Eric Biggers [Sun, 21 Jul 2024 18:38:40 +0000 (11:38 -0700)]
scsi: ufs: exynos: Don't resume FMP when crypto support is disabled
If exynos_ufs_fmp_init() did not enable FMP support, then
exynos_ufs_fmp_resume() should not execute the FMP-related SMC calls.
Fixes: c96499fcb403 ("scsi: ufs: exynos: Add support for Flash Memory Protector (FMP)") Signed-off-by: Eric Biggers <ebiggers@google.com> Link: https://lore.kernel.org/r/20240721183840.209284-1-ebiggers@kernel.org Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Damien Le Moal [Fri, 19 Jul 2024 07:39:12 +0000 (16:39 +0900)]
scsi: mpt3sas: Avoid IOMMU page faults on REPORT ZONES
Some firmware versions of the 9600 series SAS HBA byte-swap the REPORT
ZONES command reply buffer from ATA-ZAC devices by directly accessing the
buffer in the host memory. This does not respect the default command DMA
direction and causes IOMMU page faults on architectures with an IOMMU
enforcing write-only mappings for DMA_FROM_DEVICE DMA driection (e.g. AMD
hosts).
Avoid such issue by always mapping the buffer of REPORT ZONES commands
using DMA_BIDIRECTIONAL (read+write IOMMU mapping). This is done by
introducing the helper function _base_scsi_dma_map() and using this helper
in _base_build_sg_scmd() and _base_build_sg_scmd_ieee() instead of calling
directly scsi_dma_map().
Fixes: 471ef9d4e498 ("mpt3sas: Build MPI SGL LIST on GEN2 HBAs and IEEE SGL LIST on GEN3 HBAs") Cc: stable@vger.kernel.org Signed-off-by: Damien Le Moal <dlemoal@kernel.org> Link: https://lore.kernel.org/r/20240719073913.179559-3-dlemoal@kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Damien Le Moal [Fri, 19 Jul 2024 07:39:11 +0000 (16:39 +0900)]
scsi: mpi3mr: Avoid IOMMU page faults on REPORT ZONES
Some firmware versions of the 9600 series SAS HBA byte-swap the REPORT
ZONES command reply buffer from ATA-ZAC devices by directly accessing the
buffer in the host memory. This does not respect the default command DMA
direction and causes IOMMU page faults on architectures with an IOMMU
enforcing write-only mappings for DMA_FROM_DEVICE DMA direction (e.g. AMD
hosts), leading to the device capacity to be dropped to 0:
Avoid this issue by always mapping the buffer of REPORT ZONES commands
using DMA_BIDIRECTIONAL, that is, using a read-write IOMMU mapping.
Suggested-by: Christoph Hellwig <hch@lst.de> Fixes: 023ab2a9b4ed ("scsi: mpi3mr: Add support for queue command processing") Cc: stable@vger.kernel.org Signed-off-by: Damien Le Moal <dlemoal@kernel.org> Link: https://lore.kernel.org/r/20240719073913.179559-2-dlemoal@kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
scsi: ufs: core: Do not set link to OFF state while waking up from hibernation
UFS link is just put into hibern8 state during the 'freeze' process of the
hibernation. Afterwards, the system may get powered down. But that doesn't
matter during wakeup. Because during wakeup from hibernation, UFS link is
again put into hibern8 state by the restore kernel and then the control is
handed over to the to image kernel.
So in both the places, UFS link is never turned OFF. But
ufshcd_system_restore() just assumes that the link will be in OFF state and
sets the link state accordingly. And this breaks hibernation wakeup:
The offending commit tried to suppress a double "Starting disk" message for
some drivers, but instead started spamming the log with bogus messages
every five seconds:
[ 311.798956] sd 0:0:0:0: [sda] Starting disk
[ 316.919103] sd 0:0:0:0: [sda] Starting disk
[ 322.040775] sd 0:0:0:0: [sda] Starting disk
[ 327.161140] sd 0:0:0:0: [sda] Starting disk
[ 332.281352] sd 0:0:0:0: [sda] Starting disk
[ 337.401878] sd 0:0:0:0: [sda] Starting disk
[ 342.521527] sd 0:0:0:0: [sda] Starting disk
[ 345.850401] sd 0:0:0:0: [sda] Starting disk
[ 350.967132] sd 0:0:0:0: [sda] Starting disk
[ 356.090454] sd 0:0:0:0: [sda] Starting disk
...
on machines that do not actually stop the disk on runtime suspend (e.g.
the Qualcomm sc8280xp CRD with UFS).
Let's just revert for now to address the regression.
Fixes: 7a6bbc2829d4 ("scsi: sd: Do not repeat the starting disk message") Cc: stable@vger.kernel.org Signed-off-by: Johan Hovold <johan+linaro@kernel.org> Link: https://lore.kernel.org/r/20240716161101.30692-1-johan+linaro@kernel.org Reviewed-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Damien Le Moal <dlemoal@kernel.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Kyoungrul Kim [Tue, 9 Jul 2024 23:25:20 +0000 (08:25 +0900)]
scsi: ufs: core: Check LSDBS cap when !mcq
If the user sets use_mcq_mode to 0, the host will try to activate the LSDB
mode unconditionally even when the LSDBS of device HCI cap is 1. This makes
commands time out and causes device probing to fail.
To prevent that problem, check the LSDBS cap when MCQ is not supported.
Zhongqiu Han [Fri, 5 Jul 2024 10:36:14 +0000 (18:36 +0800)]
scsi: aha152x: Use DECLARE_COMPLETION_ONSTACK for non-constant completion
The _ONSTACK variant should be used for on-stack completion, otherwise it
will break lockdep. See also commit 6e9a4738c9fa ("[PATCH] completions:
lockdep annotate on stack completions").
For fabric scan, current code uses switch scan opcode and flags as the
method to iterate through different commands to carry out the process.
This makes it hard to read. This patch convert those opcode and flags into
steps. In addition, this help reduce some duplicate code.
scsi: qla2xxx: Fix optrom version displayed in FDMI
Bios version was popluated for FDMI response. Systems with EFI would show
optrom version as 0. EFI version is populated here and BIOS version is
already displayed under FDMI_HBA_BOOT_BIOS_NAME.
scsi: qla2xxx: During vport delete send async logout explicitly
During vport delete, it is observed that during unload we hit a crash
because of stale entries in outstanding command array. For all these stale
I/O entries, eh_abort was issued and aborted (fast_fail_io = 2009h) but
I/Os could not complete while vport delete is in process of deleting.
The command completion was done early while aborting the commands in driver
unload path but outside lock to avoid the WARN_ON condition of performing
dma_free_attr within the lock. However this caused race condition while
command completion via multiple paths causing system crash.
Hence complete the command early in unload path but within the lock to
avoid race condition.
Link up failure is observed as a result of flash read failure. Current
code does not check flash read return code where it relies on FW checksum
to detect the problem.
Add check of flash read failure to detect the problem sooner.
Reported-by: kernel test robot <lkp@intel.com> Reported-by: Dan Carpenter <dan.carpenter@linaro.org> Closes: https://lore.kernel.org/all/202406210815.rPDRDMBi-lkp@intel.com/ Cc: stable@vger.kernel.org Signed-off-by: Quinn Tran <qutran@marvell.com> Signed-off-by: Nilesh Javali <njavali@marvell.com> Link: https://lore.kernel.org/r/20240710171057.35066-6-njavali@marvell.com Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
scsi: qla2xxx: Return ENOBUFS if sg_cnt is more than one for ELS cmds
Firmware only supports single DSDs in ELS Pass-through IOCB (0x53h), sg cnt
is decided by the SCSI ML. User is not aware of the cause of an acutal
error.
Return the appropriate return code that will be decoded by API and
application and proper error message will be displayed to user.
Fixes: 6e98016ca077 ("[SCSI] qla2xxx: Re-organized BSG interface specific code.") Cc: stable@vger.kernel.org Signed-off-by: Saurav Kashyap <skashyap@marvell.com> Signed-off-by: Nilesh Javali <njavali@marvell.com> Link: https://lore.kernel.org/r/20240710171057.35066-5-njavali@marvell.com Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
scsi: qla2xxx: Unable to act on RSCN for port online
The device does not come online when the target port is online. There were
multiple RSCNs indicating multiple devices were affected. Driver is in the
process of finishing a fabric scan. A new RSCN (device up) arrived at the
tail end of the last fabric scan. Driver mistakenly thinks the new RSCN is
being taken care of by the previous fabric scan, where this notification is
cleared and not acted on. The laser needs to be blinked again to get the
device to show up.
To prevent driver from accidentally clearing the RSCN notification, each
RSCN is given a generation value. A fabric scan will scan for that
generation(s). Any new RSCN arrive after the scan start will have a new
generation value. This will trigger another scan to get latest data. The
RSCN notification flag will be cleared when the scan is associate to that
generation.
Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202406210538.w875N70K-lkp@intel.com/ Fixes: bb2ca6b3f09a ("scsi: qla2xxx: Relogin during fabric disturbance") Cc: stable@vger.kernel.org Signed-off-by: Quinn Tran <qutran@marvell.com> Signed-off-by: Nilesh Javali <njavali@marvell.com> Link: https://lore.kernel.org/r/20240710171057.35066-2-njavali@marvell.com Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Martin K. Petersen [Thu, 11 Jul 2024 02:33:34 +0000 (22:33 -0400)]
Merge patch series "Basic inline encryption support for ufs-exynos"
Eric Biggers <ebiggers@kernel.org> says:
Add support for Flash Memory Protector (FMP), which is the inline
encryption hardware on Exynos and Exynos-based SoCs.
Specifically, add support for the "traditional FMP mode" that works on
many Exynos-based SoCs including gs101. This is the mode that uses
"software keys" and is compatible with the upstream kernel's existing
inline encryption framework in the block and filesystem layers. I
plan to add support for the wrapped key support on gs101 at a later
time.
Tested on gs101 (specifically Pixel 6) by running the 'encrypt' group
of xfstests on a filesystem mounted with the 'inlinecrypt' mount
option.
This patchset applies to v6.10-rc6, and it has no prerequisites that
aren't already upstream.
Eric Biggers [Mon, 8 Jul 2024 23:53:30 +0000 (16:53 -0700)]
scsi: ufs: exynos: Add support for Flash Memory Protector (FMP)
Add support for Flash Memory Protector (FMP), which is the inline
encryption hardware on Exynos and Exynos-based SoCs.
Specifically, add support for the "traditional FMP mode" that works on many
Exynos-based SoCs including gs101. This is the mode that uses "software
keys" and is compatible with the upstream kernel's existing inline
encryption framework in the block and filesystem layers. I plan to add
support for the wrapped key support on gs101 at a later time.
Tested on gs101 (specifically Pixel 6) by running the 'encrypt' group of
xfstests on a filesystem mounted with the 'inlinecrypt' mount option.
Signed-off-by: Eric Biggers <ebiggers@google.com> Link: https://lore.kernel.org/r/20240708235330.103590-7-ebiggers@kernel.org Reviewed-by: Peter Griffin <peter.griffin@linaro.org> Tested-by: Peter Griffin <peter.griffin@linaro.org> Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Eric Biggers [Mon, 8 Jul 2024 23:53:29 +0000 (16:53 -0700)]
scsi: ufs: core: Add UFSHCD_QUIRK_KEYS_IN_PRDT
Since the nonstandard inline encryption support on Exynos SoCs requires
that raw cryptographic keys be copied into the PRDT, it is desirable to
zeroize those keys after each request to keep them from being left in
memory. Therefore, add a quirk bit that enables the zeroization.
We could instead do the zeroization unconditionally. However, using a
quirk bit avoids adding the zeroization overhead to standard devices.
Reviewed-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Peter Griffin <peter.griffin@linaro.org> Signed-off-by: Eric Biggers <ebiggers@google.com> Link: https://lore.kernel.org/r/20240708235330.103590-6-ebiggers@kernel.org Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Eric Biggers [Mon, 8 Jul 2024 23:53:28 +0000 (16:53 -0700)]
scsi: ufs: core: Add fill_crypto_prdt variant op
Add a variant op to allow host drivers to initialize nonstandard
crypto-related fields in the PRDT. This is needed to support inline
encryption on the "Exynos" UFS controller.
Note that this will be used together with the support for overriding the
PRDT entry size that was already added by commit ada1e653a5ea ("scsi: ufs:
core: Allow UFS host drivers to override the sg entry size").
Reviewed-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Peter Griffin <peter.griffin@linaro.org> Signed-off-by: Eric Biggers <ebiggers@google.com> Link: https://lore.kernel.org/r/20240708235330.103590-5-ebiggers@kernel.org Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Add UFSHCD_QUIRK_BROKEN_CRYPTO_ENABLE which tells the UFS core to not use
the crypto enable bit defined by the UFS specification. This is needed to
support inline encryption on the "Exynos" UFS controller.
Reviewed-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Peter Griffin <peter.griffin@linaro.org> Signed-off-by: Eric Biggers <ebiggers@google.com> Link: https://lore.kernel.org/r/20240708235330.103590-4-ebiggers@kernel.org Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Eric Biggers [Mon, 8 Jul 2024 23:53:26 +0000 (16:53 -0700)]
scsi: ufs: core: fold ufshcd_clear_keyslot() into its caller
Fold ufshcd_clear_keyslot() into its only remaining caller.
Reviewed-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Peter Griffin <peter.griffin@linaro.org> Signed-off-by: Eric Biggers <ebiggers@google.com> Link: https://lore.kernel.org/r/20240708235330.103590-3-ebiggers@kernel.org Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Add UFSHCD_QUIRK_CUSTOM_CRYPTO_PROFILE which lets UFS host drivers
initialize the blk_crypto_profile themselves rather than have it be
initialized by ufshcd-core according to the UFSHCI standard. This is
needed to support inline encryption on the "Exynos" UFS controller which
has a nonstandard interface.
