Task Management to virtual drives may timeout prematurely when using a
static default timeout.
Read Abort and Reset timeouts from Device Page 0 and apply the maximum
of the firmware value and the default.
This fixes premature TM failures on virtual drives.
Signed-off-by: Chandrakanth Patil <chandrakanth.patil@broadcom.com>
Link: https://lore.kernel.org/r/20250820084138.228471-6-chandrakanth.patil@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
On 32-bit systems, 64-bit BAR writes to admin queue registers are
performed as two 32-bit writes. Without locking, this can cause partial
writes when accessed concurrently.
Updated per-queue spinlocks is used to serialize these writes and prevent
race conditions.
Fixes: 824a156633 ("scsi: mpi3mr: Base driver code")
Cc: stable@vger.kernel.org
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Link: https://lore.kernel.org/r/20250627194539.48851-4-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The volatile qualifier is redundant for __iomem pointers.
Cleaned up usage in mpi3mr_writeq() and sysif_regs pointer as per
Upstream compliance.
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Link: https://lore.kernel.org/r/20250627194539.48851-3-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Admin reply processing can be called from multiple contexts. The driver
uses an atomic flag for synchronization among multiple threads/context for
draining the admin replies.
Upon entering the admin processing routine, the driver will set the atomic
flag and start reply processing. When exiting the routine, the driver
resets the flag. However, there is a race condition when one thread (Thread
1) has processed replies and is about to reset the flag but in the meantime
few more replies are posted and another thread (Thread 2) is called to
process replies. Since the synchronization flag is still set, Thread 2 will
return without processing replies and those new replies will not be
flushed.
Make the watchdog thread monitor cases where admin ISR/poll call returns
due to another thread processing admin replies. If such an instance is
found, make driver call admin ISR to drain replies (if any).
Co-developed-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Link: https://lore.kernel.org/r/20250220142528.20837-4-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
When the task management thread processes reply queues while the reset
thread resets them, the task management thread accesses an invalid queue ID
(0xFFFF), set by the reset thread, which points to unallocated memory,
causing a crash.
Add flag 'io_admin_reset_sync' to synchronize access between the reset,
I/O, and admin threads. Before a reset, the reset handler sets this flag to
block I/O and admin processing threads. If any thread bypasses the initial
check, the reset thread waits up to 10 seconds for processing to finish. If
the wait exceeds 10 seconds, the controller is marked as unrecoverable.
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Link: https://lore.kernel.org/r/20250129100850.25430-4-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Allocate segmented trace buffer if firmware advertises the capability in
IOCfacts.
Upon driver load, read the trace buffer size from driver page 1, calculate
the required segments for trace buffer, and allocate segmented buffers.
Each segment is 4096 bytes in size.
While posting driver diagnostic buffer to firmware, advertise that trace
buffer is segmented.
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Link: https://lore.kernel.org/r/20250129100850.25430-3-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
To avoid reply queue full condition, update the driver to check IOCFacts
capabilities for qfull.
Update the operational reply queue's Consumer Index after processing 100
replies. If pending I/Os on a reply queue exceeds a threshold
(reply_queue_depth - 200), then return I/O back to OS to retry.
Also increase default admin reply queue size to 2K.
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Link: https://lore.kernel.org/r/20250129100850.25430-2-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Replace all users of blk_mq_pci_map_queues with the more generic
blk_mq_map_hw_queues. This in preparation to retire
blk_mq_pci_map_queues.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: John Garry <john.g.garry@oracle.com>
Signed-off-by: Daniel Wagner <wagi@kernel.org>
Link: https://lore.kernel.org/r/20241202-refactor-blk-affinity-helpers-v6-5-27211e9c2cd5@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The driver, through the SAS transport, exposes a sysfs interface to
enable/disable PHYs in a controller/expander setup. When multiple PHYs
are disabled and enabled in rapid succession, the persistent and current
config pages related to SAS IO unit/SAS Expander pages could get
corrupted.
Use separate memory for each config request.
Signed-off-by: Prayas Patel <prayas.patel@broadcom.com>
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Link: https://lore.kernel.org/r/20241110194405.10108-3-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Fixes all in drivers. The largest is the mpi3mr which corrects a phy
count limit that should only apply to the controller but was being
incorrectly applied to expander phys.
