Skip to contentSkip to conversation

Udev and Mylex Dac960

DISCUSSION claude 2 participants 13 messages Feb 5, 2004 – Feb 16, 2004

lore
Share & tools
lore

claude

author
#0 lore raw

Hi all,

Udev reports my DAC960 devices as /dev/rd!c0d0 ? Is this ok ?

After reading http://www.lanana.org/docs/device-list/devices.txt, I think it should be /dev/rd/c0d0. Or is there any other naming scheme ?

Thanks,
claude

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

#1 lore raw
Quoted 3 lines · claude

Hi all,

Udev reports my DAC960 devices as /dev/rd!c0d0 ? Is this ok ?

Depends, do you like that name? :)

Can you give me the output of 'tree -d /sys/block' on your system? I think udev just needs to change that '!' character to a '/' to make it correct.

thanks,

greg k-h

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

author
#2 lore raw
In reply to claude
Quoted 4 lines · Greg KH

On Fri, Feb 06, 2004 at 12:02:21AM +0100, claude wrote:

Depends, do you like that name? :)

I can't say :). I just want to have right naming. Devices list at lanana.org is pretty old, and things change too quickly. I wanted to know which name is the right today.

Quoted 4 lines

Can you give me the output of 'tree -d /sys/block' on your system? I think udev just needs to change that '!' character to a '/' to make it correct.

Yes, the output is below, and I think you're right too ;)

Quoted 2 lines

thanks,

Thanks to you all for this great work :)

Quoted 2 lines

greg k-h

Claude

[email protected]# tree -d /sys/block
/sys/block
|-- fd0
|   `-- queue
|       `-- iosched
|-- hda
|   |-- device -> ../../devices/pci0000:00/0000:00:07.1/ide0/0.0
|   `-- queue
|       `-- iosched
|-- loop0
|-- loop1
|-- loop2
|-- loop3
|-- loop4
|-- loop5
|-- loop6
|-- loop7
|-- md0
|-- ram0
|-- ram1
|-- ram10
|-- ram11
|-- ram12
|-- ram13
|-- ram14
|-- ram15
|-- ram2
|-- ram3
|-- ram4
|-- ram5
|-- ram6
|-- ram7
|-- ram8
|-- ram9
|-- rd!c0d0
|   |-- queue
|   |   `-- iosched
|   |-- rd!c0d01
|   |-- rd!c0d02
|   |-- rd!c0d03
|   |-- rd!c0d04
|   |-- rd!c0d05
|   `-- rd!c0d06
|-- rd!c0d1
|   `-- queue
|       `-- iosched
|-- rd!c0d10
|   `-- queue
|       `-- iosched
|-- rd!c0d11
|   `-- queue
|       `-- iosched
|-- rd!c0d12
|   `-- queue
|       `-- iosched
|-- rd!c0d13
|   `-- queue
|       `-- iosched
|-- rd!c0d14
|   `-- queue
|       `-- iosched
|-- rd!c0d15
|   `-- queue
|       `-- iosched
|-- rd!c0d16
|   `-- queue
|       `-- iosched
|-- rd!c0d17
|   `-- queue
|       `-- iosched
|-- rd!c0d18
|   `-- queue
|       `-- iosched
|-- rd!c0d19
|   `-- queue
|       `-- iosched
|-- rd!c0d2
|   `-- queue
|       `-- iosched
|-- rd!c0d20
|   `-- queue
|       `-- iosched
|-- rd!c0d21
|   `-- queue
|       `-- iosched
|-- rd!c0d22
|   `-- queue
|       `-- iosched
|-- rd!c0d23
|   `-- queue
|       `-- iosched
|-- rd!c0d24
|   `-- queue
|       `-- iosched
|-- rd!c0d25
|   `-- queue
|       `-- iosched
|-- rd!c0d26
|   `-- queue
|       `-- iosched
|-- rd!c0d27
|   `-- queue
|       `-- iosched
|-- rd!c0d28
|   `-- queue
|       `-- iosched
|-- rd!c0d29
|   `-- queue
|       `-- iosched
|-- rd!c0d3
|   `-- queue
|       `-- iosched
|-- rd!c0d30
|   `-- queue
|       `-- iosched
|-- rd!c0d31
|   `-- queue
|       `-- iosched
|-- rd!c0d4
|   `-- queue
|       `-- iosched
|-- rd!c0d5
|   `-- queue
|       `-- iosched
|-- rd!c0d6
|   `-- queue
|       `-- iosched
|-- rd!c0d7
|   `-- queue
|       `-- iosched
|-- rd!c0d8
|   `-- queue
|       `-- iosched
`-- rd!c0d9
     `-- queue
         `-- iosched

134 directories

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

author
#3 lore raw
In reply to claude

Hi,

I have another question about Mylex Dac960.