Reviewed-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Peter Griffin <peter.griffin@linaro.org> Signed-off-by: Eric Biggers <ebiggers@google.com> Link: https://lore.kernel.org/r/20240708235330.103590-2-ebiggers@kernel.org Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
UFSHCI controllers that are compliant with the UFSHCI 4.0 standard report
the maximum number of supported commands in the controller capabilities
register. Use that value if .get_hba_mac == NULL.
Reviewed-by: Peter Wang <peter.wang@mediatek.com> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Link: https://lore.kernel.org/r/20240708211716.2827751-11-bvanassche@acm.org Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Make ufshcd_mcq_decide_queue_depth() easier to read by inlining
ufshcd_mcq_vops_get_hba_mac().
Reviewed-by: Peter Wang <peter.wang@mediatek.com> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Link: https://lore.kernel.org/r/20240708211716.2827751-10-bvanassche@acm.org Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Move the ufshcd_mcq_enable() call from inside ufshcd_config_mcq() to the
callers of this function. No functionality is changed by this patch. This
patch makes a later patch easier to read ("scsi: ufs: Make .get_hba_mac()
optional").
Cc: Peter Wang <peter.wang@mediatek.com> Cc: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Link: https://lore.kernel.org/r/20240708211716.2827751-9-bvanassche@acm.org Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
scsi: ufs: mcq: Move the "hba->mcq_enabled = true" assignment
Move the "hba->mcq_enabled = true" assignment to prevent that it gets
duplicated by a later patch that will introduce more ufshcd_mcq_enable()
calls. No functionality is changed by this patch.
Improve code readability by inlining is_mcq_enabled().
Cc: Peter Wang <peter.wang@mediatek.com> Cc: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Link: https://lore.kernel.org/r/20240708211716.2827751-7-bvanassche@acm.org Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Move the hba->reserved_slot and the host->can_queue assignments from
ufshcd_config_mcq() into ufshcd_alloc_mcq(). The advantages of this change
are as follows:
- It becomes easier to verify that these two parameters are updated if
hba->nutrs is updated.
- It prevents unnecessary assignments to these two parameters. While
ufshcd_config_mcq() is called during host reset, ufshcd_alloc_mcq() is
not.
Cc: Can Guo <quic_cang@quicinc.com> Reviewed-by: Peter Wang <peter.wang@mediatek.com> Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Link: https://lore.kernel.org/r/20240708211716.2827751-6-bvanassche@acm.org Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
scsi: ufs: core: Rename the MASK_TRANSFER_REQUESTS_SLOTS constant
Rename this constant to prepare for the introduction of the
MASK_TRANSFER_REQUESTS_SLOTS_MCQ constant. The acronym "SDB" stands for
"single doorbell" (mode).
Reviewed-by: Peter Wang <peter.wang@mediatek.com> Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Link: https://lore.kernel.org/r/20240708211716.2827751-5-bvanassche@acm.org Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The SCSI host template members .cmd_per_lun and .can_queue are copied into
the SCSI host data structure. Before these are used, these are overwritten
by ufshcd_init(). Hence, this patch does not change any functionality.
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Reviewed-by: Keoseong Park <keosung.park@samsung.com> Reviewed-by: Peter Wang <peter.wang@mediatek.com> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Link: https://lore.kernel.org/r/20240708211716.2827751-4-bvanassche@acm.org Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
scsi: ufs: core: Initialize struct uic_command once
Instead of first zero-initializing struct uic_command and next initializing
it memberwise, initialize all members at once.
Reviewed-by: Daejun Park <daejun7.park@samsung.com> Reviewed-by: Avri Altman <avri.altman@wdc.com> Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Reviewed-by: Peter Wang <peter.wang@mediatek.com> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Link: https://lore.kernel.org/r/20240708211716.2827751-3-bvanassche@acm.org Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Several functions are declared in include/ufs/ufshcd.h and also in
drivers/ufs/core/ufshcd-priv.h. Remove the duplicate declarations.
Reviewed-by: Peter Wang <peter.wang@mediatek.com> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Link: https://lore.kernel.org/r/20240708211716.2827751-2-bvanassche@acm.org Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Reviewed-by: Keoseong Park <keosung.park@samsung.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Martin K. Petersen [Fri, 5 Jul 2024 03:38:16 +0000 (23:38 -0400)]
Merge patch series "mpi3mr: Support PCI Error Recovery"
Sumit Saxena <sumit.saxena@broadcom.com> says:
This patch series contains the changes done in the driver to support
PCI error recovery. It is rework of older patch series from Ranjan
Kumar, see [1].
Sumit Saxena [Thu, 27 Jun 2024 10:17:33 +0000 (15:47 +0530)]
scsi: mpi3mr: Support PCI Error Recovery callback handlers
PCI Error recovery support is required to recover the controller upon
detection of PCI errors. Add support for the PCI error recovery callback
handlers in mpi3mr driver.
Justin Tee [Fri, 28 Jun 2024 17:20:10 +0000 (10:20 -0700)]
scsi: lpfc: Revise lpfc_prep_embed_io routine with proper endian macro usages
On big endian architectures, it is possible to run into a memory out of
bounds pointer dereference when FCP targets are zoned.
In lpfc_prep_embed_io, the memcpy(ptr, fcp_cmnd, sgl->sge_len) is
referencing a little endian formatted sgl->sge_len value. So, the memcpy
can cause big endian systems to crash.
Redefine the *sgl ptr as a struct sli4_sge_le to make it clear that we are
referring to a little endian formatted data structure. And, update the
routine with proper le32_to_cpu macro usages.
Justin Tee [Fri, 28 Jun 2024 17:20:09 +0000 (10:20 -0700)]
scsi: lpfc: Fix incorrect request len mbox field when setting trunking via sysfs
When setting trunk modes through sysfs, the SLI_CONFIG mailbox command's
command payload length is incorrectly hardcoded to 12 bytes. SLI_CONFIG's
payload length field should be specified large enough to encompass both the
submailbox command header and the submailbox request itself.
Thus, replace the hardcoded 12 bytes with a clearer calculation by way of
sizeof(struct lpfc_mbx_set_trunk_mode) - sizeof(struct lpfc_sli4_cfg_mhdr).
Justin Tee [Fri, 28 Jun 2024 17:20:08 +0000 (10:20 -0700)]
scsi: lpfc: Handle mailbox timeouts in lpfc_get_sfp_info
The MBX_TIMEOUT return code is not handled in lpfc_get_sfp_info and the
routine unconditionally frees submitted mailbox commands regardless of
return status. The issue is that for MBX_TIMEOUT cases, when firmware
returns SFP information at a later time, that same mailbox memory region
references previously freed memory in its cmpl routine.
