Linux USB
 help / color / mirror / Atom feed
From: Greg KH <gregkh@linuxfoundation.org>
To: Patrik Nilsson <nipatriknilsson@gmail.com>
Cc: linux-usb@vger.kernel.org
Subject: Re: high speed data to usb disk makes the kernel think that is has been unmounted
Date: Tue, 11 Aug 2020 19:56:55 +0200	[thread overview]
Message-ID: <20200811175655.GB828356@kroah.com> (raw)
In-Reply-To: <a161e45d-6ce8-2077-c513-380b67c4dfd8@gmail.com>

On Tue, Aug 11, 2020 at 07:19:00PM +0200, Patrik Nilsson wrote:
> Hi!
> 
> I have an issue with a usb to sata adapter when much high speed data is
> sent. If I erase my 4Tb SSD disk connected to the usb to sata interface with
> a "slower" /dev/urandom it works, but if I exchanges this to /dev/zero the
> disk is after a while disconnected.
> 
> For logs from udev and more debug attempts see:
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1886172
> 
> Description:
> 
> When the speed of data with a usb disk device is too high, the probes that
> check if the disk still exists is missed. This makes the disk to be
> unmounted and is remounted with an other drive letter (/dev/sdX). If this
> disk is the root ("/"), the computer hangs.
> 
> A usb SSD-disk (4 TB) connected to a USB port reproduce this bug easily.
> This has never happened with a mechanical USB-drive.
> 
> Symptoms and how to reproduce:
> 
> 1) (Disk mounted as root) Computer hangs. Only way to get started again is a
> hard reset.
> 
> 2) (Erasing disk, not mounted as root) If you erase a disk with dd ("dd
> if=/dev/zero of=/dev/sdX"), the disk disappears from the mounted list and
> reappears as an other drive letter. In my case the data is redirected to
> /dev/null and the fans speed up. The apparent data transfer speed is also
> increased. Usually this happens after about 45 to 60 minutes.

Can you provide kernel logs?  Odds are the disk disconnects itself as it
takes too much power or something.

> 3) Fill RAM with programs and make sure the swap to this disk is used.
> 
> Bug found in:
> 
> $ uname -a
> Linux pb-189b1884 5.4.0-40-generic #44-Ubuntu SMP Tue Jun 23 00:01:04 UTC
> 2020 x86_64 x86_64 x86_64 GNU/Linux
> 
> $ cat /proc/version_signature
> Ubuntu 5.4.0-40.44-generic 5.4.44

That's a really old kernel, any chance you can try 5.8?

thanks,

greg k-h

  reply	other threads:[~2020-08-11 17:56 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-08-11 17:19 high speed data to usb disk makes the kernel think that is has been unmounted Patrik Nilsson
2020-08-11 17:56 ` Greg KH [this message]
2020-08-11 18:34   ` Patrik Nilsson
2020-08-11 18:39     ` Greg KH
2020-08-11 19:31     ` Alan Stern
2020-08-11 21:14       ` Patrik Nilsson
2020-08-12  6:14         ` Greg KH
2020-08-16 15:31           ` Patrik Nilsson
2020-08-18  8:55             ` Greg KH
2020-08-18 18:22               ` Patrik Nilsson
2020-08-18 18:34                 ` Alan Stern
2020-09-15 16:43                   ` Patrik Nilsson
2020-09-15 18:24                     ` Alan Stern
2020-12-05 11:49                       ` Patrik Nilsson
2020-12-05 15:39                         ` Alan Stern

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200811175655.GB828356@kroah.com \
    --to=gregkh@linuxfoundation.org \
    --cc=linux-usb@vger.kernel.org \
    --cc=nipatriknilsson@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox