Driver initiates OCR if a DCMD command times out. But there is a deadlock
if the driver attempts to invoke another OCR before the mutex lock
(reset_mutex) is released from the previous session of OCR.
This patch takes care of the above scenario using new flag
MEGASAS_FUSION_OCR_NOT_POSSIBLE to indicate if OCR is possible.
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/1579000882-20246-9-git-send-email-anand.lodnoor@broadcom.com
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Anand Lodnoor <anand.lodnoor@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Disable WRITE_SAME (no_write_same) for Virtual Disks only. For System PDs
and EPDs (Enhanced PDs), WRITE_SAME need not be disabled by default.
Link: https://lore.kernel.org/r/1579000882-20246-3-git-send-email-anand.lodnoor@broadcom.com
Signed-off-by: Anand Lodnoor <anand.lodnoor@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
This is mostly update of the usual drivers: qla2xxx, hpsa, lpfc, ufs,
mpt3sas, ibmvscsi, megaraid_sas, bnx2fc and hisi_sas as well as the
removal of the osst driver (I heard from Willem privately that he
would like the driver removed because all his test hardware has
failed). Plus number of minor changes, spelling fixes and other
trivia.
Signed-off-by: James E.J. Bottomley <jejb@linux.ibm.com>
-----BEGIN PGP SIGNATURE-----
iJwEABMIAEQWIQTnYEDbdso9F2cI+arnQslM7pishQUCXSTl4yYcamFtZXMuYm90
dG9tbGV5QGhhbnNlbnBhcnRuZXJzaGlwLmNvbQAKCRDnQslM7pishdcxAQDCJVbd
fPUX76/V1ldupunF97+3DTharxxbst+VnkOnCwD8D4c0KFFFOI9+F36cnMGCPegE
fjy17dQLvsJ4GsidHy8=
=aS5B
-----END PGP SIGNATURE-----
Merge tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Pull SCSI updates from James Bottomley:
"This is mostly update of the usual drivers: qla2xxx, hpsa, lpfc, ufs,
mpt3sas, ibmvscsi, megaraid_sas, bnx2fc and hisi_sas as well as the
removal of the osst driver (I heard from Willem privately that he
would like the driver removed because all his test hardware has
failed). Plus number of minor changes, spelling fixes and other
trivia.
The big merge conflict this time around is the SPDX licence tags.
Following discussion on linux-next, we believe our version to be more
accurate than the one in the tree, so the resolution is to take our
version for all the SPDX conflicts"
Note on the SPDX license tag conversion conflicts: the SCSI tree had
done its own SPDX conversion, which in some cases conflicted with the
treewide ones done by Thomas & co.
In almost all cases, the conflicts were purely syntactic: the SCSI tree
used the old-style SPDX tags ("GPL-2.0" and "GPL-2.0+") while the
treewide conversion had used the new-style ones ("GPL-2.0-only" and
"GPL-2.0-or-later").
In these cases I picked the new-style one.
In a few cases, the SPDX conversion was actually different, though. As
explained by James above, and in more detail in a pre-pull-request
thread:
"The other problem is actually substantive: In the libsas code Luben
Tuikov originally specified gpl 2.0 only by dint of stating:
* This file is licensed under GPLv2.
In all the libsas files, but then muddied the water by quoting GPLv2
verbatim (which includes the or later than language). So for these
files Christoph did the conversion to v2 only SPDX tags and Thomas
converted to v2 or later tags"
So in those cases, where the spdx tag substantially mattered, I took the
SCSI tree conversion of it, but then also took the opportunity to turn
the old-style "GPL-2.0" into a new-style "GPL-2.0-only" tag.
Similarly, when there were whitespace differences or other differences
to the comments around the copyright notices, I took the version from
the SCSI tree as being the more specific conversion.
Finally, in the spdx conversions that had no conflicts (because the
treewide ones hadn't been done for those files), I just took the SCSI
tree version as-is, even if it was old-style. The old-style conversions
are perfectly valid, even if the "-only" and "-or-later" versions are
perhaps more descriptive.
* tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi: (185 commits)
scsi: qla2xxx: move IO flush to the front of NVME rport unregistration
scsi: qla2xxx: Fix NVME cmd and LS cmd timeout race condition
scsi: qla2xxx: on session delete, return nvme cmd
scsi: qla2xxx: Fix kernel crash after disconnecting NVMe devices
scsi: megaraid_sas: Update driver version to 07.710.06.00-rc1
scsi: megaraid_sas: Introduce various Aero performance modes
scsi: megaraid_sas: Use high IOPS queues based on IO workload
scsi: megaraid_sas: Set affinity for high IOPS reply queues
scsi: megaraid_sas: Enable coalescing for high IOPS queues
scsi: megaraid_sas: Add support for High IOPS queues
scsi: megaraid_sas: Add support for MPI toolbox commands
scsi: megaraid_sas: Offload Aero RAID5/6 division calculations to driver
scsi: megaraid_sas: RAID1 PCI bandwidth limit algorithm is applicable for only Ventura
scsi: megaraid_sas: megaraid_sas: Add check for count returned by HOST_DEVICE_LIST DCMD
scsi: megaraid_sas: Handle sequence JBOD map failure at driver level
scsi: megaraid_sas: Don't send FPIO to RL Bypass queue
scsi: megaraid_sas: In probe context, retry IOC INIT once if firmware is in fault
scsi: megaraid_sas: Release Mutex lock before OCR in case of DCMD timeout
scsi: megaraid_sas: Call disable_irq from process IRQ poll
scsi: megaraid_sas: Remove few debug counters from IO path
...
The driver will use round-robin method for IO submission in batches within
the high IOPS queues when the number of in-flight ios on the target device
is larger than 8. Otherwise the driver will use low latency reply queues.
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Chandrakanth Patil <chandrakanth.patil@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
For RAID5/RAID6 volumes configured behind Aero, driver will be doing 64bit
division operations on behalf of firmware as controller's ARM CPU is very
slow in this division. Later, driver calculates Q-ARM, P-ARM and Log-ARM and
passes those values to firmware by writing these values to RAID_CONTEXT.
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Chandrakanth Patil <chandrakanth.patil@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
RAID1 PCI bandwidth limit algorithm is not applicable to Aero as it's PCIe
Gen4 adapter.
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Chandrakanth Patil <chandrakanth.patil@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Create a debugfs interface for megaraid_sas driver. Provide interface to
dump driver RAID map in debugfs.
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Issue Description:
We have seen cpu lock up issues from field if system has a large (more than
96) logical cpu count. SAS3.0 controller (Invader series) supports max 96
MSI-X vector and SAS3.5 product (Ventura) supports max 128 MSI-X vectors.
This may be a generic issue (if PCI device support completion on multiple
reply queues).
Let me explain it w.r.t megaraid_sas supported h/w just to simplify the
problem and possible changes to handle such issues. MegaRAID controller
supports multiple reply queues in completion path. Driver creates MSI-X
vectors for controller as "minimum of (FW supported Reply queues, Logical
CPUs)". If submitter is not interrupted via completion on same CPU, there
is a loop in the IO path. This behavior can cause hard/soft CPU lockups, IO
timeout, system sluggish etc.
Example - one CPU (e.g. CPU A) is busy submitting the IOs and another CPU
(e.g. CPU B) is busy with processing the corresponding IO's reply
descriptors from reply descriptor queue upon receiving the interrupts from
HBA. If CPU A is continuously pumping the IOs then always CPU B (which is
executing the ISR) will see the valid reply descriptors in the reply
descriptor queue and it will be continuously processing those reply
descriptor in a loop without quitting the ISR handler.
megaraid_sas driver will exit ISR handler if it finds unused reply
descriptor in the reply descriptor queue. Since CPU A will be continuously
sending the IOs, CPU B may always see a valid reply descriptor (posted by
HBA Firmware after processing the IO) in the reply descriptor queue. In
worst case, driver will not quit from this loop in the ISR handler.
Eventually, CPU lockup will be detected by watchdog.
Above mentioned behavior is not common if "rq_affinity" set to 2 or
affinity_hint is honored by irqbalancer as "exact". If rq_affinity is set
to 2, submitter will be always interrupted via completion on same CPU. If
irqbalancer is using "exact" policy, interrupt will be delivered to
submitter CPU.
Problem statement:
If CPU count to MSI-X vectors (reply descriptor Queues) count ratio is not
1:1, we still have exposure of issue explained above and for that we don't
have any solution.
Exposure of soft/hard lockup is seen if CPU count is more than MSI-X
supported by device.