Fix by adding checks for the MBX_TIMEOUT return code. During mailbox
resource cleanup, check the mbox flag to make sure that the wait did not
timeout. If the MBOX_WAKE flag is not set, then do not free the resources
because it will be freed when firmware completes the mailbox at a later
time in its cmpl routine.
Also, increase the timeout from 30 to 60 seconds to accommodate boot
scripts requiring longer timeouts.
Justin Tee [Fri, 28 Jun 2024 17:20:07 +0000 (10:20 -0700)]
scsi: lpfc: Fix handling of fully recovered fabric node in dev_loss callbk
In rare cases when a fabric node is recovered after a link bounce and
before dev_loss_tmo callbk is reached, the driver may leave the fabric node
in an inconsistent state with the NLP_IN_DEV_LOSS flag perpetually set.
In lpfc_dev_loss_tmo_callbk, a check is added for a recovered fabric node.
If the node is recovered, then don't queue the lpfc_dev_loss_tmo_handler
work. In lpfc_dev_loss_tmo_handler, the path taken for the recovered fabric
nodes is updated to clear the NLP_IN_DEV_LOSS flag.
Justin Tee [Fri, 28 Jun 2024 17:20:06 +0000 (10:20 -0700)]
scsi: lpfc: Relax PRLI issue conditions after GID_FT response
If previously in REG_LOGIN_ISSUE state, then remove the requirement that
PLOGI must have been received from the remote port before issuing a PRLI.
After GID_FT completes, it does not matter whether the driver itself sent a
PLOGI or received one. The fact that we're in REG_LOGIN_ISSUE state simply
means that the next state should be issuing the PRLI to continue discovery
of the remote port.
Justin Tee [Fri, 28 Jun 2024 17:20:05 +0000 (10:20 -0700)]
scsi: lpfc: Allow DEVICE_RECOVERY mode after RSCN receipt if in PRLI_ISSUE state
Certain vendor specific targets initially register with the fabric as an
initiator function first and then re-register as a target function
afterwards.
The timing of the target function re-registration can cause a race
condition such that the driver is stuck assuming the remote port as an
initiator function and never discovers the target's hosted LUNs.
Expand the nlp_state qualifier to also include NLP_STE_PRLI_ISSUE because
the state means that PRLI was issued but we have not quite reached
MAPPED_NODE state yet. If we received an RSCN in the PRLI_ISSUE state,
then we should restart discovery again by going into DEVICE_RECOVERY.
Fixes: dded1dc31aa4 ("scsi: lpfc: Modify when a node should be put in device recovery mode during RSCN") Cc: <stable@vger.kernel.org> # v6.6+ Signed-off-by: Justin Tee <justin.tee@broadcom.com> Link: https://lore.kernel.org/r/20240628172011.25921-3-justintee8345@gmail.com Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Justin Tee [Fri, 28 Jun 2024 17:20:04 +0000 (10:20 -0700)]
scsi: lpfc: Cancel ELS WQE instead of issuing abort when SLI port is inactive
During SLI port errata events, there should be no expectation that
submitted outstanding WQEs will return back CQEs. In these situations, the
driver should not rely on receiving CQEs from the SLI port to signal WQE
resource clean up.
Put an sli_flag LPFC_SLI_ACTIVE check in lpfc_els_flush_cmd() when walking
the txcmplq. The sli_flag check helps determine whether to issue an abort
or driver based cancel on outstanding WQEs. If !LPFC_SLI_ACTIVE, then
there's no point to issue anything to the SLI port. Instead, let the
driver based cancel logic clean up the submitted WQE resources.
Also, enhance some abort log messages that help with future debugging.
Damien Le Moal [Mon, 1 Jul 2024 21:53:26 +0000 (06:53 +0900)]
scsi: sd: Do not repeat the starting disk message
The SCSI disk message "Starting disk" to signal resuming of a suspended
disk is printed in both sd_resume() and sd_resume_common() which results
in this message being printed twice when resuming from e.g. autosuspend:
$ echo 5000 > /sys/block/sda/device/power/autosuspend_delay_ms
$ echo auto > /sys/block/sda/device/power/control
[ 4972.805851] sd 0:0:0:0: [sda] Starting disk
[ 4980.558806] sd 0:0:0:0: [sda] Starting disk
Fix this double print by removing the call to sd_printk() from sd_resume()
and moving the call to sd_printk() in sd_resume_common() earlier in the
function, before the check using sd_do_start_stop(). Doing so, the message
is printed once regardless if sd_resume_common() actually executes
sd_start_stop_device() (i.e. SCSI device case) or not (libsas and libata
managed ATA devices case).
Fixes: 0c76106cb975 ("scsi: sd: Fix TCG OPAL unlock on system resume") Cc: stable@vger.kernel.org Signed-off-by: Damien Le Moal <dlemoal@kernel.org> Link: https://lore.kernel.org/r/20240701215326.128067-1-dlemoal@kernel.org Reviewed-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: John Garry <john.g.garry@oracle.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Peter Wang [Fri, 28 Jun 2024 07:00:30 +0000 (15:00 +0800)]
scsi: ufs: core: Fix ufshcd_abort_one racing issue
When ufshcd_abort_one is racing with the completion ISR, the completed tag
of the request's mq_hctx pointer will be set to NULL by ISR. Return
success when request is completed by ISR because ufshcd_abort_one does not
need to do anything.
Below is KE back trace.
ufshcd_try_to_abort_task: cmd at tag 41 not pending in the device.
ufshcd_try_to_abort_task: cmd at tag=41 is cleared.
Aborting tag 41 / CDB 0x28 succeeded
Unable to handle kernel NULL pointer dereference at virtual address 0000000000000194
pc : [0xffffffddd7a79bf8] blk_mq_unique_tag+0x8/0x14
lr : [0xffffffddd6155b84] ufshcd_mcq_req_to_hwq+0x1c/0x40 [ufs_mediatek_mod_ise]
do_mem_abort+0x58/0x118
el1_abort+0x3c/0x5c
el1h_64_sync_handler+0x54/0x90
el1h_64_sync+0x68/0x6c
blk_mq_unique_tag+0x8/0x14
ufshcd_err_handler+0xae4/0xfa8 [ufs_mediatek_mod_ise]
process_one_work+0x208/0x4fc
worker_thread+0x228/0x438
kthread+0x104/0x1d4
ret_from_fork+0x10/0x20
Fixes: 93e6c0e19d5b ("scsi: ufs: core: Clear cmd if abort succeeds in MCQ mode") Suggested-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Peter Wang <peter.wang@mediatek.com> Link: https://lore.kernel.org/r/20240628070030.30929-3-peter.wang@mediatek.com Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Peter Wang [Fri, 28 Jun 2024 07:00:29 +0000 (15:00 +0800)]
scsi: ufs: core: Fix ufshcd_clear_cmd racing issue
When ufshcd_clear_cmd is racing with the completion ISR, the completed tag
of the request's mq_hctx pointer will be set to NULL by the ISR. And
ufshcd_clear_cmd's call to ufshcd_mcq_req_to_hwq will get NULL pointer KE.
