)
pata_amd        : nVidia Corporation|MCP51 IDE [STORAGE_IDE] (vendor:10de device:0265 subv:1025 subd:0112) (rev: f1)

So we do agree, the tests are to do with the official file initrd in /boot/ and not with the "test stanza" created above in comment 78 & 79 ?
Because this package doesn't fix this bug.
Comment 96 Thomas Backlund 2011-10-17 18:03:01 CEST
For already installed systems the final fix is to manually edit
/etc/modprobe.conf and replace amd74xx with pata_amd.

That way it will get used automatically on next update kernel install.

The change to ldetect-lst is to make sure new installs get the correct module used directly.
Comment 97 David GEIGER 2011-10-17 18:29:47 CEST
Ok Thomas ,

-So, just once I change the file "modprobe.conf" ,I replace :

install ide-controller /sbin/modprobe amd74xx; /bin/true
to:
install ide-controller /sbin/modprobe pata_amd; /bin/true

-I install the update package "ldetect-lst-0.1.291-9.1.mga1"

-I install the next update kernel

-And then, finally, the bug is fixed permanently?


For I must relate this operation to French users on the forum MLO.
Comment 98 David GEIGER 2011-10-17 20:47:59 CEST
In the end, we didn't need the module "amd74xx" ? He has served as nothing?
Comment 99 Thomas Backlund 2011-10-17 20:54:09 CEST
amd74xx is the (old) ide based driver for your disc controller.

pata_amd is the new ata based driver for your disc controller, coming with more features and better performance.
Comment 100 David GEIGER 2011-10-17 22:32:42 CEST
# LC_ALL=C ls -l /dev/*cd*
lrwxrwxrwx 1 root root  3 Oct 17 20:43 /dev/cdrom1 -> sr0
lrwxrwxrwx 1 root root  3 Oct 17 20:43 /dev/cdrw1 -> sr0
lrwxrwxrwx 1 root root  3 Oct 17 20:43 /dev/scd0 -> sr0

/dev/pktcdvd:
total 0
crw-rw---- 1 root cdrom 10, 62 Oct 17 20:24 control


# LC_ALL=C getfacl /dev/*cd*
getfacl: Removing leading '/' from absolute path names
# file: dev/cdrom1
# owner: root
# group: cdrom
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---

# file: dev/cdrw1
# owner: root
# group: cdrom
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---

# file: dev/pktcdvd
# owner: root
# group: root
user::rwx
group::r-x
other::r-x

# file: dev/scd0
# owner: root
# group: cdrom
user::rw-
user:david:rw-
group::rw-
mask::rw-
other::---


# LC_ALL=C ls -l /dev/dvd*
lrwxrwxrwx 1 root root 3 Oct 17 20:43 /dev/dvd1 -> sr0
lrwxrwxrwx 1 root root 3 Oct 17 20:43 /dev/dvdrw1 -> sr0


# LC_ALL=C ls -l /dev/sr*
brw-rw----+ 1 root cdrom 11, 0 Oct 17 20:43 /dev/sr0


# LC_ALL=C ls -l /dev/cdrom
ls: cannot access /dev/cdrom: No such file or directory


Why the output of "ls -l /dev/cdrom" returns: 
cannot access /dev/cdrom: No such file or directory

Is it normal ,or not ?


With this ,it's ok :

# LC_ALL=C ls -l /dev/cdrom1
lrwxrwxrwx 1 root root 3 Oct 17 20:43 /dev/cdrom1 -> sr0
Comment 101 Dave Hodgins 2011-10-17 23:19:11 CEST
(In reply to comment #100)
> # LC_ALL=C ls -l /dev/*cd*
> lrwxrwxrwx 1 root root  3 Oct 17 20:43 /dev/cdrom1 -> sr0
> lrwxrwxrwx 1 root root  3 Oct 17 20:43 /dev/cdrw1 -> sr0
> lrwxrwxrwx 1 root root  3 Oct 17 20:43 /dev/scd0 -> sr0

At some point the drive changed location on the pci bus,
(slave/master change for example), so the system sees
it as a new drive.