If CPUs count to MSI-X vectors count ratio is not 1:1, (Other way, if
CPU counts to MSI-X vector count ratio is something like X:1, where X > 1)
then 'exact' irqbalance policy OR rq_affinity = 2 won't help to avoid CPU
hard/soft lockups. There won't be any one to one mapping between
CPU to MSI-X vector instead one MSI-X interrupt (or reply descriptor queue)
is shared with group/set of CPUs and there is a possibility of having a
loop in the IO path within that CPU group and may observe lockups.
For example: Consider a system having two NUMA nodes and each node having
four logical CPUs and also consider that number of MSI-X vectors enabled on
the HBA is two, then CPUs count to MSI-X vector count ratio as 4:1.
e.g.
MSI-X vector 0 is affinity to CPU 0, CPU 1, CPU 2 & CPU 3 of NUMA node 0 and
MSI-X vector 1 is affinity to CPU 4, CPU 5, CPU 6 & CPU 7 of NUMA node 1.
numactl --hardware
available: 2 nodes (0-1)
node 0 cpus: 0 1 2 3 --> MSI-X 0
node 0 size: 65536 MB
node 0 free: 63176 MB
node 1 cpus: 4 5 6 7 --> MSI-X 1
node 1 size: 65536 MB
node 1 free: 63176 MB
Assume that user started an application which uses all the CPUs of NUMA
node 0 for issuing the IOs. Only one CPU from affinity list (it can be any
cpu since this behavior depends upon irqbalance) CPU0 will receive the
interrupts from MSI-X 0 for all the IOs. Eventually, CPU 0 IO submission
percentage will be decreasing and ISR processing percentage will be
increasing as it is more busy with processing the interrupts. Gradually IO
submission percentage on CPU 0 will be zero and it's ISR processing
percentage will be 100% as IO loop has already formed within the
NUMA node 0, i.e. CPU 1, CPU 2 & CPU 3 will be continuously busy with
submitting the heavy IOs and only CPU 0 is busy in the ISR path as it
always find the valid reply descriptor in the reply descriptor queue.
Eventually, we will observe the hard lockup here.
Chances of occurring of hard/soft lockups are directly proportional to
value of X. If value of X is high, then chances of observing CPU lockups is
high.
Solution:
Use IRQ poll interface defined in "irq_poll.c".
megaraid_sas driver will execute ISR routine in softirq context and it will
always quit the loop based on budget provided in IRQ poll interface.
Driver will switch to IRQ poll only when more than a threshold number of
reply descriptors are handled in one ISR. Currently threshold is set as
1/4th of HBA queue depth.
In these scenarios (i.e. where CPUs count to MSI-X vectors count ratio is
X:1 (where X > 1)), IRQ poll interface will avoid CPU hard lockups due to
voluntary exit from the reply queue processing based on budget.
Note - Only one MSI-X vector is busy doing processing.
Select CONFIG_IRQ_POLL from driver Kconfig for driver compilation.
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Based on 2 normalized pattern(s):
this program is free software you can redistribute it and or modify
it under the terms of the gnu general public license as published by
the free software foundation either version 2 of the license or at
your option any later version this program is distributed in the
hope that it will be useful but without any warranty without even
the implied warranty of merchantability or fitness for a particular
purpose see the gnu general public license for more details you
should have received a copy of the gnu general public license along
with this program if not see http www gnu org licenses
this program is free software you can redistribute it and or modify
it under the terms of the gnu general public license as published by
the free software foundation either version 2 of the license or at
your option any later version this program is distributed in the
hope that it will be useful but without any warranty without even
the implied warranty of merchantability or fitness for a particular
purpose see the gnu general public license for more details [based]
[from] [clk] [highbank] [c] you should have received a copy of the
gnu general public license along with this program if not see http
www gnu org licenses
extracted by the scancode license scanner the SPDX license identifier
GPL-2.0-or-later
has been chosen to replace the boilerplate/reference in 355 file(s).
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Jilayne Lovejoy <opensource@jilayne.com>
Reviewed-by: Steve Winslow <swinslow@gmail.com>
Reviewed-by: Allison Randal <allison@lohutok.net>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190519154041.837383322@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch adds support for the new DEVICE_LIST DCMD.
Driver currently sends two separate DCMDs for getting the list of PDs and
LDs that are exposed to host. The new DCMD provides a single interface to
get a list of both PDs and LDs that are exposed to the host. Based on the
list of target IDs that are returned by this DCMD, driver will add the
devices (PD/LD) to SML. Driver will check for FW support for this new DCMD
and based on the support will either send the new DCMD or will fall back to
the earlier method of sending two separate DCMDs for PD and LD list.