Return success when the request is completed by ISR because sq does not
need cleanup.
ufshcd_try_to_abort_task: cmd pending in the device. tag = 6
Unable to handle kernel NULL pointer dereference at virtual address 0000000000000194
pc : [0xffffffd589679bf8] blk_mq_unique_tag+0x8/0x14
lr : [0xffffffd5862f95b4] ufshcd_mcq_sq_cleanup+0x6c/0x1cc [ufs_mediatek_mod_ise]
Workqueue: ufs_eh_wq_0 ufshcd_err_handler [ufs_mediatek_mod_ise]
Call trace:
dump_backtrace+0xf8/0x148
show_stack+0x18/0x24
dump_stack_lvl+0x60/0x7c
dump_stack+0x18/0x3c
mrdump_common_die+0x24c/0x398 [mrdump]
ipanic_die+0x20/0x34 [mrdump]
notify_die+0x80/0xd8
die+0x94/0x2b8
__do_kernel_fault+0x264/0x298
do_page_fault+0xa4/0x4b8
do_translation_fault+0x38/0x54
do_mem_abort+0x58/0x118
el1_abort+0x3c/0x5c
el1h_64_sync_handler+0x54/0x90
el1h_64_sync+0x68/0x6c
blk_mq_unique_tag+0x8/0x14
ufshcd_clear_cmd+0x34/0x118 [ufs_mediatek_mod_ise]
ufshcd_try_to_abort_task+0x2c8/0x5b4 [ufs_mediatek_mod_ise]
ufshcd_err_handler+0xa7c/0xfa8 [ufs_mediatek_mod_ise]
process_one_work+0x208/0x4fc
worker_thread+0x228/0x438
kthread+0x104/0x1d4
ret_from_fork+0x10/0x20
Fixes: 8d7290348992 ("scsi: ufs: mcq: Add supporting functions for MCQ abort") Suggested-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Peter Wang <peter.wang@mediatek.com> Link: https://lore.kernel.org/r/20240628070030.30929-2-peter.wang@mediatek.com Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Terrence Adams [Thu, 27 Jun 2024 15:59:24 +0000 (15:59 +0000)]
scsi: pm8001: Update log level when reading config table
Reading the main config table occurs as a part of initialization in
pm80xx_chip_init(). Because of this it makes more sense to have it be a
part of the INIT logging.
Igor Pylypiv [Thu, 27 Jun 2024 15:59:23 +0000 (15:59 +0000)]
scsi: pm80xx: Set phy->enable_completion only when we wait for it
pm8001_phy_control() populates the enable_completion pointer with a stack
address, sends a PHY_LINK_RESET / PHY_HARD_RESET, waits 300 ms, and
returns. The problem arises when a phy control response comes late. After
300 ms the pm8001_phy_control() function returns and the passed
enable_completion stack address is no longer valid. Late phy control
response invokes complete() on a dangling enable_completion pointer which
leads to a kernel crash.
Signed-off-by: Igor Pylypiv <ipylypiv@google.com> Signed-off-by: Terrence Adams <tadamsjr@google.com> Link: https://lore.kernel.org/r/20240627155924.2361370-2-tadamsjr@google.com Acked-by: Jack Wang <jinpu.wang@ionos.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Kyoungrul Kim [Thu, 27 Jun 2024 08:51:04 +0000 (17:51 +0900)]
scsi: ufs: core: Remove SCSI host only if added
If host tries to remove ufshcd driver from a UFS device it would cause a
kernel panic if ufshcd_async_scan fails during ufshcd_probe_hba before
adding a SCSI host with scsi_add_host and MCQ is enabled since SCSI host
has been defered after MCQ configuration introduced by commit 0cab4023ec7b
("scsi: ufs: core: Defer adding host to SCSI if MCQ is supported").
To guarantee that SCSI host is removed only if it has been added, set the
scsi_host_added flag to true after adding a SCSI host and check whether it
is set or not before removing it.
scsi: ufs: core: Suspend clk scaling on no request
Currently UFS clk scaling is getting suspended only when the clks are
scaled down. When high load is generated, a huge amount of latency is added
due to scaling up the clk and completing the request post that.
Suspending the scaling in its existing state when high load is generated
improves the random performance KPI by 28%. So suspending the scaling when
there are no requests. And the clk would be put in low scaled state when
the actual request load is low.
Make this change optional by having the check enabled using vops since for
some devices suspending without bringing the clk in low scaled state might
have impact on power consumption of the SoC.
Martin K. Petersen [Thu, 27 Jun 2024 03:30:54 +0000 (23:30 -0400)]
Merge patch series "mpi3mr: Host diag buffer support"
Ranjan Kumar <ranjan.kumar@broadcom.com> says:
The controllers managed by mpi3mr driver requires system memory to
save hardware and firmware diagnostic information, this patch set
enhances the drivers to provide host memory to the controller for
diagnostic information. This patch set also provides driver changes
to push kernel messages into the diagnostic buffers reserved for the
driver, so that the information will be available as part of debug
data fetched from the controller. In addition, support for
configuring automatic diagnostic information is added in the driver.
Ranjan Kumar [Wed, 26 Jun 2024 10:26:44 +0000 (15:56 +0530)]
scsi: mpi3mr: Trigger support
Add functions to process automatic diag triggers. If a condition defined in
the triggers is met, the driver will call appropriate controller functions
to save the diagnostic information.
Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202405151955.BiAWI1SY-lkp@intel.com/ Co-developed-by: Sathya Prakash <sathya.prakash@broadcom.com> Signed-off-by: Sathya Prakash <sathya.prakash@broadcom.com> Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com> Link: https://lore.kernel.org/r/20240626102646.14298-3-ranjan.kumar@broadcom.com Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Ranjan Kumar [Wed, 26 Jun 2024 10:26:43 +0000 (15:56 +0530)]
scsi: mpi3mr: HDB allocation and posting for hardware and firmware buffers
To be able to debug controller problems it is beneficial to allocate and
configure system/host memory buffers which can be used to capture hardware
and firmware diagnostic information.
Add functions required to allocate and post firmware and hardware
diagnostic buffers to the controller and to set up automatic diagnostic
capture triggers.
Captures will be triggered under the following circumstances:
1. Firmware is in FAULT state.
2. Admin commands time out.
3. Controller reset caused due to I/O timeout
Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202405151758.7xrJz6rp-lkp@intel.com/ Co-developed-by: Sathya Prakash <sathya.prakash@broadcom.com> Signed-off-by: Sathya Prakash <sathya.prakash@broadcom.com> Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com> Link: https://lore.kernel.org/r/20240626102646.14298-2-ranjan.kumar@broadcom.com Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Huai-Yuan Liu [Fri, 21 Jun 2024 08:25:45 +0000 (16:25 +0800)]
scsi: lpfc: Fix a possible null pointer dereference
In function lpfc_xcvr_data_show, the memory allocation with kmalloc might
fail, thereby making rdp_context a null pointer. In the following context
and functions that use this pointer, there are dereferencing operations,
leading to null pointer dereference.
To fix this issue, a null pointer check should be added. If it is null,
use scnprintf to notify the user and return len.
Fixes: 479b0917e447 ("scsi: lpfc: Create a sysfs entry called lpfc_xcvr_data for transceiver info") Signed-off-by: Huai-Yuan Liu <qq810974084@gmail.com> Link: https://lore.kernel.org/r/20240621082545.449170-1-qq810974084@gmail.com Reviewed-by: Justin Tee <justin.tee@broadcom.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Target debugfs entry is removed via async_schedule() which isn't drained
when adding same name target, so failure of "Directory 'target11:0:0' with
parent 'scsi_debug' already present!" can be triggered easily.
Fix it by switching to domain async schedule, and draining it before
adding new target debugfs entry.
Cc: Wenchao Hao <haowenchao2@huawei.com> Fixes: f084fe52c640 ("scsi: scsi_debug: Add debugfs interface to fail target reset") Signed-off-by: Ming Lei <ming.lei@redhat.com> Acked-by: Wenchao Hao <haowenchao22@gmail.com> Link: https://lore.kernel.org/r/20240619013803.3008857-1-ming.lei@redhat.com Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Bart Van Assche [Thu, 13 Jun 2024 21:18:27 +0000 (14:18 -0700)]
scsi: usb: uas: Do not query the IO Advice Hints Grouping mode page for USB/UAS devices
Recently it was reported that the following USB storage devices are
unusable with Linux kernel 6.9:
* Kingston DataTraveler G2
* Garmin FR35
This is because attempting to read the IO Advice Hints Grouping mode page
causes these devices to reset. Hence do not read the IO Advice Hints
Grouping mode page from USB/UAS storage devices.
Acked-by: Alan Stern <stern@rowland.harvard.edu> Cc: stable@vger.kernel.org Fixes: 4f53138fffc2 ("scsi: sd: Translate data lifetime information") Reported-by: Joao Machado <jocrismachado@gmail.com> Closes: https://lore.kernel.org/linux-scsi/20240130214911.1863909-1-bvanassche@acm.org/T/#mf4e3410d8f210454d7e4c3d1fb5c0f41e651b85f Tested-by: Andy Shevchenko <andy.shevchenko@gmail.com> Bisected-by: Christian Heusel <christian@heusel.eu> Reported-by: Andy Shevchenko <andy.shevchenko@gmail.com> Closes: https://lore.kernel.org/linux-scsi/CACLx9VdpUanftfPo2jVAqXdcWe8Y43MsDeZmMPooTzVaVJAh2w@mail.gmail.com/ Signed-off-by: Bart Van Assche <bvanassche@acm.org> Link: https://lore.kernel.org/r/20240613211828.2077477-3-bvanassche@acm.org Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Joel Slebodnick [Thu, 13 Jun 2024 18:27:28 +0000 (14:27 -0400)]
scsi: ufs: core: Free memory allocated for model before reinit
Under the conditions that a device is to be reinitialized within
ufshcd_probe_hba(), the device must first be fully reset.
Resetting the device should include freeing U8 model (member of dev_info)
but does not, and this causes a memory leak. ufs_put_device_desc() is
responsible for freeing model.
Fixes: 96a7141da332 ("scsi: ufs: core: Add support for reinitializing the UFS device") Cc: <stable@vger.kernel.org> Reviewed-by: Andrew Halaney <ahalaney@redhat.com> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Joel Slebodnick <jslebodn@redhat.com> Link: https://lore.kernel.org/r/20240613200202.2524194-1-jslebodn@redhat.com Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Bart Van Assche [Wed, 12 Jun 2024 17:15:21 +0000 (10:15 -0700)]
scsi: core: Fix an incorrect comment
The comment that scsi_static_device_list would go away was added more than
18 years ago. Today, that list is still there and a large number of
additional entries have been added. This shows that this comment is
incorrect. Hence fix that comment.
Damien Le Moal [Tue, 11 Jun 2024 08:34:35 +0000 (17:34 +0900)]
scsi: mpi3mr: Fix ATA NCQ priority support
The function mpi3mr_qcmd() of the mpi3mr driver is able to indicate to
the HBA if a read or write command directed at an ATA device should be
translated to an NCQ read/write command with the high prioiryt bit set
when the request uses the RT priority class and the user has enabled NCQ
priority through sysfs.
However, unlike the mpt3sas driver, the mpi3mr driver does not define
the sas_ncq_prio_supported and sas_ncq_prio_enable sysfs attributes, so
the ncq_prio_enable field of struct mpi3mr_sdev_priv_data is never
actually set and NCQ Priority cannot ever be used.
Fix this by defining these missing atributes to allow a user to check if
an ATA device supports NCQ priority and to enable/disable the use of NCQ
priority. To do this, lift the function scsih_ncq_prio_supp() out of the
mpt3sas driver and make it the generic SCSI SAS transport function
sas_ata_ncq_prio_supported(). Nothing in that function is hardware
specific, so this function can be used in both the mpt3sas driver and
the mpi3mr driver.
Reported-by: Scott McCoy <scott.mccoy@wdc.com> Fixes: 023ab2a9b4ed ("scsi: mpi3mr: Add support for queue command processing") Cc: stable@vger.kernel.org Signed-off-by: Damien Le Moal <dlemoal@kernel.org> Link: https://lore.kernel.org/r/20240611083435.92961-1-dlemoal@kernel.org Reviewed-by: Niklas Cassel <cassel@kernel.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Jeff Johnson [Mon, 10 Jun 2024 16:16:15 +0000 (09:16 -0700)]
scsi: Add missing MODULE_DESCRIPTION() macros
On x86, make allmodconfig && make W=1 C=1 reports:
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/scsi/scsi_common.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/scsi/advansys.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/scsi/BusLogic.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/scsi/aha1740.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/scsi/isci/isci.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/scsi/elx/efct.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/scsi/atp870u.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/scsi/ppa.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/scsi/imm.o
Add all missing invocations of the MODULE_DESCRIPTION() macro.
This updates all files which have a MODULE_LICENSE() but which do not have
a MODULE_DESCRIPTION(), even ones which did not produce the x86
allmodconfig warnings.
Ziqi Chen [Fri, 7 Jun 2024 10:06:23 +0000 (18:06 +0800)]
scsi: ufs: core: Quiesce request queues before checking pending cmds
In ufshcd_clock_scaling_prepare(), after SCSI layer is blocked,
ufshcd_pending_cmds() is called to check whether there are pending
transactions or not. And only if there are no pending transactions can we
proceed to kickstart the clock scaling sequence.
ufshcd_pending_cmds() traverses over all SCSI devices and calls
sbitmap_weight() on their budget_map. sbitmap_weight() can be broken down
to three steps:
1. Calculate the nr outstanding bits set in the 'word' bitmap.
2. Calculate the nr outstanding bits set in the 'cleared' bitmap.
3. Subtract the result from step 1 by the result from step 2.
This can lead to a race condition as outlined below:
Assume there is one pending transaction in the request queue of one SCSI
device, say sda, and the budget token of this request is 0, the 'word' is
0x1 and the 'cleared' is 0x0.
1. When step 1 executes, it gets the result as 1.
2. Before step 2 executes, block layer tries to dispatch a new request to
sda. Since the SCSI layer is blocked, the request cannot pass through
SCSI but the block layer would do budget_get() and budget_put() to
sda's budget map regardless, so the 'word' has become 0x3 and 'cleared'
has become 0x2 (assume the new request got budget token 1).
3. When step 2 executes, it gets the result as 1.
4. When step 3 executes, it gets the result as 0, meaning there is no
pending transactions, which is wrong.
Thread A Thread B
ufshcd_pending_cmds() __blk_mq_sched_dispatch_requests()
| |
sbitmap_weight(word) |
| scsi_mq_get_budget()
| |
| scsi_mq_put_budget()
| |
sbitmap_weight(cleared)
...
When this race condition happens, the clock scaling sequence is started
with transactions still in flight, leading to subsequent hibernate enter
failure, broken link, task abort and back to back error recovery.
Fix this race condition by quiescing the request queues before calling
ufshcd_pending_cmds() so that block layer won't touch the budget map when
ufshcd_pending_cmds() is working on it. In addition, remove the SCSI layer
blocking/unblocking to reduce redundancies and latencies.
Fixes: 8d077ede48c1 ("scsi: ufs: Optimize the command queueing code") Co-developed-by: Can Guo <quic_cang@quicinc.com> Signed-off-by: Can Guo <quic_cang@quicinc.com> Signed-off-by: Ziqi Chen <quic_ziqichen@quicinc.com> Link: https://lore.kernel.org/r/1717754818-39863-1-git-send-email-quic_ziqichen@quicinc.com Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Damien Le Moal [Fri, 7 Jun 2024 01:25:07 +0000 (10:25 +0900)]
scsi: core: Disable CDL by default
For SCSI devices supporting the Command Duration Limits feature set, the
user can enable/disable this feature use through the sysfs device attribute
"cdl_enable". This attribute modification triggers a call to
scsi_cdl_enable() to enable and disable the feature for ATA devices and set
the scsi device cdl_enable field to the user provided bool value. For SCSI
devices supporting CDL, the feature set is always enabled and
scsi_cdl_enable() is reduced to setting the cdl_enable field.
However, for ATA devices, a drive may spin-up with the CDL feature enabled
by default. But the SCSI device cdl_enable field is always initialized to
false (CDL disabled), regardless of the actual device CDL feature
state. For ATA devices managed by libata (or libsas), libata-core always
disables the CDL feature set when the device is attached, thus syncing the
state of the CDL feature on the device and of the SCSI device cdl_enable
field. However, for ATA devices connected to a SAS HBA, the CDL feature is
not disabled on scan for ATA devices that have this feature enabled by
default, leading to an inconsistent state of the feature on the device with
the SCSI device cdl_enable field.
Avoid this inconsistency by adding a call to scsi_cdl_enable() in
scsi_cdl_check() to make sure that the device-side state of the CDL feature
set always matches the scsi device cdl_enable field state. This implies
that CDL will always be disabled for ATA devices connected to SAS HBAs,
which is consistent with libata/libsas initialization of the device.
Reported-by: Scott McCoy <scott.mccoy@wdc.com> Fixes: 1b22cfb14142 ("scsi: core: Allow enabling and disabling command duration limits") Cc: stable@vger.kernel.org Signed-off-by: Damien Le Moal <dlemoal@kernel.org> Link: https://lore.kernel.org/r/20240607012507.111488-1-dlemoal@kernel.org Reviewed-by: Niklas Cassel <cassel@kernel.org> Reviewed-by: Igor Pylypiv <ipylypiv@google.com> Reviewed-by: Hannes Reinecke <hare@suse.de> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Breno Leitao [Wed, 5 Jun 2024 08:55:29 +0000 (01:55 -0700)]
scsi: mpt3sas: Avoid test/set_bit() operating in non-allocated memory
There is a potential out-of-bounds access when using test_bit() on a single
word. The test_bit() and set_bit() functions operate on long values, and
when testing or setting a single word, they can exceed the word
boundary. KASAN detects this issue and produces a dump:
BUG: KASAN: slab-out-of-bounds in _scsih_add_device.constprop.0 (./arch/x86/include/asm/bitops.h:60 ./include/asm-generic/bitops/instrumented-atomic.h:29 drivers/scsi/mpt3sas/mpt3sas_scsih.c:7331) mpt3sas
Write of size 8 at addr ffff8881d26e3c60 by task kworker/u1536:2/2965
For full log, please look at [1].
Make the allocation at least the size of sizeof(unsigned long) so that
set_bit() and test_bit() have sufficient room for read/write operations
without overwriting unallocated memory.
Martin K. Petersen [Wed, 5 Jun 2024 02:25:21 +0000 (22:25 -0400)]
scsi: sd: Use READ(16) when reading block zero on large capacity disks
Commit 321da3dc1f3c ("scsi: sd: usb_storage: uas: Access media prior
to querying device properties") triggered a read to LBA 0 before
attempting to inquire about device characteristics. This was done
because some protocol bridge devices will return generic values until
an attached storage device's media has been accessed.
Pierre Tomon reported that this change caused problems on a large
capacity external drive connected via a bridge device. The bridge in
question does not appear to implement the READ(10) command.
Issue a READ(16) instead of READ(10) when a device has been identified
as preferring 16-byte commands (use_16_for_rw heuristic).
Martin K. Petersen [Wed, 5 Jun 2024 02:22:11 +0000 (22:22 -0400)]
Merge patch series "Declare local functions static"
Bart Van Assche <bvanassche@acm.org> says:
Hi Martin,
There are several 32-bit ARM SCSI drivers that trigger compiler warnings
about missing function declarations. This patch series fixes these
compiler warnings by declaring local functions static. Please consider
this patch series for the next merge window.
Martin K. Petersen [Wed, 5 Jun 2024 02:16:22 +0000 (22:16 -0400)]
Merge patch series "ufs: pci: Add support UFSHCI 4.0 MCQ"
Minwoo Im <minwoo.im@samsung.com> says:
This patchset introduces add support for MCQ introduced in UFSHCI 4.0.
The first patch adds a simple helper to get the address of MCQ queue
config registers. The second one enables MCQ feature by adding mandatory
vops callback functions required at MCQ initialization phase. The last
one is to prevent a case where number of MCQ is given 1 since driver
allocates poll_queues first rather than I/O queues to handle device
commands. Instead of causing exception handlers due to no I/O queue,
failfast during the initialization time.
