linux-toradex.git/include/linux/libata.h, branch v7.0-rc5

Merge tag 'ata-6.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/libata/linux

2026-02-13T01:12:43+00:00

Pull ATA updates from Damien Le Moal:

 - Cleanup IRQ masking in the handling of completed report zones
   commands (Niklas)

 - Improve the handling of Thunderbolt attached devices to speed up
   device removal (Henry)

 - Several patches to generalize the existing max_sec quirks to
   facilitates quirking the maximum command size of buggy drives, many
   of which have recently showed up with the recent increase of the
   default max_sectors block limit (Niklas)

 - Cleanup the ahci-platform and sata dt-bindings schema (Rob,
   Manivannan)

 - Improve device node scan in the ahci-dwc driver (Krzysztof)

 - Remove clang W=1 warnings with the ahci-imx and ahci-xgene drivers
   (Krzysztof)

 - Fix a long standing potential command starvation situation with
   non-NCQ commands issued when NCQ commands are on-going (me)

 - Limit max_sectors to 8191 on the INTEL SSDSC2KG480G8 SSD (Niklas)

 - Remove Vesa Local Bus (VLB) support in the pata_legacy driver (Ethan)

 - Simple fixes in the pata_cypress (typo) and pata_ftide010 (timing)
   drivers (Ethan, Linus W)

* tag 'ata-6.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/libata/linux:
  ata: pata_ftide010: Fix some DMA timings
  ata: pata_cypress: fix typo in error message
  ata: pata_legacy: remove VLB support
  ata: libata-core: Quirk INTEL SSDSC2KG480G8 max_sectors
  dt-bindings: ata: sata: Document the graph port
  ata: libata-scsi: avoid Non-NCQ command starvation
  ata: libata-scsi: refactor ata_scsi_translate()
  ata: ahci-xgene: Fix Wvoid-pointer-to-enum-cast warning
  ata: ahci-imx: Fix Wvoid-pointer-to-enum-cast warning
  ata: ahci-dwc: Simplify with scoped for each OF child loop
  dt-bindings: ata: ahci-platform: Drop unnecessary select schema
  ata: libata: Allow more quirks
  ata: libata: Add libata.force parameter max_sec
  ata: libata: Add support to parse equal sign in libata.force
  ata: libata: Change libata.force to use the generic ATA_QUIRK_MAX_SEC quirk
  ata: libata: Add ata_force_get_fe_for_dev() helper
  ata: libata: Add ATA_QUIRK_MAX_SEC and convert all device quirks
  ata: libata: avoid long timeouts on hot-unplugged SATA DAS
  ata: libata-scsi: Remove superfluous local_irq_save()

scsi: Change the return type of the .queuecommand() callback

2026-01-24T02:32:34+00:00

In clang version 21.1 and later the -Wimplicit-enum-enum-cast warning
option has been introduced. This warning is enabled by default and can
be used to catch .queuecommand() implementations that return another
value than 0 or one of the SCSI_MLQUEUE_* constants. Hence this patch
that changes the return type of the .queuecommand() implementations from
'int' into 'enum scsi_qc_status'. No functionality has been changed.

Cc: Damien Le Moal 
Cc: John Garry 
Signed-off-by: Bart Van Assche 
Link: https://patch.msgid.link/20260115210357.2501991-6-bvanassche@acm.org
Signed-off-by: Martin K. Petersen

ata: libata-scsi: avoid Non-NCQ command starvation

2026-01-14T10:07:14+00:00

When a non-NCQ command is issued while NCQ commands are being executed,
ata_scsi_qc_issue() indicates to the SCSI layer that the command issuing
should be deferred by returning SCSI_MLQUEUE_XXX_BUSY.  This command
deferring is correct and as mandated by the ACS specifications since
NCQ and non-NCQ commands cannot be mixed.

However, in the case of a host adapter using multiple submission queues,
when the target device is under a constant load of NCQ commands, there
are no guarantees that requeueing the non-NCQ command will be executed
later and it may be deferred again repeatedly as other submission queues
can constantly issue NCQ commands from different CPUs ahead of the
non-NCQ command. This can lead to very long delays for the execution of
non-NCQ commands, and even complete starvation for these commands in the
worst case scenario.

