umass: Bring in small fix from NetBSD's umass

When completing a request for UFI, don't fail the request on
non-zero asc/ascq values if we've done a request sense.

This idea is from umass.c 1.100 by

umass: Bring in small fix from NetBSD's umass

When completing a request for UFI, don't fail the request on
non-zero asc/ascq values if we've done a request sense.

This idea is from umass.c 1.100 by mycroft. He used it to help
elminate the INQUIRY_SHORT quirk that we still have. However, it will
make little difference because we treat both return values the same
for CBI. And it appears we have (maybe bogusly) some devices that
specify this quirk that aren't CBI.

Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D49475

show more ...

umass: Fix a direct call to sc_transform

Call umass_std_transform instead of following the sc_transform member to
get the standard transforms now that has moved to
umass_std_transform. Pass NULL as

umass: Fix a direct call to sc_transform

Call umass_std_transform instead of following the sc_transform member to
get the standard transforms now that has moved to
umass_std_transform. Pass NULL as the CCB so that we don't stomp on the
status that was just set (and allow that in std_transform).

Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D49474

show more ...

umass: Be more consistent about how we fail bad commands

Elsewhere, if we fail a command, we use SCSI code ILLEGAL REQUEST/24/00,
but here we were using the CAM code for illegal request.

Sponsored

umass: Be more consistent about how we fail bad commands

Elsewhere, if we fail a command, we use SCSI code ILLEGAL REQUEST/24/00,
but here we were using the CAM code for illegal request.

Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D49473

show more ...

umass: Move INQUIRY and TEST UNIT READY quirks to umass_std_transform

Only doing a short inquiry and converting TUR to START STOP UNIT are the
same everywhere, so move those quirks to umass_std_tran

umass: Move INQUIRY and TEST UNIT READY quirks to umass_std_transform

Only doing a short inquiry and converting TUR to START STOP UNIT are the
same everywhere, so move those quirks to umass_std_transform.

Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D49472

show more ...

umass: Move common code into umass_std_transform

Move the length checks, and byte copying into umass_std_transform. The
copies are typically small and this simplifies the code a lot. Move
zeroing th

umass: Move common code into umass_std_transform

Move the length checks, and byte copying into umass_std_transform. The
copies are typically small and this simplifies the code a lot. Move
zeroing the buffer now to only when we change TEST UNIT READY to START
STOP.

Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D49471

show more ...

umass: make *_transform() return a bool

These are now boolean, so make them return a boolean.

Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D49470

umass: Simplify umass_std_transform to eliminate fake success

Now that nothing returns (2) to fake the success of the command,
eliminate that magic number by eliminating the case.

Sponsored by: Ne

umass: Simplify umass_std_transform to eliminate fake success

Now that nothing returns (2) to fake the success of the command,
eliminate that magic number by eliminating the case.

Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D49469

show more ...

umass: Fail SYNCHRONIZE_CACHE for UFI

Just go ahead and fail SYNCHRONIZE_CACHE commands for UFI. This standard
doesn't have a cache to flush, and the upper layers (aka da) will notice sync
cache fai

umass: Fail SYNCHRONIZE_CACHE for UFI

Just go ahead and fail SYNCHRONIZE_CACHE commands for UFI. This standard
doesn't have a cache to flush, and the upper layers (aka da) will notice sync
cache failed and won't send it again.

Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D49468

show more ...

umass: Document the ideal that it's better to fail bad commands

Generally, lying to the periph drivers about commands working is a bad
idea. They cannot get a global insight into the actual capaciti

umass: Document the ideal that it's better to fail bad commands

Generally, lying to the periph drivers about commands working is a bad
idea. They cannot get a global insight into the actual capacities of
the device when commands just work. There's good reason to intercept
bad commands at times, and to prevent the device from being exposed to
commands that, for example, hand the device. However, upper layer
periph drivers is the better place to cope with devices that don't
support the commands since it can make other inferences about the
device when it has wider knowledge.

But this represents a change in philosphy, so document that as
well. This shift happened for some commands a while ago. Many of the
quirks existed only to avoid whining errors on boot that were benign
and were a big burden. So da and cd slowly stopped the whining. Except
in cases where the command hangs the drive, we should limit new quirks
and avoid overly broad quirks (things work, but are less performant or
reliable, again mostly to avoid harmless errors that are no longer
printed).

Other sims will likely need some help with this.

Differential Revision: https://reviews.freebsd.org/D49467
Sponsored by: Netflix

show more ...

umass: fail synchronize_cache

When we know the umass device doesn't support synchronize cache
(either probed or hard quirk), fail the command w/o sending it to the
device. The da driver is the only

umass: fail synchronize_cache

When we know the umass device doesn't support synchronize cache
(either probed or hard quirk), fail the command w/o sending it to the
device. The da driver is the only driver that sends this command, and
since this filter was written has grown the ability to catch the
failure and never send the command again.

In an ideal world, we'd let the da driver work this out. However,
there's a lot of devices that hang when this command is sent in a way
that's impossible to detect other than a black list. There's also a
number that require a hard USB endpoint reset to recover, which the
autoquirk code takes care of. So umass has to act as a filter.

By failing this command, we're able to avoid upper level BIO_FLUSH
commands by tagging the underlying disk as not supporting that.

Differential Revision: https://reviews.freebsd.org/D49466
Sponsored by: Netflix

show more ...

umass: Move illegal opcode return to a function

Move illegal opcode completion of ccb to its own function. da can cope
with a number of different failures now (it didn't used to), so we
should retur

umass: Move illegal opcode return to a function

Move illegal opcode completion of ccb to its own function. da can cope
with a number of different failures now (it didn't used to), so we
should return the unnsupported commands as illegal rather than faking
completion. This will allow da to, in the future, do more intelligent
things in response to these quirks. Other SIMs will need to do this
too.

Differential Revision: https://reviews.freebsd.org/D49465
Sponsored by: Netflix

show more ...