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Change copyright to Broadcom Inc. Also update any references to Avago with
Broadcom. Update copyright duration wherever required.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Latest firmware adds a mechanism to save firmware logs just before
controller reset on pre-allocated internal controller DRAM. This feature is
called snapdump which will help debugging firmware issues. This feature
requires extra time and firmware reports these values through new driver
interface. Before initiating an OCR, driver needs to inform FW to save a
snapdump and then wait for a specified time for the snapdump to complete.
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
No functional changes.
This patch is a re-work of DCMD refire code to better manage all the
different cases to decide whether to REFIRE or SKIP or COMPLETE certain
DCMD.
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The latest MegaRAID Firmware (for Invader series) has support for 64bit
DMA for both streaming and consistent DMA buffers. All Ventura series
controller FW always support 64 bit consistent DMA. Also, on a few
architectures 32bit DMA is not supported.
Current driver always prefers 32bit for consistent DMA and 64bit for
streaming DMA. This behavior was unintentional and carried forwarded
from legacy controller FW. Need to enhance the driver to support 64bit
consistent DMA buffers based on the firmware capability.
Below is the DMA setting strategy in driver with this patch. For
Ventura series, always try to set 64bit DMA mask. If it fails fall back
to 32bit DMA mask. For Invader series and earlier generation
controllers, first try to set to 32bit consistent DMA mask irrespective
of FW capability. This is needed to ensure firmware downgrades do not
break. If 32bit DMA setting fails, check FW capability and try seting to
64bit DMA mask.
There are certain restrictions in the hardware for having all sense
buffers and all reply descriptors to be in the same 4GB memory region.
This limitation is h/w dependent and can not be changed in firmware.
This limitation needs to be taken care in driver while allocating the
buffers. There was a discussion regarding this - find details at below
link. https://www.spinics.net/lists/linux-scsi/msg108251.html
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
In certain cases, the host memory is limited and with FW supporting
higher queue depths there are increasing chances of IO request frame
allocation failures that we are seeing. In case of request frame
allocation failures, retry allocation with reduced queue depth (in steps
of 64) to continue to configure the controller with a reduced
performance rather than failing load.
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Older firmware version unconditionally pulls 4k frame for IOC INIT MFA
frame. But driver allocates 1k or 4k max_chain_frame_sz based on FW
capability. During boot time, this results in DMA read errors.
Workaround fix in driver by allocating separate ioc_init frame of 4k
size to support older firmware.
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Cc: stable@vger.kernel.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Pre-allocate few of the frequently used DMA buffers during load time.
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
fusion_context structure is very large around 180kB and most of the size
is contributed by log_to_span array. Move log_to_span out of fusion
context and have separate allocation for log_to_span. And use kmalloc to
allocate fusion_context. Currently kmemleak reports 1000s of false
positives for fusion->cmd_list[]. kmemleak does not track page
allocation for fusion_context. This change will also fix the false
positives reported by kmemleak.
Ref: https://marc.info/?l=linux-scsi&m=150545293900917
Reported-by: Shu Wang <shuwang@redhat.com>
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
No functional change.
Refactor adapter_type to set for all generation controllers, not
just for fusion controllers.
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
For RAID1 FastPath writes, driver needs to allocate extra commands
internally to accommodate for the extra peer command being sent.
Currently driver is allocating 2 extra commands for each but only one
extra command is necessary. Set RAID_1_10_RMW_CMDS to 2 and also change
macro name to RAID_1_PEER_CMDS.
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Fix indentation issues and smatch warning reported by Dan Carpenter
for previous series as discussed below.
http://www.spinics.net/lists/linux-scsi/msg103635.htmlhttp://www.spinics.net/lists/linux-scsi/msg103603.html
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Sasikumar Chandrasekaran <sasikumar.pc@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Fix endiannes fixes for Ventura specific.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Currently fusion context has fixed array load_balance_info. Use dynamic
allocation. In few places, driver do not want physically contigious
memory. Attempt to use vmalloc if physical contiguous memory is not
available.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
This patch provide true fast path IO support. Driver creates PRP for
NVME drives and send Fast Path for performance. Certain h/w requirement
needs to be taken care in driver.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
This patch fetch true values of NVME property from FW using New DCMD
interface MR_DCMD_DEV_GET_TARGET_PROP
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Adding detection logic for NVME device attached behind Ventura
controller. Driver set HostPageSize in IOC_INIT frame to inform about
page size for NVME devices. Firmware reports NVME page size to the
driver. PD INFO DCMD provide new interface type NVME_PD. Driver set
property of NVME device.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
No functional change. Code refactor.