Is there any way for udev to create device only for existing array ?

(eg : I only have /dev/rd!c0d0 but udev create /dev/rd!c0d[1-31]).

Thanks for your response !

claude

Quoted 141 lines
[email protected]# tree -d /sys/block
/sys/block
|-- fd0
|   `-- queue
|       `-- iosched
|-- hda
|   |-- device -> ../../devices/pci0000:00/0000:00:07.1/ide0/0.0
|   `-- queue
|       `-- iosched
|-- loop0
|-- loop1
|-- loop2
|-- loop3
|-- loop4
|-- loop5
|-- loop6
|-- loop7
|-- md0
|-- ram0
|-- ram1
|-- ram10
|-- ram11
|-- ram12
|-- ram13
|-- ram14
|-- ram15
|-- ram2
|-- ram3
|-- ram4
|-- ram5
|-- ram6
|-- ram7
|-- ram8
|-- ram9
|-- rd!c0d0
|   |-- queue
|   |   `-- iosched
|   |-- rd!c0d01
|   |-- rd!c0d02
|   |-- rd!c0d03
|   |-- rd!c0d04
|   |-- rd!c0d05
|   `-- rd!c0d06
|-- rd!c0d1
|   `-- queue
|       `-- iosched
|-- rd!c0d10
|   `-- queue
|       `-- iosched
|-- rd!c0d11
|   `-- queue
|       `-- iosched
|-- rd!c0d12
|   `-- queue
|       `-- iosched
|-- rd!c0d13
|   `-- queue
|       `-- iosched
|-- rd!c0d14
|   `-- queue
|       `-- iosched
|-- rd!c0d15
|   `-- queue
|       `-- iosched
|-- rd!c0d16
|   `-- queue
|       `-- iosched
|-- rd!c0d17
|   `-- queue
|       `-- iosched
|-- rd!c0d18
|   `-- queue
|       `-- iosched
|-- rd!c0d19
|   `-- queue
|       `-- iosched
|-- rd!c0d2
|   `-- queue
|       `-- iosched
|-- rd!c0d20
|   `-- queue
|       `-- iosched
|-- rd!c0d21
|   `-- queue
|       `-- iosched
|-- rd!c0d22
|   `-- queue
|       `-- iosched
|-- rd!c0d23
|   `-- queue
|       `-- iosched
|-- rd!c0d24
|   `-- queue
|       `-- iosched
|-- rd!c0d25
|   `-- queue
|       `-- iosched
|-- rd!c0d26
|   `-- queue
|       `-- iosched
|-- rd!c0d27
|   `-- queue
|       `-- iosched
|-- rd!c0d28
|   `-- queue
|       `-- iosched
|-- rd!c0d29
|   `-- queue
|       `-- iosched
|-- rd!c0d3
|   `-- queue
|       `-- iosched
|-- rd!c0d30
|   `-- queue
|       `-- iosched
|-- rd!c0d31
|   `-- queue
|       `-- iosched
|-- rd!c0d4
|   `-- queue
|       `-- iosched
|-- rd!c0d5
|   `-- queue
|       `-- iosched
|-- rd!c0d6
|   `-- queue
|       `-- iosched
|-- rd!c0d7
|   `-- queue
|       `-- iosched
|-- rd!c0d8
|   `-- queue
|       `-- iosched
`-- rd!c0d9
    `-- queue
        `-- iosched

134 directories

Quoted 10 lines

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

#4 lore raw
In reply to claude
Quoted 9 lines · claude
Quoted 4 lines · Greg KH

On Fri, Feb 06, 2004 at 12:02:21AM +0100, claude wrote:

Depends, do you like that name? :)

I can't say :). I just want to have right naming. Devices list at lanana.org is pretty old, and things change too quickly. I wanted to know which name is the right today.

