Add module parameter to control logging levels of async event
notifications from firmware that get logged to system log. Also, allow
changing the value from sysfs after driver load.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Load balancing IO completions across all available MSI-X vectors should be
enabled for Invader and later generation controllers only. This needs to
be disabled for older controllers. Add an adapter type check before
setting msix_load_balance flag.
Fixes: 1d15d9098a ("scsi: megaraid_sas: Load balance completions across all MSI-X")
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
In megasas_get_target_prop(), driver is incorrectly calculating the target
ID for devices with channel 1 and 3. Due to this, firmware will either
fail the command (if there is no device with the target id sent from
driver) or could return the properties for a target which was not
intended. Devices could end up with the wrong queue depth due to this.
Fix target id calculation for channel 1 and 3.
Fixes: 96188a89cc ("scsi: megaraid_sas: NVME interface target prop added")
Cc: stable@vger.kernel.org
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Fix sparse warnings:
drivers/scsi/megaraid/megaraid_sas_base.c:271:1: warning: symbol 'megasas_issue_dcmd' was not declared. Should it be static?
drivers/scsi/megaraid/megaraid_sas_base.c:2227:6: warning: symbol 'megasas_do_ocr' was not declared. Should it be static?
drivers/scsi/megaraid/megaraid_sas_base.c:3194:25: warning: symbol 'megaraid_host_attrs' was not declared. Should it be static?
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
For Aero adapters, driver provides three different performance modes
controlled through module parameter named 'perf_mode'. Below are those
performance modes:
0: Balanced - Additional high IOPS reply queues will be enabled along with
low latency queues. Interrupt coalescing will be enabled only for these
high IOPS reply queues.
1: IOPS - No additional high IOPS queues are enabled. Interrupt coalescing
will be enabled on all reply queues.
2: Latency - No additional high IOPS queues are enabled. Interrupt
coalescing will be disabled on all reply queues. This is a legacy
behavior similar to Ventura & Invader Series.
Default performance mode settings:
- Performance mode set to 'Balanced', if Aero controller is working in
16GT/s PCIe speed.
- Performance mode will be set to 'Latency' mode for all other cases.
Through module parameter 'perf_mode', user can override default performance
mode to desired one.
Captured some performance numbers with these performance modes. 4k Random
Read IO performance numbers on 24 SAS SSD drives for above three
performance modes. Performance data is from Intel Skylake and HGST SS300
(drive model SDLL1DLR400GCCA1).
IOPS:
-----------------------------------------------------------------------
|perf_mode | qd = 1 | qd = 64 | note |
|-------------|--------|---------|-------------------------------------
|balanced | 259K | 3061k | Provides max performance numbers |
| | | | both on lower QD workload & |
| | | | also on higher QD workload |
|-------------|--------|---------|-------------------------------------
|iops | 220K | 3100k | Provides max performance numbers |
| | | | only on higher QD workload. |
|-------------|--------|---------|-------------------------------------
|latency | 246k | 2226k | Provides good performance numbers |
| | | | only on lower QD worklaod. |
-----------------------------------------------------------------------
Average Latency:
-----------------------------------------------------
|perf_mode | qd = 1 | qd = 64 |
|-------------|--------------|----------------------|
|balanced | 92.05 usec | 501.12 usec |
|-------------|--------------|----------------------|
|iops | 108.40 usec | 498.10 usec |
|-------------|--------------|----------------------|
|latency | 97.10 usec | 689.26 usec |
-----------------------------------------------------
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>
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>
High iops queues are mapped to non-managed IRQs. Set affinity of
non-managed irqs to local numa node. Low latency queues are mapped to
managed IRQs.
Driver reserves some reply queues for high IOPS queues (through
pci_alloc_irq_vectors_affinity and .pre_vectors interface). The rest of
queues are for low latency.
Based on IO workload, driver will decide which group of reply queues
(either high IOPS queues or low latency queues) to be used.
High IOPS queues will be mapped to local numa node of controller and
low latency queues will be mapped to CPUs across numa nodes. In general,
high IOPS and low latency queues should fit into 128 reply queues
which is the max number of reply queues supported by Aero adapters.
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>
Driver should enable interrupt coalescing (during driver load and after
Controller Reset) for High IOPS queues by masking appropriate bits in IOC
INIT frame.
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>
Aero controllers support balanced performance mode through the ability to
configure queues with different properties.