Remove function megasas_fpio_to_ldio as we never require to convert fpio
to ldio because of frame unavailability. Grab extra frame of raid 1
write fast path before it creates first frame as Fast Path. Removed
is_raid_1_fp_write flag as raid 1 write fast path command is decided
using r1_alt_dev_handle only. Move resetting megasas_cmd_fusion fields
at common function megasas_return_cmd_fusion.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
This reverts commit "3e5eadb1a881" ("scsi: megaraid_sas: Enable or
Disable Fast path based on the PCI Threshold Bandwidth")
This patch was aimed to increase performance of R1 Write operation for
large IO size. Since this method used timer approach, it turn on/off
fast path did not work as expected. Patch 0013 describes new algorithm
and performance number.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Update Linux driver to use new pdTargetId field for JBOD target ID
Signed-off-by: Sasikumar Chandrasekaran <sasikumar.pc@broadcom.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Large SEQ IO workload should sent as non fast path commands
Signed-off-by: Sasikumar Chandrasekaran <sasikumar.pc@broadcom.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The Megaraid driver has to support the SAS3.5 Generic Megaraid Controllers Firmware functionality.
Signed-off-by: Sasikumar Chandrasekaran <sasikumar.pc@broadcom.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
SAS3.5 Generic Megaraid Controllers FW will support new dynamic RaidMap to have different
sizes for different number of supported VDs.
Signed-off-by: Sasikumar Chandrasekaran <sasikumar.pc@broadcom.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
To improve RAID 1/10 Write performance, OS drivers need to issue the
required Write IOs as Fast Path IOs (after the appropriate checks
allowing Fast Path to be used) to the appropriate physical drives
(translated from the OS logical IO) and wait for all Write IOs to complete.
Design: A write IO on RAID volume will be examined if it can be sent in
Fast Path based on IO size and starting LBA and ending LBA falling on to
a Physical Drive boundary. If the underlying RAID volume is a RAID 1/10,
driver issues two fast path write IOs one for each corresponding physical
drive after computing the corresponding start LBA for each physical drive.
Both write IOs will have the same payload and are posted to HW such that
replies land in the same reply queue.
If there are no resources available for sending two IOs, driver will send
the original IO from SCSI layer to RAID volume through the Firmware.
Based on PCI bandwidth and write payload, every second this feature is
enabled/disabled.
When both IOs are completed by HW, the resources will be released
and SCSI IO completion handler will be called.
Signed-off-by: Sasikumar Chandrasekaran <sasikumar.pc@broadcom.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Detect sequential Write IOs and pass the hint that it is part of sequential
stream to help HBA Firmware do the Full Stripe Writes. For read IOs on
certain RAID volumes like Read Ahead volumes,this will help driver to
send it to Firmware even if the IOs can potentially be sent to
hardware directly (called fast path) bypassing firmware.
Design: 8 streams are maintained per RAID volume as per the combined
firmware/driver design. When there is no stream detected the LRU stream
is used for next potential stream and LRU/MRU map is updated to make this
as MRU stream. Every time a stream is detected the MRU map
is updated to make the current stream as MRU stream.
Signed-off-by: Sasikumar Chandrasekaran <sasikumar.pc@broadcom.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
An UNMAP command on a PI formatted device will leave the Logical Block Application
Tag and Logical Block Reference Tag as all F's (for those LBAs that are unmapped).
To avoid IO errors if those LBAs are subsequently read before they are written with
valid tag fields, the MPI SCSI IO requests need to set the EEDPFlags element EEDP
Escape Mode field, Bits [7:6] appropriately. A value of 2 should be set to disable
all PI checks if the Logical Block Application Tag is 0xFFFF for PI types 1 and 2.
A value of 3 should be set to disable all PI checks if the Logical Block Application
Tag is 0xFFFF and the Logical Block Reference Tag is 0xFFFFFFFF for PI type 3.
Signed-off-by: Sasikumar Chandrasekaran <sasikumar.pc@broadcom.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
We get a few warnings when building kernel with W=1:
drivers/scsi/megaraid/megaraid_sas_fusion.c:281:1: warning: no previous prototype for 'megasas_free_cmds_fusion' [-Wmissing-prototypes]
drivers/scsi/megaraid/megaraid_sas_fusion.c:714:1: warning: no previous prototype for 'megasas_ioc_init_fusion' [-Wmissing-prototypes]
....