Quoted 5 lines
Quoted 4 lines

Can you give me the output of 'tree -d /sys/block' on your system? I think udev just needs to change that '!' character to a '/' to make it correct.

Yes, the output is below, and I think you're right too ;)

Thanks for the output. I've fixed this in the current udev tree, so the next release should work just fine for this kind of device.

Please let me know if it doesn't.

Thanks again for reporting this.

greg k-h

-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id56&alloc_id438&op=click
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

#5 lore raw
In reply to claude
Quoted 7 lines · claude

Hi,

I have another question about Mylex Dac960.

Is there any way for udev to create device only for existing array ?

(eg : I only have /dev/rd!c0d0 but udev create /dev/rd!c0d[1-31]).

Well your kernel is saying that you really do have all of those other devices. You might want to talk to the kernel driver about this...

But yes, you can ignore devices with the following rule:

	KERNEL="*c0d[1-31]", NAME=""

The NAME="" section is optional, but a bit nicer to try to figure out what is happening.

The ability to ignore devices is broken in udev 016, but is now fixed in the current tree. Try this with the next release (hopefully in a few days...)

thanks,

greg k-h

-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id56&alloc_id438&op=click
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

author
#6 lore raw
In reply to claude
Quoted 6 lines · Greg KH
Quoted 4 lines · claude

Hi,

I have another question about Mylex Dac960.

Quoted 8 lines
Quoted 3 lines

Is there any way for udev to create device only for existing array ?

(eg : I only have /dev/rd!c0d0 but udev create /dev/rd!c0d[1-31]).

Well your kernel is saying that you really do have all of those other devices. You might want to talk to the kernel driver about this...

Yes i will do, but is there any docs about sysfs and how it works ? I've only found sysfs.txt in linux-2.6.2/Documentation, and it was too light explanation for my poor knowledge. Any pointer ?

Quoted 9 lines

But yes, you can ignore devices with the following rule:

	KERNEL="*c0d[1-31]", NAME=""

The NAME="" section is optional, but a bit nicer to try to figure out what is happening.

The ability to ignore devices is broken in udev 016, but is now fixed in the current tree. Try this with the next release (hopefully in a few days...)

ok, I will wait for it.

Quoted 3 lines

thanks,

thanks too and thanks to all others for this great work :)

Quoted 2 lines

greg k-h

claude

-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id56&alloc_id438&op=click
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

#7 lore raw
In reply to claude
Quoted 3 lines · claude

Yes i will do, but is there any docs about sysfs and how it works ? I've only found sysfs.txt in linux-2.6.2/Documentation, and it was too light explanation for my poor knowledge. Any pointer ?

What do you want to know about sysfs?

How to write code to use it? If so, lwn.net has has a few good articles about it, and there is a OLS and linux.conf.au paper about it.

If you just want to know what in the world these different directories are, and how they relate to each other, from a user point of view, that's not so well documentated, yet :(

thanks,

greg k-h

-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id56&alloc_id438&op=click
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

author
#8 lore raw
In reply to claude
Quoted 7 lines · Greg KH
Quoted 3 lines · claude

Yes i will do, but is there any docs about sysfs and how it works ? I've only found sysfs.txt in linux-2.6.2/Documentation, and it was too light explanation for my poor knowledge. Any pointer ?

Quoted 5 lines

What do you want to know about sysfs?

How to write code to use it? If so, lwn.net has has a few good articles about it, and there is a OLS and linux.conf.au paper about it.

Many thanks for links, that's just what I want.

Also a bad news :(. I've downloaded udev-017.tar.gz, compiled and installed it, but my Mylex logical device are always created as /dev/rd!c0d0. Is there anything I can do ?

Quoted 9 lines

If you just want to know what in the world these different directories are, and how they relate to each other, from a user point of view, that's not so well documentated, yet :(

thanks,

greg k-h

Quoted 10 lines

-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id56&alloc_id438&op=click
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id56&alloc_id438&op=click
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

#9 lore raw
In reply to claude
Quoted 14 lines · claude
Quoted 7 lines · Greg KH
Quoted 3 lines · claude

Yes i will do, but is there any docs about sysfs and how it works ? I've only found sysfs.txt in linux-2.6.2/Documentation, and it was too light explanation for my poor knowledge. Any pointer ?