Signed-off-by: James E.J. Bottomley <James.Bottomley@HansenPartnership.com>
-----BEGIN PGP SIGNATURE-----
iJwEABMIAEQWIQTnYEDbdso9F2cI+arnQslM7pishQUCZxQINSYcamFtZXMuYm90
dG9tbGV5QGhhbnNlbnBhcnRuZXJzaGlwLmNvbQAKCRDnQslM7pishfuiAP9+gxzU
MYpyNFYmS8iwbma+wagKMr2ZxpiIAnQIukBtugD7Baqw7YjNbAPtCcQMgC96y8ac
X7zeMkLTyMqkaIjOQFk=
=eHKZ
-----END PGP SIGNATURE-----
Merge tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Pull SCSI fixes from James Bottomley:
"Fixes all in drivers. The largest is the mpi3mr which corrects a phy
count limit that should only apply to the controller but was being
incorrectly applied to expander phys"
* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
scsi: target: core: Fix null-ptr-deref in target_alloc_device()
scsi: mpi3mr: Validate SAS port assignments
scsi: ufs: core: Set SDEV_OFFLINE when UFS is shut down
scsi: ufs: core: Requeue aborted request
scsi: ufs: core: Fix the issue of ICU failure
A sanity check on phy_mask was added in commit 3668651def ("scsi:
mpi3mr: Sanitise num_phys"). This causes warning messages when more than
64 phys are detected and devices connected to phys greater than 64 are
dropped.
The phy_mask bitmap is only needed for controller phys and not required
for expander phys. Controller phys can go up to a maximum of 64 and
therefore u64 is good enough to contain phy_mask bitmap.
To suppress those warnings and allow devices to be discovered as before
the offending commit, restrict the phy_mask setting and lowest phy
setting only to the controller phys.
Fixes: 3668651def ("scsi: mpi3mr: Sanitise num_phys")
Cc: stable@vger.kernel.org
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202410051943.Mp9o5DlF-lkp@intel.com/
Reported-by: Alexander Motin <mav@ixsystems.com>
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Link: https://lore.kernel.org/r/20241008074353.200379-1-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
asm/unaligned.h is always an include of asm-generic/unaligned.h;
might as well move that thing to linux/unaligned.h and include
that - there's nothing arch-specific in that header.
auto-generated by the following:
for i in `git grep -l -w asm/unaligned.h`; do
sed -i -e "s/asm\/unaligned.h/linux\/unaligned.h/" $i
done
for i in `git grep -l -w asm-generic/unaligned.h`; do
sed -i -e "s/asm-generic\/unaligned.h/linux\/unaligned.h/" $i
done
git mv include/asm-generic/unaligned.h include/linux/unaligned.h
git mv tools/include/asm-generic/unaligned.h tools/include/linux/unaligned.h
sed -i -e "/unaligned.h/d" include/asm-generic/Kbuild
sed -i -e "s/__ASM_GENERIC/__LINUX/" include/linux/unaligned.h tools/include/linux/unaligned.h
Make driver use the timestamp update interval value provided by firmware in
the driver page 1. If firmware fails to provide non-zero value, then the
driver will fall back to the driver defined macro.
Signed-off-by: Prayas Patel <prayas.patel@broadcom.com>
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Link: https://lore.kernel.org/r/20240905102753.105310-3-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Commit 0c52310f26 ("hrtimer: Ignore slack time for RT tasks in
schedule_hrtimeout_range()") effectivelly shortens a sleep in a polling
function in the driver. That is causing a performance regression as the
new value of just 2us is too low, in certain tests the perf drop is ~30%.
Fix this by adjusting the sleep to 20us (close to the previous value).
Reported-by: Jan Jurca <jjurca@redhat.com>
Signed-off-by: Tomas Henzl <thenzl@redhat.com>
Acked-by: Sumit Saxena <sumit.saxena@broadcom.com>
Link: https://lore.kernel.org/r/20240903144729.37218-1-thenzl@redhat.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Bart Van Assche <bvanassche@acm.org> says:
Hi Martin,
Multiple SCSI drivers use snprintf() to format a workqueue name before
invoking one of the create*_workqueue() macros. This patch series
simplifies such code by passing the format string and arguments to
alloc_workqueue(). Additionally, the structure members that are only
used as a temporary buffer for formatting workqueue names are
removed. Please consider this patch series for the next merge window.
