Description of problem: When connecting my iRiver device to the USB disk I get the following messages in /var/log/messages: May 31 23:22:28 grower kernel: usb 3-2: new full speed USB device using ohci_hcd and address 2 May 31 23:22:28 grower kernel: scsi3 : SCSI emulation for USB Mass Storage devices May 31 23:22:57 grower kernel: usb 3-2: reset full speed USB device using ohci_hcd and address 2 May 31 23:22:57 grower kernel: usb 3-2: device not accepting address 2, error -32 May 31 23:22:57 grower kernel: usb 3-2: reset full speed USB device using ohci_hcd and address 2 May 31 23:22:58 grower kernel: usb 3-2: device not accepting address 2, error -32 May 31 23:22:58 grower kernel: usb 3-2: reset full speed USB device using ohci_hcd and address 2 May 31 23:22:58 grower kernel: usb 3-2: device not accepting address 2, error -32 May 31 23:22:58 grower kernel: usb 3-2: reset full speed USB device using ohci_hcd and address 2 May 31 23:22:59 grower kernel: usb 3-2: device not accepting address 2, error -32 May 31 23:22:59 grower kernel: scsi: Device offlined - not ready after error recovery: host 3 channel 0 id 0 lun 0 May 31 23:22:59 grower kernel: usb 3-2: USB disconnect, address 2 May 31 23:22:59 grower kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000060 May 31 23:22:59 grower kernel: printing eip: May 31 23:22:59 grower kernel: c01cdcf8 May 31 23:22:59 grower kernel: *pde = 16d30067 May 31 23:22:59 grower kernel: Oops: 0000 [#1] May 31 23:22:59 grower kernel: Modules linked in: parport_pc lp parport autofs4 i2c_dev i2c_core sunrpc vfat fat dm_mod video button battery ac usb_storage usblp md5 ipv6 joydev ohci_hcd ehci_hcd snd_intel8x0 snd_ac97_codec snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd soundcore snd_page_alloc sis900 floppy ext3 jbd sata_sis libata sd_mod scsi_mod May 31 23:22:59 grower kernel: CPU: 0 May 31 23:22:59 grower kernel: EIP: 0060:[<c01cdcf8>] Not tainted VLI May 31 23:22:59 grower kernel: EFLAGS: 00010296 (2.6.11-1.1366_FC4) May 31 23:22:59 grower kernel: EIP is at sysfs_hash_and_remove+0xe/0xf1 May 31 23:22:59 grower kernel: eax: 00000000 ebx: d63eb664 ecx: 00000000 edx: c039215c May 31 23:22:59 grower kernel: esi: d63eb65c edi: e0883830 ebp: e08837a0 esp: c171edd8 May 31 23:22:59 grower kernel: ds: 007b es: 007b ss: 0068 May 31 23:22:59 grower kernel: Process khubd (pid: 86, threadinfo=c171e000 task=c170a000) May 31 23:22:59 grower kernel: Stack: c039215c 00000000 c03c9ab8 d63eb664 d63eb65c e0883830 e08837a0 c0288f4d May 31 23:22:59 grower kernel: e0883838 d63eb65c d63eb400 d63eb858 00000202 c0288f89 d63eb800 e086c954 May 31 23:22:59 grower kernel: d63eb7f8 d63eb800 db707880 e086ca6a 00000003 d4f42d4c c019f823 00000002 May 31 23:22:59 grower kernel: Call Trace: May 31 23:22:59 grower kernel: [<c0288f4d>] class_device_del+0x93/0xc7 May 31 23:22:59 grower kernel: [<c0288f89>] class_device_unregister+0x8/0x10 May 31 23:22:59 grower kernel: [<e086c954>] scsi_remove_device+0x49/0x8c [scsi_mod] May 31 23:22:59 grower kernel: [<e086ca6a>] __scsi_remove_target+0xd3/0x333 [scsi_mod] May 31 23:22:59 grower kernel: [<c019f823>] generic_delete_inode+0x142/0x346 May 31 23:22:59 grower kernel: [<e086ccd5>] scsi_remove_target+0xb/0x5a [scsi_mod] May 31 23:22:59 grower kernel: [<e086b9e3>] scsi_forget_host+0x91/0x2e9 [scsi_mod] May 31 23:22:59 grower kernel: [<e0862056>] scsi_remove_host+0x9/0x4b [scsi_mod] May 31 23:22:59 grower kernel: [<e09aaf86>] storage_disconnect+0x5f/0x79 [usb_storage] May 31 23:22:59 grower kernel: [<c02cbcd2>] usb_unbind_interface+0x34/0x5d May 31 23:22:59 grower kernel: [<c0287e89>] device_release_driver+0x45/0x50 May 31 23:22:59 grower kernel: [<c028808f>] bus_remove_device+0x65/0x9e May 31 23:22:59 grower kernel: [<c02872e3>] device_del+0x65/0x92 May 31 23:22:59 grower kernel: [<c02d557e>] usb_disable_device+0x97/0xe4 May 31 23:22:59 grower kernel: [<c02ce5d1>] usb_disconnect+0xa5/0x296 May 31 23:22:59 grower kernel: [<c02cf5f4>] hub_port_connect_change+0x51/0x626 May 31 23:22:59 grower kernel: [<c02cfec0>] hub_events+0x2f7/0x577 May 31 23:22:59 grower kernel: [<c02d018d>] hub_thread+0x4d/0x10a May 31 23:22:59 grower kernel: [<c0140052>] autoremove_wake_function+0x0/0x37 May 31 23:22:59 grower kernel: [<c02d0140>] hub_thread+0x0/0x10a May 31 23:22:59 grower kernel: [<c01012ad>] kernel_thread_helper+0x5/0xb May 31 23:22:59 grower kernel: Code: 0c c7 44 24 08 8d 00 00 00 c7 44 24 04 84 ae 38 c0 c7 04 24 d0 1a 38 c0 e8 d9 2b f5 ff 55 57 56 53 83 ec 0c 89 44 24 04 89 14 24 <8b> 70 60 8b 58 20 ba 66 00 00 00 b8 eb 51 38 c0 e8 ae f8 f4 ff Version-Release number of selected component (if applicable): uname -a yields: Linux grower.grobnet 2.6.11-1.1366_FC4 #1 Sun May 29 23:58:27 EDT 2005 i686 i686 i386 GNU/Linux How reproducible: Every time. Steps to Reproduce: 1. Connect iRiver to USB 2. 3. Actual results: See above. Expected results: I expected the device to appear in /media/usbdisk , which is what happened in Fedora Core 3. Additional info: Please let me know if you need more.
Dov, can you tell me what kernel was that worked last (I presume, on FC3)?
In this case I have to admit that in my stupidity I did not save the FC3 partition when I upgraded. Thus the only information that I can give is that I kept doing "yum update" while I was using FC3 and it never broke. Perhaps there are some log files that can help, but you would have to tell me which. If it helps I can try mounting the device on my older PIII 800MBox which I also upgraded to FC4, and then find out whether the problem is with the device, or the USB controller?
Yes, please do it with the PIII. While it looks like a regression when FC3 worked and FC4 fails on the same hardware, trying various combinations of hardware gives useful hints. There's too much difference between FC3 and FC4 kernels for me to guess outright from the .diff.
I'm ashamed to admit this, but when I now tried to mount the device again with a fully charged battery, it mounted flawlessly. I really should have checked this before. My apologies. But I am happy that it works. Closing the bug.
I'm glad the workaround exists, but the kernel must not oops. That's what the bug is. I'll pass this over to the upstream for now.
Tried this on the latest kernel: Linux grower.grobnet 2.6.12-1.1400_FC5 #1 Thu Jun 23 23:19:41 EDT 2005 i686 i686 i386 GNU/Linux and the device mounts ok, even with a half-empty battery. :-)