Quoted 5 lines

What do you want to know about sysfs?

How to write code to use it? If so, lwn.net has has a few good articles about it, and there is a OLS and linux.conf.au paper about it.

Quoted 7 lines

Many thanks for links, that's just what I want.

Also a bad news :(. I've downloaded udev-017.tar.gz, compiled and installed it,
but my Mylex logical device are always created as /dev/rd!c0d0. Is there anything
I can do ?

Ick. Can you build udev by doing the following:
make clean
make DEBUG=true USE_LOG=true

And then run the udevtest binary in that directory that gets built for your block device:
./udevtest block/rd!c0d0
(you might have to do this as:

	./udevtest block/rd\!c0d0

depending on your shell)

and then send the output that this program creates to us?

thanks,

greg k-h

-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id56&alloc_id438&op=click
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

author
#10 lore raw
In reply to claude
Quoted 13 lines · Greg KH
Quoted 5 lines · claude

Also a bad news :(. I've downloaded udev-017.tar.gz, compiled and installed it,
but my Mylex logical device are always created as /dev/rd!c0d0. Is there anything
I can do ?

Ick. Can you build udev by doing the following:
make clean
make DEBUG=true USE_LOG=true

Quoted 6 lines

And then run the udevtest binary in that directory that gets built for your block device:
./udevtest block/rd!c0d0
(you might have to do this as:

	./udevtest block/rd\!c0d0

depending on your shell)

Yeah, that's why '!' bother me ! ;)

Quoted 2 lines

and then send the output that this program creates to us?

Yes output follow (i have to type :

	udevtest /block/rd\!c0d0 )

main: version 017
udev_hotplug: looking at '/block/rd!c0d0'
get_dirs: sysfs_path='/sys'
parse_config_file: reading '/etc/udev/udev.conf' as config file
namedev_init_rules: reading '/etc/udev/udev.rules' as rules file
namedev_init_permissions: reading '/etc/udev/udev.permissions' as permissions file
sleep_for_dev: looking for '/sys/block/rd!c0d0/dev'
get_class_dev: looking at '/sys/block/rd!c0d0'
get_class_dev: class_dev->name='rd!c0d0'
get_major_minor: dev='48:0
'
get_major_minor: found majorH, minor=0
get_blockdev_parent: rd!c0d0 not a partition
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
get_sysfs_device: timed out waiting for device symlink, continuing on anyway...
namedev_name_device: class_dev->name = 'rd!c0d0'
namedev_name_device: udev->kernel_name = 'rd/c0d0'
namedev_name_device: kernel_number='0'
namedev_name_device: name, 'rd!c0d0' is going to have owner='root', group='root', mode = 0600
udev_add_device: name='rd!c0d0'
creating device node '/dev/rd!c0d0'
create_node: mknod(/dev/rd!c0d0, 020600, 48, 0)
create_node: chmod(/dev/rd!c0d0, 020600)

Quoted 4 lines

thanks,

greg k-h

thanks,
claude -

-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id56&alloc_id438&op=click
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

#11 lore raw
In reply to claude
Quoted 2 lines · claude

Yes output follow (i have to type :

	udevtest /block/rd\!c0d0 )

Oops, sorry, forgot the leading '/' I need to document udevtest, I made it for my testing but it looks like it really is handy to find problems like this.

