kernel/linux.git/drivers/ata, branch master

Merge tag 'ata-7.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/libata/linux

2026-04-15T22:03:01+00:00

Pull ata updates from Niklas Cassel: - Misc code cleanups related to tag checking and tag command completion (Damien) - Remove Baikal bt1-ahci DT binding since the upstreaming for this SoC is not going to be finalized (Andy) - Only call the libata port error handler from the SCSI error handler if there were command timeouts or if EH was scheduled for the port (Damien) - Refactor ata_scsiop_maint_in() to more clearly show that there is only one service action implemented for the MAINTENANCE IN command (me) - Clean up the handling of sysfs attributes exposed by libata (Heiner) - Let libahci_platform use a flexible array member for platform PHYs to avoid multiple allocations (Rosen) - Do not retry reset if the device has been removed/hot-unplugged (Igor) - Add missing newlines to error prints in pata_arasan_cf driver (Haoyu) - Use the correct SCSI host byte when completing deferred ATA PASS-THROUGH commands, to avoid the SCSI mid-layer from failing the commands instead of requeuing (Igor) * tag 'ata-7.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/libata/linux: ata: libata-scsi: fix requeue of deferred ATA PASS-THROUGH commands ata: pata_arasan_cf: fix missing newline in dev_err() messages ata: libata-transport: remove static variable ata_scsi_transport_template ata: libata-transport: split struct ata_internal ata: libata-transport: use static struct ata_transport_internal to simplify match functions ata: libata-transport: inline ata_attach|release_transport ata: libata-transport: instantiate struct ata_internal statically ata: libata-eh: Do not retry reset if the device is gone ata: libahci_platform: use flex array for platform PHYs ata: libata-transport: remove redundant dynamic sysfs attributes ata: libata-scsi: refactor ata_scsiop_maint_in() ata: libata-eh: avoid unnecessary calls to ata_scsi_port_error_handler() ata: ahci-dwc: Remove not-going-to-be-supported code for Baikal SoC ata: libata-scsi: rename and improve ata_qc_done() ata: libata-scsi: make ata_scsi_simulate() static ata: libata-scsi: simplify ata_scsi_requeue_deferred_qc() ata: libata-sata: simplify ata_sas_queuecmd() ata: libata-core: improve tag checks in ata_qc_issue()

ata: libata-scsi: fix requeue of deferred ATA PASS-THROUGH commands

2026-04-12T17:57:59+00:00

Commit 0ea84089dbf6 ("ata: libata-scsi: avoid Non-NCQ command starvation") introduced ata_scsi_requeue_deferred_qc() to handle commands deferred during resets or NCQ failures. This deferral logic completed commands with DID_SOFT_ERROR to trigger a retry in the SCSI mid-layer. However, DID_SOFT_ERROR is subject to scsi_cmd_retry_allowed() checks. ATA PASS-THROUGH commands sent via SG_IO ioctl have scmd->allowed set to zero. This causes the mid-layer to fail the command immediately instead of retrying, even though the command was never actually issued to the hardware. Switch to DID_REQUEUE to ensure these commands are inserted back into the request queue regardless of retry limits. Fixes: 0ea84089dbf6 ("ata: libata-scsi: avoid Non-NCQ command starvation") Reviewed-by: Damien Le Moal Signed-off-by: Igor Pylypiv Signed-off-by: Niklas Cassel

ata: ahci: force 32-bit DMA for JMicron JMB582/JMB585

2026-04-07T07:36:46+00:00

The JMicron JMB585 (and JMB582) SATA controllers advertise 64-bit DMA support via the S64A bit in the AHCI CAP register, but their 64-bit DMA implementation is defective. Under sustained I/O, DMA transfers targeting addresses above 4GB silently corrupt data -- writes land at incorrect memory addresses with no errors logged. The failure pattern is similar to the ASMedia ASM1061 (commit 20730e9b2778 ("ahci: add 43-bit DMA address quirk for ASMedia ASM1061 controllers")), which also falsely advertised full 64-bit DMA support. However, the JMB585 requires a stricter 32-bit DMA mask rather than 43-bit, as corruption occurs with any address above 4GB. On the Minisforum N5 Pro specifically, the combination of the JMB585's broken 64-bit DMA with the AMD Family 1Ah (Strix Point) IOMMU causes silent data corruption that is only detectable via checksumming filesystems (BTRFS/ZFS scrub). The corruption occurs when 32-bit IOVA space is exhausted and the kernel transparently switches to 64-bit DMA addresses. Add device-specific PCI ID entries for the JMB582 (0x0582) and JMB585 (0x0585) before the generic JMicron class match, using a new board type that combines AHCI_HFLAG_IGN_IRQ_IF_ERR (preserving existing behavior) with AHCI_HFLAG_32BIT_ONLY to force 32-bit DMA masks. Signed-off-by: Arthur Husband Reviewed-by: Damien Le Moal Signed-off-by: Niklas Cassel

ata: pata_arasan_cf: fix missing newline in dev_err() messages

2026-04-03T07:16:00+00:00

Add missing trailing newlines to dev_err() messages in pata_arasan_cf.c. This keeps the error output as properly terminated log lines. Acked-by: Viresh Kumar Signed-off-by: Haoyu Lu Reviewed-by: Damien Le Moal Signed-off-by: Niklas Cassel

ata: libata-transport: remove static variable ata_scsi_transport_template

2026-04-02T19:01:10+00:00

Simplify the code by making struct ata_scsi_transportt public, instead of using separate variable ata_scsi_transport_template. Reviewed-by: Damien Le Moal Signed-off-by: Heiner Kallweit Signed-off-by: Niklas Cassel

ata: libata-transport: split struct ata_internal

2026-04-02T19:01:10+00:00

There's no need for an umbrella struct, so remove it. It's also a prerequisite for making the embedded struct scsi_transport_template public. Reviewed-by: Damien Le Moal Signed-off-by: Heiner Kallweit Signed-off-by: Niklas Cassel

ata: libata-transport: use static struct ata_transport_internal to simplify match functions

2026-04-02T19:01:02+00:00

Both matching functions can make use of static struct ata_transport_internal. This eliminates the dependency on static variable ata_scsi_transport_template, and it allows to remove helper to_ata_internal(). Small drawback is that a forward declaration of both functions is needed. Reviewed-by: Damien Le Moal Signed-off-by: Heiner Kallweit Signed-off-by: Niklas Cassel

ata: libata-transport: inline ata_attach|release_transport

2026-04-02T18:51:00+00:00

Both functions are helpers which are used only once. So remove them and merge their code into libata_transport_init() and libata_transport_exit() respectively. Reviewed-by: Damien Le Moal Signed-off-by: Heiner Kallweit Signed-off-by: Niklas Cassel

ata: libata-transport: instantiate struct ata_internal statically

2026-04-02T18:51:00+00:00

Struct ata_internal is only instantiated once, in module init code. So we can also instantiate it statically, which allows simplifying the code. Reviewed-by: Damien Le Moal Signed-off-by: Heiner Kallweit Signed-off-by: Niklas Cassel

ata: libata-eh: Do not retry reset if the device is gone

2026-04-02T18:18:29+00:00

If a device is hot-unplugged or otherwise disappears during error handling, ata_eh_reset() may fail with -ENODEV. Currently, the error handler will continue to retry the reset operation up to max_tries times. Prevent unnecessary reset retries by exiting the loop early when ata_do_reset() returns -ENODEV. Reviewed-by: Damien Le Moal Signed-off-by: Igor Pylypiv Signed-off-by: Niklas Cassel