In fact, these functions are declared in
drivers/scsi/megaraid/megaraid_sas_base.c, but should be declared in a
header file, thus can be recognized in other file.
So this patch adds the declarations into
drivers/scsi/megaraid/megaraid_sas_fusion.h.
Signed-off-by: Baoyou Xie <baoyou.xie@linaro.org>
Acked-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
This patch will create a reply queue pool for each MSI-X index and will
provide an array of base addresses instead of the single address of
legacy mode. Using this new interface the driver can support higher
queue depths through scattered DMA pools.
If array mode is not supported driver will fall back to the legacy
method of reply pool allocation. This limits controller queue depth to
1K max. To enable a queue depth of more than 1K driver requires firmware
to support array mode and scratch_pad3 will provide the new queue depth
value.
When RDPQ is used, downgrading to an older firmware release should not
be permitted. This may cause firmware fault and is not supported.
Signed-off-by: Kashyap Desai <kashyap.desai@avagotech.com>
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Firmware will fill out per-LD data to tell driver whether a particular
LD supports region lock bypass. If yes, then driver will send non-FP
LDIO to region lock bypass FIFO. With this change in driver, firmware
will optimize certain code to improve performance.
Signed-off-by: Kashyap Desai <kashyap.desai@avagotech.com>
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
This patch adds task management for SCSI commands. Added functions are
task abort and target reset.
1. Currently, megaraid_sas driver performs controller reset when any IO
times out. With task management support added, task abort and target
reset will be tried to recover timed out IO. If task management fails,
then controller reset will be performaned. If the task management
request times out, fail the request and escalate to the next
level (controller reset).
2. mr_device_priv_data will be allocated for all generations of
controller, but is_tm_capable flag will never be set for
controllers (prior to Invader series) as firmware support is not
available for task management.
3. Task management capable firmware will set is_tm_capable flag in
firmware API.
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Signed-off-by: Kashyap Desai <kashyap.desai@avagotech.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Remove PCI id based checks and use instance->ctrl_context to decide
whether controller is MFI-based or a Fusion adapter. Additionally,
Fusion adapters are divided into two categories: Thunderbolt and
Invader.
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Signed-off-by: Kashyap Desai <kashyap.desai@avagotech.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Driver will expose max sge = 256 (earlier it was 64) if firmware
supports extended IO size (1M).
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Signed-off-by: Kashyap Desai <kashyap.desai@avagotech.com>
Reviewed-by: Martin Petersen <martin.petersen@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Implemented JBOD map which will provide quick access for JBOD path and
also provide sequence number. This will help hardware to fail command
to the FW in case of any sequence mismatch.
Fast Path I/O for JBOD will refer JBOD map (which has sequence number
per JBOD device) instead of RAID map. Previously, the driver used RAID
map to get device handle for fast path I/O and this not have sequence
number information. Now, driver will use JBOD map instead. As part of
error handling, if JBOD map is failed/not supported by firmware, driver
will continue using legacy behavior.
Now there will be three IO paths for JBOD (syspd):
- JBOD map with sequence number (Fast Path)
- RAID map without sequence number (Fast Path)
- FW path via h/w exception queue deliberately setup devhandle
0xFFFF (FW path).
Relevant data structures:
- Driver send new DCMD MR_DCMD_SYSTEM_PD_MAP_GET_INFO for this purpose.
- struct MR_PD_CFG_SEQ- This structure represent map of single physical
device.
- struct MR_PD_CFG_SEQ_NUM_SYNC- This structure represent whole JBOD
map in general(size, count of sysPDs configured, struct MR_PD_CFG_SEQ
of syspD with 0 index).
- JBOD sequence map size is: sizeof(struct MR_PD_CFG_SEQ_NUM_SYNC)
+ (sizeof(struct MR_PD_CFG_SEQ) * (MAX_PHYSICAL_DEVICES - 1)) which
is allocated while setting up JBOD map at driver load time.