Since the block layer and the SCSI layer do not distinguish between
queueable (NCQ) and non queueable (non-NCQ) commands, libata-scsi SAT
implementation must ensure forward progress for non-NCQ commands in the
presence of NCQ command traffic. This is similar to what SAS HBAs with a
hardware/firmware based SAT implementation do.

Implement such forward progress guarantee by limiting requeueing of
non-NCQ commands from ata_scsi_qc_issue(): when a non-NCQ command is
received and NCQ commands are in-flight, do not force a requeue of the
non-NCQ command by returning SCSI_MLQUEUE_XXX_BUSY and instead return 0
to indicate that the command was accepted but hold on to the qc using
the new deferred_qc field of struct ata_port.

This deferred qc will be issued using the work item deferred_qc_work
running the function ata_scsi_deferred_qc_work() once all in-flight
commands complete, which is checked with the port qc_defer() callback
return value indicating that no further delay is necessary. This check
is done using the helper function ata_scsi_schedule_deferred_qc() which
is called from ata_scsi_qc_complete(). This thus excludes this mechanism
from all internal non-NCQ commands issued by ATA EH.

When a port deferred_qc is non NULL, that is, the port has a command
waiting for the device queue to drain, the issuing of all incoming
commands (both NCQ and non-NCQ) is deferred using the regular busy
mechanism. This simplifies the code and also avoids potential denial of
service problems if a user issues too many non-NCQ commands.

Finally, whenever ata EH is scheduled, regardless of the reason, a
deferred qc is always requeued so that it can be retried once EH
completes. This is done by calling the function
ata_scsi_requeue_deferred_qc() from ata_eh_set_pending(). This avoids
the need for any special processing for the deferred qc in case of NCQ
error, link or device reset, or device timeout.

Reported-by: Xingui Yang 
Reported-by: Igor Pylypiv 
Fixes: bdb01301f3ea ("scsi: Add host and host template flag 'host_tagset'")
Cc: stable@vger.kernel.org
Signed-off-by: Damien Le Moal 
Reviewed-by: Niklas Cassel 
Reviewed-by: Martin K. Petersen 
Reviewed-by: John Garry 
Tested-by: Igor Pylypiv 
Tested-by: Xingui Yang

ata: libata: Allow more quirks

2025-12-15T07:32:05+00:00

We have currently used up 30 out of the 32-bits in the struct ata_device
struct member quirks. Thus, it is only possible to add two more quirks.

Change the struct ata_device struct member quirks from an unsigned int to
an u64.

Doing this core level change now, will make it easier for us now, as we
will not need to also do core level changes once the final two bits are
used as well.

Signed-off-by: Niklas Cassel 
Reviewed-by: Martin K. Petersen 
Reviewed-by: Damien Le Moal 
Signed-off-by: Damien Le Moal

ata: libata: Change libata.force to use the generic ATA_QUIRK_MAX_SEC quirk

2025-12-15T07:32:05+00:00

Modify the existing libata.force parameters "max_sec_128" and
"max_sec_1024" to use the generic ATA_QUIRK_MAX_SEC quirk rather than
individual quirks.

This also allows us to remove the individual quirks ATA_QUIRK_MAX_SEC_128
and ATA_QUIRK_MAX_SEC_1024.

Signed-off-by: Niklas Cassel 
Reviewed-by: Martin K. Petersen 
Reviewed-by: Damien Le Moal 
Signed-off-by: Damien Le Moal

ata: libata: Add ATA_QUIRK_MAX_SEC and convert all device quirks

2025-12-15T07:32:05+00:00

Add a new quirk ATA_QUIRK_MAX_SEC, which has a separate table with device
specific values.

Convert all existing ATA_QUIRK_MAX_SEC_XXX device quirks in
__ata_dev_quirks to the new format.

Quirks ATA_QUIRK_MAX_SEC_128 and ATA_QUIRK_MAX_SEC_1024 cannot be removed
yet, since they are also used by libata.force, which functionally, is a
separate user of the quirks. The quirks will be removed once all users
have been converted to use the new format.

The quirk ATA_QUIRK_MAX_SEC_8191 can be removed since it has no equivalent
libata.force parameter.