The name /dev/cdrom is reserved for when the drive is in
the old location.

To get the drive assigned as /dev/cdrom, instead of as
/dev/cdrom1, delete all of the non-comment lines from
/etc/udev/rules.d/70 and reboot.
Comment 102 Dave Hodgins 2011-10-17 23:25:53 CEST
Sorry, copy/paste error.

The file is /etc/udev/rules.d/70-persistent-cd.rules
Comment 103 David GEIGER 2011-10-17 23:44:24 CEST
Yes ,thank you Dave .I had guessed.

So,for me it's good and Ok.

I think this bug is fixed ,what do you think ?

In any case, thank you very much Dave and Thomas for you to be busy of this bug, which I think was not so easy.
Comment 104 Philippe Didier 2011-10-18 00:06:28 CEST
Hats off !
Perseverance , method, knowledge , good willing, respect...

You gave a lesson on how to proceed for diagnosis and therapy!

I have followed the whole story ... waiting for each next episode ... each question each answer 


Thanks !
Best regards !

CC: (none) => philippedidier

Comment 105 Dave Hodgins 2011-10-18 01:45:15 CEST
Thomas, is ldetect-0.11.1-5.mga1.src.rpm ready for validation,
or do you have other changes to make first?
Comment 106 Thomas Backlund 2011-10-18 06:50:11 CEST
I'm waiting for Marianne Lombard,reporter of bug #933 to confirm that we haven't broken her setup.

CC: (none) => marianne

Comment 107 Gert Ereth 2011-10-18 09:31:41 CEST
after installing ldetect-lst-0.1.291-9.1.mga1 :

$ lspcidrake -v | grep STORAGE_IDE
sata_nv         : nVidia Corporation|MCP51 Serial ATA Controller [STORAGE_IDE] (vendor:10de device:0266 subv:17c0 subd:4076) (rev: f1)
pata_amd        : nVidia Corporation|MCP51 IDE [STORAGE_IDE] (vendor:10de device:0265 subv:17c0 subd:4076) (rev: f1)

ok !
Comment 108 David GEIGER 2011-10-18 19:55:17 CEST
Thomas ,Just a suggestion:

Why not use (or start at boot time) the two drivers amd74xx and pata_amd?

Modprobe.conf file would be:

options snd-ac97-codec power_save=1
install scsi_hostadapter /sbin/modprobe sata_nv; /bin/true
remove tifm_7xx1 /sbin/modprobe -r mmc_block; /sbin/modprobe -r tifm_sd; /sbin/modprobe --first-time -r --ignore-remove tifm_7xx1
install tifm_7xx1 /sbin/modprobe --first-time --ignore-install tifm_7xx1 && { /sbin/modprobe mmc_block; /sbin/modprobe tifm_sd; /bin/true; }
install usb-interface /sbin/modprobe ehci_hcd; /sbin/modprobe ohci_hcd; /bin/true
install ide-controller /sbin/modprobe pata_amd; /bin/true
install ide-controller /sbin/modprobe amd74xx; /bin/true

This could work or not ? 

And it could prevent future conflicts for the driver from the hardware configuration of each.As if bug #1525 and bug #933.

What do you think ? good suggestion or not.
Comment 109 Thomas Backlund 2011-10-18 20:03:42 CEST
(In reply to comment #108)
> Thomas ,Just a suggestion:
> 
> Why not use (or start at boot time) the two drivers amd74xx and pata_amd?
> 
[...]

> 
> This could work or not ? 
> 

It wont work.

As both drivers support the exact same hw, the first one that loads will drive the hw, and the other will fail to load because it finds no "available" hw.

But this is not a big issue as we try to move all users to ata based drivers, as ide drivers are considered legacy, and only to be used as a fallback.