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Signed-off-by: Kashyap Desai <kashyap.desai@avagotech.com>
Reviewed-by: Martin Petersen <martin.petersen@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
This adds endianness annotations to all data structures, and a few
variables directly referencing them.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: James Bottomley <JBottomley@Odin.com>
megaraid_sas driver will use block layer provided tag for indexing internal
MPT frames to get any unique MPT frame tied with tag. Each IO request
submitted from SCSI mid layer will get associated MPT frame from MPT framepool
(retrieved and return back using spinlock inside megaraid_sas driver's
submission/completion call back). Getting MPT frame from MPT Frame pool is
very expensive operation because of associated spin lock operation (spinlock
overhead increase on multi NUMA node). This type of locking in driver is very
expensive call considering each IO request need - Acquire and Release of the
same lock.
With this support, in IO path driver will directly provide the unique command
index(which is based on block layer tag) and will get the MPT frame tied to
the tag and this way driver can get rid off lock, which synchronizes the
access to MPT frame pool while fetching and returning MPT frame from the pool.
This support in driver provides siginificant performance improvement(on multi
NUMA node system)on latest upstream with SCSI.MQ as well as on existing linux
distributions.
Here is the data for test executed at Avago-
- IO Tool- FIO
- 4 Socket SMC server. (4 NUMA node server)
- 12 SSDs in JBOD mode .
- 4K Rand READ, QD=32
- SCSI MQ x86_64 (Latest Upstream kernel)
- upto 300% Performance Improvement.
If IOs are running on single Node, perfromance gain is less, but as soon as
increase number of nodes, performance improvement is significant. IOs running
on all 4 NUMA nodes, with this patch applied IOPs observed was 1170K vs 344K
IOPs seen without this patch.
Logically, there are two parts of this patch- 1) Block layer tag support 2)
changes in calling convention of return_cmd. part 2 will revert the changes
done by patch- 90dc9d9 megaraid_sas : MFI MPT linked list corruption fix
because changes done in part 1 has fixed the problem of MFI MPT linked list
corruption. part 2 is very much dependent on part 1, so we decided to have
single patch for these two logical changes.
[jejb: remove chatty printk pointed out by hch]
Signed-off-by: Kashyap Desai <kashyap.desai@avagotech.com>
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: James Bottomley <JBottomley@Odin.com>
This patch addresses below issues:
1) Few endianness bug fixes.
2) Break the iteration after (MAX_LOGICAL_DRIVES_EXT - 1)),
instead of MAX_LOGICAL_DRIVES_EXT.
3) Optimization in MFI INIT frame before firing.
4) MFI IO frame should be 256bytes aligned. Code is optimized to reduce
the size of frame for fusion adapters and make the MFI frame size
calculation a bit transparent and readable.
Cc: <stable@vger.kernel.org>
Signed-off-by: Kashyap Desai <kashyap.desai@avagotech.com>
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Signed-off-by: Chaitra Basappa <chaitra.basappa@avagotech.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
In OCR (Online Controller Reset) path, driver sets adapter state to
MEGASAS_HBA_OPERATIONAL before getting new RAID map. There will be a small
window where IO will come from OS with old RAID map. This patch will
update adapter state to MEGASAS_HBA_OPERATIONAL, only after driver has new
RAID map to avoid any IOs getting build using old RAID map.
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Signed-off-by: Kashyap Desai <kashyap.desai@avagotech.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Update MAINTAINERS list and copyright information for megaraid_sas driver.
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Resending the patch. Addressed the review comments from Tomas Henzl.
Added comment for to-do work.
Problem statement:
MFI link list in megaraid_sas driver is used from mfi-mpt pass-through commands.
This list can be corrupted due to many possible race conditions in driver and
eventually we may see kernel panic.
One example -
MFI frame is freed from calling process as driver send command via polling method and interrupt
for that command comes after driver free mfi frame (actually even after some other context reuse
the mfi frame). When driver receive MPT frame in ISR, driver will be using the index of MFI and
access that MFI frame and finally in-used MFI frame’s list will be corrupted.
High level description of new solution -
Free MFI and MPT command from same context.
Free both the command either from process (from where mfi-mpt pass-through was called) or from
ISR context. Do not split freeing of MFI and MPT, because it creates the race condition which
will do MFI/MPT list corruption.
Renamed the cmd_pool_lock which is used in instance as well as fusion with below name.
mfi_pool_lock and mpt_pool_lock to add more code readability.
Signed-off-by: Sumit Saxena <sumit.saxena@avagotech.com>
Signed-off-by: Kashyap Desai <kashyap.desai@avagotech.com>
Reviewed-by: Tomas Henzl <thenzl@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Anand Lodnoor