Thanks,
Bart.
Link: https://lore.kernel.org/r/20240822195944.654691-1-bvanassche@acm.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Let alloc_ordered_workqueue() format the workqueue name instead of calling
snprintf() explicitly.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240822195944.654691-9-bvanassche@acm.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Instead of updating the ConsumerIndex of the Admin and Operational
ReplyQueues after processing all replies in the queue, the index will now
be periodically updated after processing every 100 replies.
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/20240808125418.8832-3-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
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.
Co-developed-by: Sathya Prakash <sathya.prakash@broadcom.com>
Signed-off-by: Sathya Prakash <sathya.prakash@broadcom.com>
Co-developed-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Link: https://lore.kernel.org/r/20240627101735.18286-2-sumit.saxena@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
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>
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>
Switch to the ->device_configure method instead of ->slave_configure and
update the block limits on the passed in queue_limits instead of using the
per-limit accessors.
Note that mpi3mr also updates the limits from an event handler that
iterates all SCSI devices. This is also updated to use the queue_limits,
but the complete locking of this path probably means it already is
completely broken and needs a proper audit.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20240409143748.980206-22-hch@lst.de
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The driver did not remove the virtual disk that was exposed as hidden and
offline after the controller was reset.
Drive is removed from OS when firmware sends "device added" event with
hidden bit set or access status indicating inability to accept I/Os.
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Signed-off-by: Sathya Prakash <sathya.prakash@broadcom.com>
Link: https://lore.kernel.org/r/20240313100746.128951-2-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The driver acquires the required NVMe SGLs from the pre-allocated
pool.
Co-developed-by: Sathya Prakash <sathya.prakash@broadcom.com>
Signed-off-by: Sathya Prakash <sathya.prakash@broadcom.com>
Signed-off-by: Chandrakanth patil <chandrakanth.patil@broadcom.com>
Link: https://lore.kernel.org/r/20231205191630.12201-4-chandrakanth.patil@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The driver acquires the required SGLs from the pre-allocated pool.
Co-developed-by: Sathya Prakash <sathya.prakash@broadcom.com>
Signed-off-by: Sathya Prakash <sathya.prakash@broadcom.com>
Signed-off-by: Chandrakanth patil <chandrakanth.patil@broadcom.com>
Link: https://lore.kernel.org/r/20231205191630.12201-3-chandrakanth.patil@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The driver now supports SGLs for BSG data transfer. Upon loading, the
driver pre-allocates SGLs in chunks of 8k, results in a total of 256 * 8k,
equal to 2MB. These pre-allocated SGLs are reserved for handling BSG
commands and are deallocated during driver unload.
Co-developed-by: Sathya Prakash <sathya.prakash@broadcom.com>
Signed-off-by: Sathya Prakash <sathya.prakash@broadcom.com>
Signed-off-by: Chandrakanth patil <chandrakanth.patil@broadcom.com>
Link: https://lore.kernel.org/r/20231205191630.12201-2-chandrakanth.patil@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
SAS5116 controllers supports maximum 48 physical PHYs. Modify driver to
accommodate up to 64 PHYs (though current need is to support 48 PHYs).
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Link: https://lore.kernel.org/r/20231123160132.4155-4-sumit.saxena@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Enhance the driver to get the maximum data length per I/O request from IOC
Facts data and report that to the upper layers. If the IOC facts data is
not reported then a default I/O size of 1MB is reported to the OS.
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Link: https://lore.kernel.org/r/20230804104248.118924-4-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Update driver version to 8.4.1.0.0.
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Signed-off-by: Sreekanth Reddy <sreekanth.reddy@broadcom.com>
Link: https://lore.kernel.org/r/20230316110209.60145-9-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Update copyright year from 2022 to 2023.
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Signed-off-by: Sreekanth Reddy <sreekanth.reddy@broadcom.com>
Link: https://lore.kernel.org/r/20230316110209.60145-8-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Modify Message Unit Reset timeout value to 120 seconds from the previous
value of 30 seconds.
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Signed-off-by: Sreekant Reddy <sreekanth.reddy@broadcom.com>
Link: https://lore.kernel.org/r/20230316110209.60145-4-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
