kernel: xhci_hcd WARN: Stalled endpoint

[Marko Polo]

I get this weird error from time to time without any possible reason:

Aug  6 07:58:49 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 07:58:49 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 07:58:54 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 07:58:54 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 07:59:02 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 07:59:02 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 07:59:04 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 07:59:04 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 07:59:07 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 07:59:07 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 07:59:15 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 07:59:15 kernel: xhci_hcd 0000:00:0c.0: WARN: Stalled endpoint
Aug  6 08:16:51 kernel: sd 0:0:0:0: timing out command, waited 180s
Aug  6 08:16:51 kernel: sd 0:0:0:0: [sda] Unhandled sense code
Aug  6 08:16:51 kernel: sd 0:0:0:0: [sda]  Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Aug  6 08:16:51 kernel: sd 0:0:0:0: [sda]  Sense Key : Hardware Error [current] 
Aug  6 08:16:51 kernel: sd 0:0:0:0: [sda]  ASC=0x44 <<vendor>> ASCQ=0x81
Aug  6 08:16:51 kernel: sd 0:0:0:0: [sda] CDB: Read(10): 28 00 00 00 01 e2 00 00 08 00
Aug  6 08:16:51 kernel: end_request: I/O error, dev sda, sector 482
Aug  6 08:16:51 kernel: tntfs error (device sda1, pid 9): ntfs_end_buffer_async_read(): Buffer I/O error, logical block 0xe0.
Aug  6 08:16:51 kernel: tntfs error (device sda1, pid 9): ntfs_end_buffer_async_read(): Buffer I/O error, logical block 0xe1.
Aug  6 08:16:51 kernel: tntfs error (device sda1, pid 9): ntfs_end_buffer_async_read(): Buffer I/O error, logical block 0xe2.
Aug  6 08:16:51 kernel: tntfs error (device sda1, pid 9): ntfs_end_buffer_async_read(): Buffer I/O error, logical block 0xe3.
Aug  6 08:16:51 kernel: tntfs error (device sda1, pid 14427): ntfs_mft_record_map(): Failed to map page containing mft record 0xd3 (error 5).
Aug  6 08:16:51 kernel: tntfs error (device sda1, pid 14427): ntfs_mft_record_map(): Failed with error code 5.
Aug  6 08:16:51 kernel: tntfs error (device sda1, pid 14427): ntfs_index_lookup_init(): Failed to map mft record (error 5).
Aug  6 08:16:51 kernel: tntfs error (device sda1, pid 14427): ntfs_index_lookup_by_position(): Failed (error 5).
Aug  6 08:16:51 kernel: tntfs error (device sda1, pid 14427): ntfs_readdir(): Failed to look up index entry with position 0x0 in directory 0xd3 (error 5).

I cannot access Samba shares and cannot complete SSH-connection successfully. Only router reboot helps.
Any suggestions? Is attached drive is dying?

 

[Marko Polo]

Is there any disk check tool I can utilyze just from router (from Entware)? I checked ntfs partitions with chkdsk but I have no physical linux machine to check EXT4 partitions.

 

[RMerlin]

Try a surface scan check (many HDD manufacturers provide such tools on their website), it will scan the whole surface rather than just the NTFS partitions.

Also try replacing the USB cable, and make sure the HDD has its own power supply rather than relying strictly on USB power.

The logged errors aren't at the filesystem level (and in fact those logged were on the NTFS partition), so you don't need to run any ext filesystem check.

 

[Marko Polo]

Try a surface scan check (many HDD manufacturers provide such tools on their website), it will scan the whole surface rather than just the NTFS partitions.

I checked the HDD (including surface scan) on several utilities (including those from the HDD vendor) and found no errors. It is absolutely healthy. So it seems to be strictly (mis)configuration issue. Are you really thinking this is insufficient power supply?
It's odd I never faced such issues in previous months of using Merlin.

 

[sfx2000]

Also try replacing the USB cable, and make sure the HDD has its own power supply rather than relying strictly on USB power.

The logged errors aren't at the filesystem level (and in fact those logged were on the NTFS partition), so you don't need to run any ext filesystem check.

Concur - good power and a high quality cable are a must.

Do not use bus-powered HDD's and ensure a direct connection - no USB hubs in between.

The fs errors are due to the stalled xhci endpoint.

 

[Marko Polo]

Still struggling with this problem. Now I get such errors:

Sep  8 13:18:54 kernel: gro enabled with interval 2
Sep  8 13:18:58 kernel: usb 2-2: device not accepting address 10, error -110
Sep  8 13:18:59 kernel: usb 2-2: new high speed USB device using ehci_hcd and address 11
Sep  8 13:19:02 custom_script: Running /jffs/scripts/smb.postconf (args: /etc/smb.conf)
Sep  8 13:19:02 hotplug[2455]: USB ext4 fs at /dev/sda4 mounted on /tmp/mnt/IMAGES
Sep  8 13:19:02 usb: USB ext4 fs at /dev/sda4 mounted on /tmp/mnt/IMAGES.
Sep  8 13:19:02 kernel: EXT4-fs (sda4): recovery complete
Sep  8 13:19:02 kernel: EXT4-fs (sda4): mounted filesystem with ordered data mode. Opts: user_xattr
Sep  8 13:19:03 custom_script: Running /jffs/scripts/post-mount (args: /tmp/mnt/IMAGES)
Sep  8 13:19:07 kernel: gro enabled with interval 2
Sep  8 13:19:08 custom_script: Running /jffs/scripts/smb.postconf (args: /etc/smb.conf)
Sep  8 13:19:09 Samba_Server: daemon is started
Sep  8 13:19:09 kernel: usb 2-2: device not accepting address 11, error -110
Sep  8 13:19:09 kernel: hub 2-0:1.0: unable to enumerate USB device on port 2

Can you say what they mean?

 

[john9527]

https://stackoverflow.com/questions/13653692/device-descriptor-read-64-error-110

If you have a device plugged in the other USB port, try unplugging it. Also power down the router, and replug it's power cable. Last resort, try a different router supply supply.

 

[jerry6]

Don't let the SSD go into sleep mode , had problems with mine and they went away after I set always on , something to do with trim in SSD , at least with mine . Had one go bad because of this

 

[Marko Polo]

Don't let the SSD go into sleep mode

How to do that in Merlin?

 

[ColinTaylor]

How to do that in Merlin?

He's probably referring to the spin-down settings at Tools > Other Settings.

 

[Marko Polo]

He's probably referring to the spin-down settings at Tools > Other Settings.

Haven't found such setting in Other settings.

Spoiler: Miscellaneous

Spoiler: Advanced Tweaks settings

 

[ColinTaylor]

Asuswrt-Merlin 384 Changelog
============================

General changes in the 384/NG branch vs 380.xx:
- HDD spindown settings moved to the System settings page

 

[Marko Polo]

I suppose now it is called Enable HDD Hibernation? I do not see anything similar to spindown or mentioning this word.
If yes, then it was disabled, so it is no-change.