The problem here is that the initial fix for #933 casted a too wide net, catching all nVidia MCP51 based ide controllers, instead of that specific system, but that should now be fixed.
Comment 110 Dave Hodgins 2011-10-19 00:47:57 CEST
Comment 73 still applies for cauldron.

I just installed kernel-server-3.1.0-0.rc9.1.mga2.
# lsinitrd /boot/initrd-3.1.0-server-0.rc10.1.mga2.img|grep pata
-rw-r--r--   1 root     root         5717 Oct 18 02:25 lib/modules/3.1.0-server-0.rc10.1.mga2/pata_via.ko.gz
-rw-r--r--   1 root     root         2747 Oct 18 02:25 lib/modules/3.1.0-server-0.rc10.1.mga2/pata_acpi.ko.gz
echo "Loading pata_acpi module"
modprobe -q pata_acpi
echo "Loading pata_via module"
modprobe -q pata_via

Again, it's loading pata_acpi before pata_via, so my dvd drive is hdd
instead of sr0.

Obviously it's a different problem then what's affecting the amd modules,
but with the same result.  Should I open a separate bug report for the
pata_via module loading after pata_acpi?
Comment 111 Dave Hodgins 2011-10-19 02:37:38 CEST
*** Bug 3064 has been marked as a duplicate of this bug. ***

CC: (none) => nextwavecs

Comment 112 Bill Wilkinson 2011-10-20 21:43:19 CEST
Another question:  

I've gone through Dave's recommended changes to initrd and modprobe.conf changing amd74xx to pata_amd.

Just for grins, I did a lsmod and got:
lsmod | grep amd
amd74xx                12977  0
pata_amd               14130  0
ide_core              121334  3 amd74xx,ide_pci_generic,ide_gd_mod
libata                218441  4 pata_amd,ata_generic,pata_acpi,sata_nv

So am I missing something in the init scripts?  Or is something else hiding in the background?
Comment 113 David GEIGER 2011-10-20 22:01:49 CEST
Indeed, it's weird I get:

# lsmod | grep amd
pata_amd               14130  0
libata                218405  4 pata_amd,ata_generic,pata_acpi,sata_nv

By cons, edit init file was just a test (test stanza in Comment 78 and Comment 79).
-In final, I have boot on file "initrd" original and not on the "test stanza".
-I edit the modprobe.conf file.
-I install the new update kernel.
-Reboot the system and it's Ok.
Comment 114 Thomas Backlund 2011-10-20 22:06:09 CEST
(In reply to comment #112)
> Another question:  
> 
> I've gone through Dave's recommended changes to initrd and modprobe.conf
> changing amd74xx to pata_amd.
> 
> Just for grins, I did a lsmod and got:
> lsmod | grep amd
> amd74xx                12977  0
> pata_amd               14130  0
> ide_core              121334  3 amd74xx,ide_pci_generic,ide_gd_mod
> libata                218441  4 pata_amd,ata_generic,pata_acpi,sata_nv
> 
> So am I missing something in the init scripts?  Or is something else hiding in
> the background?

If you changed the initrd before you changed modprobe.conf, then this is normal.

the "--preload pata_amd" just added pata_amd to the initrd, it it didn't remove amd74xx

But on next kernel install (or generation of initrd) the amd74xx will be gone.
Comment 115 Bill Wilkinson 2011-10-20 22:08:56 CEST
Thanks, Thomas!  

Should I gather from David's comment that there is a kernel update in Updates Testing?
Comment 116 David GEIGER 2011-10-20 22:32:15 CEST
(In reply to comment #115)
  
> 
> Should I gather from David's comment that there is a kernel update in Updates
> Testing?

Yes Bill, in Core_Updates_Testing is an update kernel 2.6.38.8-6.1.mga.
Comment 117 Dave Hodgins 2011-10-21 10:33:25 CEST
(In reply to comment #116)
> (In reply to comment #115)
> 
> > 
> > Should I gather from David's comment that there is a kernel update in Updates
> > Testing?
> 
> Yes Bill, in Core_Updates_Testing is an update kernel 2.6.38.8-6.1.mga.