Quoted 28 lines

main: version 017
udev_hotplug: looking at '/block/rd!c0d0'
get_dirs: sysfs_path='/sys'
parse_config_file: reading '/etc/udev/udev.conf' as config file
namedev_init_rules: reading '/etc/udev/udev.rules' as rules file
namedev_init_permissions: reading '/etc/udev/udev.permissions' as permissions file
sleep_for_dev: looking for '/sys/block/rd!c0d0/dev'
get_class_dev: looking at '/sys/block/rd!c0d0'
get_class_dev: class_dev->name='rd!c0d0'
get_major_minor: dev='48:0
'
get_major_minor: found majorH, minor=0
get_blockdev_parent: rd!c0d0 not a partition
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
get_sysfs_device: timed out waiting for device symlink, continuing on anyway...
namedev_name_device: class_dev->name = 'rd!c0d0'

Quoted 8 lines

namedev_name_device: udev->kernel_name = 'rd/c0d0'
namedev_name_device: kernel_number='0'
namedev_name_device: name, 'rd!c0d0' is going to have owner='root', group='root', mode = 0600
udev_add_device: name='rd!c0d0'
creating device node '/dev/rd!c0d0'
create_node: mknod(/dev/rd!c0d0, 020600, 48, 0)
create_node: chmod(/dev/rd!c0d0, 020600)

Ah, found it. I was testing this with the following rule in my rules file:

	BUS="scsi", NAME="%k"

If you put that at the end of your udev.rules file, you should get the proper name generated.

Or you can apply the following patch, which will solve the problem if you do not have a rule in udev.rules for this device.

thanks a lot for helping me find the problem.

greg k-h

# Fix bug where we did not use the "converted" kernel name if we had no rule. #
# This fixes the bug with names that have a ! in them and no rule to match.

namedev.c +1 -1
diff -Nru a/namedev.c b/namedev.c
@@ -745,7 +745,7 @@
 	}
 
 	/* no rule was found so we use the kernel name */
-	strfieldcpy(udev->name, class_dev->name);
+	strfieldcpy(udev->name, udev->kernel_name);
 	goto done;
 
 found:

-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id56&alloc_id438&op=click
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

author
#12 lore raw
In reply to claude

I've applied patch and it works great ! many thanks again !

Quoted 70 lines · Greg KH
Quoted 2 lines · claude

Yes output follow (i have to type :

	udevtest /block/rd\!c0d0 )

Oops, sorry, forgot the leading '/' I need to document udevtest, I made it for my testing but it looks like it really is handy to find problems like this.

Quoted 27 lines

main: version 017
udev_hotplug: looking at '/block/rd!c0d0'
get_dirs: sysfs_path='/sys'
parse_config_file: reading '/etc/udev/udev.conf' as config file
namedev_init_rules: reading '/etc/udev/udev.rules' as rules file
namedev_init_permissions: reading '/etc/udev/udev.permissions' as permissions file
sleep_for_dev: looking for '/sys/block/rd!c0d0/dev'
get_class_dev: looking at '/sys/block/rd!c0d0'
get_class_dev: class_dev->name='rd!c0d0'
get_major_minor: dev='48:0
'
get_major_minor: found majorH, minor=0
get_blockdev_parent: rd!c0d0 not a partition
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
sysfs_path_is_link: stat() failed
get_sysfs_device: timed out waiting for device symlink, continuing on anyway...
namedev_name_device: class_dev->name = 'rd!c0d0'

Quoted 8 lines

namedev_name_device: udev->kernel_name = 'rd/c0d0'
namedev_name_device: kernel_number='0'
namedev_name_device: name, 'rd!c0d0' is going to have owner='root', group='root', mode = 0600
udev_add_device: name='rd!c0d0'
creating device node '/dev/rd!c0d0'
create_node: mknod(/dev/rd!c0d0, 020600, 48, 0)
create_node: chmod(/dev/rd!c0d0, 020600)

Ah, found it. I was testing this with the following rule in my rules file:

	BUS="scsi", NAME="%k"

If you put that at the end of your udev.rules file, you should get the proper name generated.

Or you can apply the following patch, which will solve the problem if you do not have a rule in udev.rules for this device.

thanks a lot for helping me find the problem.

greg k-h

# Fix bug where we did not use the "converted" kernel name if we had no rule. #
# This fixes the bug with names that have a ! in them and no rule to match.

diff -Nru a/namedev.c b/namedev.c

namedev.c line 751

Quoted 8 lines
namedev.c +1 -1
@@ -745,7 +745,7 @@
 	}
 
 	/* no rule was found so we use the kernel name */
-	strfieldcpy(udev->name, class_dev->name);
+	strfieldcpy(udev->name, udev->kernel_name);
 	goto done;
 
 found:

-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id56&alloc_id438&op=click
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel