kobject_add failed for uhci_hcd with -EEXIST

dark* · 06-09-2008, 11:37 AM

I have been watching this message from kernel (slack 12 hugesmp) at booting, ever since I installed it (I even had it on my previous system but didnt bother to ask).

kobject_add failed for uhci_hcd with -EEXIST, don't try to register things with the same name in the same directory.
[<c03e86f7>] kobject_shadow_add+0x117/0x1a0
[<c013fba4>] mod_sysfs_setup+0x24/0xb0
[<c0141458>] sys_init_module+0x1648/0x1940
[<c0102ae8>] syscall_call+0x7/0xb
[<c0710000>] __mutex_lock_slowpath+0x40/0x90
=======================
kobject_add failed for uhci_hcd with -EEXIST, don't try to register things with the same name in the same directory.
[<c03e86f7>] kobject_shadow_add+0x117/0x1a0
[<c013fba4>] mod_sysfs_setup+0x24/0xb0
[<c0141458>] sys_init_module+0x1648/0x1940
[<c0102ae8>] syscall_call+0x7/0xb
[<c0710000>] __mutex_lock_slowpath+0x40/0x90
=======================
sd 2:0:0:0: Attached scsi generic sg0 type 0
shpchp: Standard Hot Plug PCI Controller Driver version: 0.4
kobject_add failed for uhci_hcd with -EEXIST, don't try to register things with the same name in the same directory.
[<c03e86f7>] kobject_shadow_add+0x117/0x1a0
[<c013fba4>] mod_sysfs_setup+0x24/0xb0
[<c0141458>] sys_init_module+0x1648/0x1940
[<c0102ae8>] syscall_call+0x7/0xb
[<c0710000>] __mutex_lock_slowpath+0x40/0x90
=======================

Whats the meaning of this, i didnt change anything (like a new kernel or driver). how to get rid of it? its not hindering any functionality but its looks like something is wrong.

Uncle_Theodore · 06-09-2008, 11:45 AM

Quote:

Originally Posted by dark*

I have been watching this message from kernel (slack 12 hugesmp) at booting, ever since I installed it (I even had it on my previous system but didnt bother to ask).

kobject_add failed for uhci_hcd with -EEXIST, don't try to register things with the same name in the same directory.
[<c03e86f7>] kobject_shadow_add+0x117/0x1a0
[<c013fba4>] mod_sysfs_setup+0x24/0xb0
[<c0141458>] sys_init_module+0x1648/0x1940
[<c0102ae8>] syscall_call+0x7/0xb
[<c0710000>] __mutex_lock_slowpath+0x40/0x90
=======================
kobject_add failed for uhci_hcd with -EEXIST, don't try to register things with the same name in the same directory.
[<c03e86f7>] kobject_shadow_add+0x117/0x1a0
[<c013fba4>] mod_sysfs_setup+0x24/0xb0
[<c0141458>] sys_init_module+0x1648/0x1940
[<c0102ae8>] syscall_call+0x7/0xb
[<c0710000>] __mutex_lock_slowpath+0x40/0x90
=======================
sd 2:0:0:0: Attached scsi generic sg0 type 0
shpchp: Standard Hot Plug PCI Controller Driver version: 0.4
kobject_add failed for uhci_hcd with -EEXIST, don't try to register things with the same name in the same directory.
[<c03e86f7>] kobject_shadow_add+0x117/0x1a0
[<c013fba4>] mod_sysfs_setup+0x24/0xb0
[<c0141458>] sys_init_module+0x1648/0x1940
[<c0102ae8>] syscall_call+0x7/0xb
[<c0710000>] __mutex_lock_slowpath+0x40/0x90
=======================

Whats the meaning of this, i didnt change anything (like a new kernel or driver). how to get rid of it? its not hindering any functionality but its looks like something is wrong.

It's described in CHANGES_AND_HINTS.TXT ftp://carroll.cac.psu.edu/pub/linux/..._AND_HINTS.TXT

It's not really a problem, just a consequence of using the huge kernel. Switch to generic (you'll need an initrd), it'll go away.