Signed-off-by: Niklas Cassel 
Reviewed-by: Damien Le Moal 
Reviewed-by: Martin K. Petersen 
Signed-off-by: Damien Le Moal

Merge tag 'ata-6.19-rc1-part2' of git://git.kernel.org/pub/scm/linux/kernel/git/libata/linux

2025-12-06T04:41:20+00:00

Pull ata fix from Niklas Cassel:

 - The DELLBOSS VD SATA controller times out when sending I/Os of size
   4096 KiB or larger, even though it claims LBA48 support, which per
   the ACS standard requires support for a maximum command size of
   65535 sectors, i.e. 32 MiB - 512. Thus, quirk the device so that it
   sets a lower maximum command size (me)

* tag 'ata-6.19-rc1-part2' of git://git.kernel.org/pub/scm/linux/kernel/git/libata/linux:
  ata: libata-core: Quirk DELLBOSS VD max_sectors
  ata: libata: Move quirk flags to their own enum

ata: libata-core: Quirk DELLBOSS VD max_sectors

2025-12-02T10:17:56+00:00

Commit 9b8b84879d4a ("block: Increase BLK_DEF_MAX_SECTORS_CAP") increased
the default max_sectors_kb from 1280 KiB to 4096 KiB.

DELLBOSS VD with FW rev MV.R00-0 times out when sending I/Os of size
4096 KiB.

Enable ATA_QUIRK_MAX_SEC, with value 8191 (sectors) for this device,
since any I/O with more sectors than that lead to I/O timeouts.

With this, the DELLBOSS VD SATA controller is usable again.

Cc: stable+noautosel@kernel.org # depends on Move quirk flags to their own enum
Fixes: 9b8b84879d4a ("block: Increase BLK_DEF_MAX_SECTORS_CAP")
Reviewed-by: Damien Le Moal 
Signed-off-by: Niklas Cassel

ata: libata: Move quirk flags to their own enum

2025-12-02T10:17:56+00:00

The anonymous enum in include/linux/libata.h that is used to store
various global constants can currently be backed by type int.
(It contains both negative and positive constants.)

__ATA_QUIRK_MAX is currently 31.
The quirk flags in the various global constants enum are defined as
"1U << quirk_flag_bit".

Thus if we simply add an additional quirk, the quirk flag will be 1 << 31,
which is a value that is too large to be represented by a signed int.
The various global constants enum will thus therefore be backed by type
long.

This will lead to error prints like e.g.:
ata_port_err(ap, "EH pending after %d tries, giving up\n",
	     ATA_EH_MAX_TRIES);

now failing to build, with build error:
error: format ‘%d’ expects argument of type ‘int’, but argument 4 has type ‘long int’ [-Werror=format=]

This is because all constants in the various global constants enum now
has to be printed as a long, as that is now the backing type of the enum.

Since the compiler will use the smallest possible backing type for an
enum, it is good practice to not mix unrelated things in a single enum.

Move the quirk flags to a separate enum, so that we don't need to change
the printf specifier for all other constants in the "various global
constants" enum when adding an additional quirk.

Reviewed-by: Damien Le Moal 
Reviewed-by: Martin K. Petersen 
Signed-off-by: Niklas Cassel

ata: libata-core: relax checks in ata_read_log_directory()

2025-10-13T07:12:36+00:00

Commit 6d4405b16d37 ("ata: libata-core: Cache the general purpose log
directory") introduced caching of a device general purpose log directory
to avoid repeated access to this log page during device scan. This
change also added a check on this log page to verify that the log page
version is 0x0001 as mandated by the ACS specifications.

And it turns out that some devices do not bother reporting this version,
instead reporting a version 0, resulting in error messages such as:

ata6.00: Invalid log directory version 0x0000

and to the device being marked as not supporting the general purpose log
directory log page.

Since before commit 6d4405b16d37 the log page version check did not
exist and things were still working correctly for these devices, relax
ata_read_log_directory() version check and only warn about the invalid
log page version number without disabling access to the log directory
page.

Fixes: 6d4405b16d37 ("ata: libata-core: Cache the general purpose log directory")
Cc: stable@vger.kernel.org
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=220635
Signed-off-by: Damien Le Moal 
Signed-off-by: Niklas Cassel