Note that comment 73 was about a new kernel in cauldron.

While there is a new kernel in Core Updates Testing for Mageia one, as usual,
many of the i586 rpm packages are missing, so it is not ready for qa testing
yet, let alone end user testing, with the exception of those testing specific
packages for hardware support.
Comment 118 David GEIGER 2011-11-11 21:02:26 CET
Hello,

Given that Marianne Lombard  has not responded on bug #933,Thomas can you
push the srpm ldetect-lst-0.1.291-9.1.mga1.src.rpm from Core_Updates_Testing
to Core_Updates ?

And this way we can finally close this bug.

Thank you very much.
Comment 119 verg 2011-12-16 15:52:30 CET
The Bug is presented in Mageia2Alpha2
More info: https://forums.mageia.org/en/viewtopic.php?p=11552#p11552

CC: (none) => vergyz

Comment 120 Colin Guthrie 2011-12-16 16:51:16 CET
Is this basically the same problem as https://bugs.mageia.org/show_bug.cgi?id=3395#c8 ?

(different drivers but basically the same issue).

Should be solvable with a simple modprobe rule and using dracut to generate the initrd.

Still need to check if we can do it automatically tho'.
Comment 121 Marja Van Waes 2012-01-31 08:51:12 CET
(In reply to comment #118)
> Hello,
> 
> Given that Marianne Lombard  has not responded on bug #933,Thomas can you
> push the srpm ldetect-lst-0.1.291-9.1.mga1.src.rpm from Core_Updates_Testing
> to Core_Updates ?
> 
> And this way we can finally close this bug.
> 
> Thank you very much.

For some reason it wasn't closed. Now closing.

@ verg
If you still have your issue in Mageia 2 alpha, please look at the link Colin gave or search for another cauldron bug about that problem

Status: NEW => RESOLVED
CC: (none) => marja11
Resolution: (none) => FIXED

Comment 122 verg 2012-02-22 20:51:00 CET
(In reply to comment #121)
> (In reply to comment #118)
> > Hello,
> > 
> > Given that Marianne Lombard  has not responded on bug #933,Thomas can you
> > push the srpm ldetect-lst-0.1.291-9.1.mga1.src.rpm from Core_Updates_Testing
> > to Core_Updates ?
> > 
> > And this way we can finally close this bug.
> > 
> > Thank you very much.
> 
> For some reason it wasn't closed. Now closing.
> 
> @ verg
> If you still have your issue in Mageia 2 alpha, please look at the link Colin
> gave or search for another cauldron bug about that problem

The problem from comment â119 is fixed in Mageia2Beta1
Thanks.
Comment 123 AL13N 2012-05-07 16:39:10 CEST
i have the same problem in mga1 with:

[root@localhost ~]# lspcidrake | grep _IDE
pata_jmicron    : JMicron Technology Corp.|JMB362/JMB363 Serial ATA Controller [STORAGE_IDE] (rev: 03)
ata_piix        : Intel Corporation|82801I (ICH9 Family) 2 port SATA IDE Controller [STORAGE_IDE] (rev: 02)
ata_piix        : Intel Corporation|82801IB (ICH9) 2 port SATA IDE Controller [STORAGE_IDE] (rev: 02)


will test mkinitrd --preload ata_piix and see if it helps.

CC: (none) => alien

Comment 124 AL13N 2012-05-07 17:01:34 CEST
ata_piix is loaded according to lsinitrd, but lsmod says pata_jmicron is still first.

no dvddrive present...
Comment 125 AL13N 2012-05-07 18:17:49 CEST
forget it, my IDE cable was not connected
Comment 126 Thierry Vignaud 2012-05-07 18:48:14 CEST
... '-|-

Note You need to log in before you can comment on or make changes to this bug.