T3slider · 06-09-2008, 09:57 PM

Please read all, or at least most, of the relevant text files included with your Slackware installation, particularly CHANGES_AND_HINTS.TXT and UPGRADE.TXT (even if you are not upgrading). CHANGES_AND_HINTS.TXT is the most quoted piece of official documentation that goes unread, even though it contains the answers to 90% of the problems asked.

fibster · 06-09-2008, 10:48 PM

Quote:

Originally Posted by T3slider

Please read all, or at least most, of the relevant text files included with your Slackware installation, particularly CHANGES_AND_HINTS.TXTCHANGES_AND_HINTS.TXT and UPGRADE.TXT (even if you are not upgrading). CHANGES_AND_HINTS.TXT is the most quoted piece of official documentation that goes unread, even though it contains the answers to 90% of the problems asked.

Yes the CHANGES_AND_HINTS is always a good wager.

dark* · 06-10-2008, 02:07 AM

Thanks, i'll definitely look it.

brianL · 06-10-2008, 06:03 AM

If you do change to the generic kernel, you can use Eric's script to give you the right command to enter to create an initrd:

http://www.slackware.com/~alien/tool...d_generator.sh

semiRocket · 06-22-2009, 09:34 PM

I have similar errors in my system start, so I decided to do change to generic kernel.

I did everything from README.initrd except the compiling line which I changed a bit to suite my situation:

Code:

mkinitrd -c -k 2.6.27.7-smp -m mbcache:jbd:ext3 -f ext3 -r /dev/sda6

After I compiled initrd.gz I changed lilo.conf:

Code:

# Linux bootable partition config begins
image = /boot/vmlinuz-generic-smp-2.6.27.7-smp  # Default vmlinuz
  initrd = /boot/initrd.gz      # Default without the whole line
  root = /dev/sda6
  label = Slack
  read-only  # Partitions should be mounted read-only for checking
# Linux bootable partition config ends

I have run the lilo command which ran whithout errors. After I have restarted it was given me error that it can't find init and that it may indicate of mounting problems. And that I have been given a chance to fix it mannualy with some sort of a command line. (If you need the exact sequence, just tell me I can make a photo)

From lsmod -vnn I've found out that my SATA controller use kernel driver via_sata, but all I've found was /usr/src/linux-2.6.27.7/drivers/ata/sata_via.c, so I think that the controller is build into kernel.

What I've did wrong?

C-Sniper · 06-22-2009, 11:03 PM

Beside the ancient thread revival did you read the CHANGES_AND_HINTS.TXT?

semiRocket · 06-22-2009, 11:18 PM

Sorry about that but I have read CHANGES_AND_HINTS.TXT on the original dvd and it's saying see /boot/README.initrd for instructions, Which I did.

It's also saying something about udev but I can' understand if it's related to my situation.

I'm no expert

EDIT:
Here's the photo: http://i43.tinypic.com/2uikz9y.jpg
It is saying also that cannot find kernel modules but I all packages all installed, I did Full install

What does this mean?

brianL · 06-23-2009, 04:07 AM

Is there an initrd.gz in /boot? Did you cd to /boot before running the mkinitrd command?

bgeddy · 06-23-2009, 06:24 AM

Quote:

It is saying also that cannot find kernel modules but I all packages all installed, I did Full install

What does this mean?

This is being reported by the loading of the initial ram disk. It can't find any modules (which it stores in the ramdisk itself). The ramdisk doesn't look to your main system at this point so the fact you did a full install has nothing to do with it.

You can have a look at the /boot/initrd-tree directory to see how this works. This is a temporary storage that is built into the ramdisk.

As it has been said it looks like something has gone wrong with creating the initial ramdisk (initrd.gz). Note that whenever you recreate the initrd you will have to rerun lilo to install it.

semiRocket · 06-23-2009, 08:56 AM

Thank you guys I'm not sure what exactly went wrong, but I did what you told me and it worked

@brianL yes I have checked many times before the mkinitrd command created /boot/initrd-tree & initrd.gz. I don't remember if I went to /boot before compiling, but I thought it doesn't matter because it always creates those files in /boot.

@bgeddy I didn't know that, thank you
I think the missing modules in initrd was the cause