[smartmontools-support] Unbreaking Seagate UAS smart reporting at the kernel level

[Hans de Goede]

Hi all,

Short self-intro: I was the maintainer of the kernel's UAS driver for a
long time and I'm still somewhat involved in its development.

Recently an email was submitted to drop the code from the kernel which
automatically disables UAS ATA passthrough for all Seagate devices:
https://lore.kernel.org/linux-usb/02f8f171-635a-4ad7-f937-78f935568de2@redhat.com/T/#t

In this thread I got pointed to your wiki page about this:
https://www.smartmontools.org/wiki/SAT-with-UAS-Linux
so I thought it would be good to reach out to you to discuss this.

I am (and the other kernel devs involved are) not in favor of the suggested
change, this behavior was added to the kernel because many Seagate
enclosures completely hang upon receiving an ATA_12 or ATA_16 command, see e.g.:
https://bugzilla.kernel.org/show_bug.cgi?id=79511
https://bbs.archlinux.org/viewtopic.php?id=183190

My reasoning here is that having drives not hang (possibly while a
filesystem is mounted, causing fs damage) is more important then
having working SMART support. Note I'm not saying that SMART is not
important, I'm merly saying that having the basic storage functionality
of the drive working is even more important.

Still it would be good if we can both have our cake and eat it.

So for now the plan is to add a new US_FL_ATA_1X_OK (1) flag to the set
of flags used by the usb_storage and uas drivers and set that for known
to work Seagate models. Which would product id 0xab03 from the starter of
the kernel mailinglist thread + the 4 devices marked as known to work from:
https://www.smartmontools.org/wiki/Supported_USB-Devices

I also noticed on the SAT-with-UAS-Linux wikipage that not being able to
detect if the US_FL_NO_ATA_1X flag is set for a drive using UAS is
problematic for smartmontools. If this would be useful for you I can write
a kernel patch to expose this info somewhere in sysfs under one of the
parent devices of the drive (either the host or the USB-interface or
USB-device, I need to check where I can cleanly add this).

I can either add a dedicated sysfs file for this, or (maybe better)
report all the flags which can be set with the quirks parameter using the
same flag characters as the on the commandline / module-parameter.

Please let me know if this would be useful for you and if you have any
preferences as to what form the userspace API for this should have.

Regards,

Hans

p.s.

I know that kernel-devs have a reputation as being hard to communicate with
and I guess that for some parts of the kernel-community that is somewhat
deserved but in general we don't bite. I'm somewhat surprised that there is
a whole wikipage about this but no one ever reached out to me to discuss
this (or maybe my memory is failing me?).


1) This will only be available inside the kernel, since a
usb_storage.quirks= setting on the kernel commandline without the
t flag already allows overriding the US_FL_NO_ATA_1X flag