Minwoo Im [Fri, 31 May 2024 21:22:44 +0000 (06:22 +0900)]
scsi: ufs: mcq: Prevent no I/O queue case for MCQ
If hba_maxq equals poll_queues, which means there are no I/O queues
(HCTX_TYPE_DEFAULT, HCTX_TYPE_READ), the very first hw queue will be
allocated as HCTX_TYPE_POLL and it will be used as the dev_cmd_queue. In
this case, device commands such as QUERY cannot be properly handled.
This patch prevents the initialization of MCQ when the number of I/O queues
is not set and only the number of POLL queues is set.
Minwoo Im [Fri, 31 May 2024 21:22:43 +0000 (06:22 +0900)]
scsi: ufs: pci: Add support MCQ for QEMU-based UFS
Recently, ufs-mcq feature has been introduced to QEMU hw/ufs device [1].
This patch adds MCQ support for upstream QEMU UFS PCI controller. This
patch provides mandatory vops callbacks to make UFS controller work
properly on MCQ mode. Operation and Runtime Config register stride is
fixed to 48bytes which is implemented by qemu.
Minwoo Im [Sun, 19 May 2024 22:14:56 +0000 (07:14 +0900)]
scsi: ufs: mcq: Fix missing argument 'hba' in MCQ_OPR_OFFSET_n
The MCQ_OPR_OFFSET_n macro takes 'hba' in the caller context without
receiving 'hba' instance as an argument. To prevent potential bugs in
future use cases, add an argument 'hba'.
Fixes: 2468da61ea09 ("scsi: ufs: core: mcq: Configure operation and runtime interface") Cc: Asutosh Das <quic_asutoshd@quicinc.com> Signed-off-by: Minwoo Im <minwoo.im@samsung.com> Link: https://lore.kernel.org/r/20240519221457.772346-2-minwoo.im@samsung.com Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Chanwoo Lee [Fri, 24 May 2024 01:59:04 +0000 (10:59 +0900)]
scsi: ufs: mcq: Fix error output and clean up ufshcd_mcq_abort()
An error unrelated to ufshcd_try_to_abort_task is being logged and can
cause confusion. Modify ufshcd_mcq_abort() to print the result of the abort
failure. For readability, return immediately instead of 'goto'.
Fixes: f1304d442077 ("scsi: ufs: mcq: Added ufshcd_mcq_abort()") Signed-off-by: Chanwoo Lee <cw9316.lee@samsung.com> Link: https://lore.kernel.org/r/20240524015904.1116005-1-cw9316.lee@samsung.com Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Martin K. Petersen [Fri, 31 May 2024 00:36:08 +0000 (20:36 -0400)]
Merge patch series "scsi: ufs: Allow RTT negotiation"
Avri Altman <avri.altman@wdc.com> says:
The rtt-upiu packets precede any data-out upiu packets, thus synchronizing
the data input to the device: this mostly applies to write operations, but
there are other operations that requires rtt as well.
There are several rules binding this rtt - data-out dialog, specifically
There can be at most outstanding bMaxNumOfRTT such packets. This might
have an effect on write performance (sequential write in particular), as
each data-out upiu must wait for its rtt sibling.
UFSHCI expects bMaxNumOfRTT to be min(bDeviceRTTCap, NORTT). However, as of
today, there does not appear to be no-one who sets it: not the host
controller nor the driver. It wasn't an issue up to now: bMaxNumOfRTT is
set to 2 after manufacturing, and wasn't limiting the write performance.
UFS4.0, and specifically gear 5 changes this, and requires the device to be
more attentive. This doesn't come free - the device has to allocate more
resources to that end, but the sequential write performance improvement is
significant. Early measurements shows 25% gain when moving from rtt 2 to
9. Therefore, set bMaxNumOfRTT to be min(bDeviceRTTCap, NORTT) as UFSHCI
expects.
Avri Altman [Thu, 30 May 2024 14:25:09 +0000 (17:25 +0300)]
scsi: ufs: sysfs: Make max_number_of_rtt read-write
Given the importance of the RTT parameter, we want to be able to configure
it via sysfs. This is because UFS users should be discouraged from change
UFS device parameters without the UFSHCI driver being aware of these
changes.
Avri Altman [Thu, 30 May 2024 14:25:08 +0000 (17:25 +0300)]
scsi: ufs: core: Maximum RTT supported by the host driver
Allow platform vendors to take precedence having their own max rtt support.
This makes sense because the host controller's nortt characteristic may
vary among vendors.
Signed-off-by: Avri Altman <avri.altman@wdc.com> Link: https://lore.kernel.org/r/20240530142510.734-3-avri.altman@wdc.com Reviewed-by: Peter Wang <peter.wang@mediatek.com> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Avri Altman [Thu, 30 May 2024 14:25:07 +0000 (17:25 +0300)]
scsi: ufs: core: Allow RTT negotiation
The rtt-upiu packets precede any data-out upiu packets, thus synchronizing
the data input to the device: this mostly applies to write operations, but
there are other operations that requires rtt as well.
There are several rules binding this rtt - data-out dialog, specifically
There can be at most outstanding bMaxNumOfRTT such packets. This might
have an effect on write performance (sequential write in particular), as
each data-out upiu must wait for its rtt sibling.
UFSHCI expects bMaxNumOfRTT to be min(bDeviceRTTCap, NORTT). However, as of
today, there does not appears to be no-one who sets it: not the host
controller nor the driver. It wasn't an issue up to now: bMaxNumOfRTT is
set to 2 after manufacturing, and wasn't limiting the write performance.
UFS4.0, and specifically gear 5 changes this, and requires the device to be
more attentive. This doesn't come free - the device has to allocate more
resources to that end, but the sequential write performance improvement is
significant. Early measurements shows 25% gain when moving from rtt 2 to
9. Therefore, set bMaxNumOfRTT to be min(bDeviceRTTCap, NORTT) as UFSHCI
expects.
Signed-off-by: Avri Altman <avri.altman@wdc.com> Link: https://lore.kernel.org/r/20240530142510.734-2-avri.altman@wdc.com Reviewed-by: Bean Huo <beanhuo@micron.com> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
John Garry [Fri, 24 May 2024 08:48:29 +0000 (08:48 +0000)]
scsi: bsg: Pass dev to blk_mq_alloc_queue()
When calling bsg_setup_queue() -> blk_mq_alloc_queue(), we don't pass
the dev as the queuedata, but rather manually set it afterwards. Just
pass dev to blk_mq_alloc_queue() to have automatically set.
Signed-off-by: John Garry <john.g.garry@oracle.com> Link: https://lore.kernel.org/r/20240524084829.2132555-3-john.g.garry@oracle.com Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Hannes Reinecke <hare@suse.de> Tested-by: Himanshu Madhani <himanshu.madhani@oracle.com> Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
projects / linux.git / log