Reply queues with interrupt coalescing enabled are called "high iops reply
queues" and reply queues with interrupt coalescing disabled are called "low
latency reply queues".
The driver configures a combination of high iops and low latency reply
queues if:
- HBA is an AERO controller;
- MSI-X vectors supported by the HBA is 128;
- Total CPU count in the system more than high iops queue count;
- Driver is loaded with default max_msix_vectors module parameter; and
- System booted in non-kdump mode.
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>
Added driver support to allow passthrough MPI toolbox type MFI commands to
firmware based on firmware capability.
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>
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>
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Chandrakanth Patil <chandrakanth.patil@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Issue: This issue is applicable to scenario when JBOD sequence map is
unavailable (memory allocation for JBOD sequence map failed) to driver but
feature is supported by firmware. If the driver sends a JBOD IO by not
adding 255 (MAX_PHYSICAL_DEVICES - 1) to device ID when underlying firmware
supports JBOD sequence map, it will lead to the IO failure.
Fix: For JBOD IOs, driver will not use the RAID map to fetch the devhandle
if JBOD sequence map is unavailable. Driver will set Devhandle to 0xffff
and Target ID to 'device ID + 255 (MAX_PHYSICAL_DEVICES - 1)'.
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>
Firmware does not expect FastPath IO sent through Region Lock Bypass queue.
Though firmware never exposes such settings when fastpath IO can be sent to
RL bypass queue but it's safer to remove dead code which directs fastpath
IO to RL Bypass queue.
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>
Issue: Under certain conditions, controller goes in FAULT state after IOC
INIT fired to firmware. Such Fault can be recovered through controller
reset.
Fix: In driver probe context, if firmware fault is observed post IOC INIT,
driver would do controller reset followed by retry logic for IOC INIT
command.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Chandrakanth Patil <chandrakanth.patil@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Issue: There is possibility of few DCMDs timing out with 'reset_mutex' lock
held. As part of DCMD timeout handling, driver calls function
megasas_reset_fusion which also tries to acquire same lock 'reset_mutex'
and end up with deadlock.
Fix: Upon timeout of DCMDs (which are fired with 'reset_mutex' lock held),
driver will release 'reset_mutex' before calling OCR function and will
acquire lock again after OCR function returns.
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>
On PowerPC architecture, calling disable_irq_nosync from IRQ context is not
providing the required effect.
In current megaraid_sas driver, disable_irq_nosync is being called from IRQ
context before enabling IRQ poll. But due to the issue seen on PPC, after
IRQ poll disable and legacy ISR is enabled, we are not seeing our ISR
getting called.
Fix: Call disable_irq from IRQ poll thread context instead of IRQ context.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Chandrakanth Patil <chandrakanth.patil@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
This patch will add support for non-secure Aero adapter PCI IDs. Driver
will throw an error message when a non-secure type controller is
detected. Purpose of this interface is to avoid interacting with any
firmware which is not secured/signed by Broadcom. Any tampering on Firmware
component will be detected by hardware and it will be communicated to the
driver to avoid any further interaction with that component.
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>
Aero adapters provides Atomic Request Descriptor as an alternative method
for posting an entry onto a request queue. The posting of an Atomic Request
Descriptor is an atomic operation, providing a safe mechanism for multiple
processors on the host to post requests without synchronization. This
Atomic Request Descriptor format is identical to first 32 bits of Default
Request Descriptor and uses only 32 bits.
If Aero adapters support Atomic descriptor, driver should use it for
posting IOs and DCMDs to firmware.
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>
One of the more common cases of allocation size calculations is finding the
size of a structure that has a zero-sized array at the end, along with
memory for some number of elements for that array. For example:
struct MR_PD_CFG_SEQ_NUM_SYNC {
...
struct MR_PD_CFG_SEQ seq[1];
} __packed;
Make use of the struct_size() helper instead of an open-coded version in
order to avoid any potential type mistakes.
So, replace the following form:
sizeof(struct MR_PD_CFG_SEQ_NUM_SYNC) + (sizeof(struct MR_PD_CFG_SEQ) * (MAX_PHYSICAL_DEVICES - 1))
with:
struct_size(pd_sync, seq, MAX_PHYSICAL_DEVICES - 1)
This code was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Acked-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Remove including <linux/version.h> that don't need it.
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Use existing macros. No functional change.
Signed-off-by: Tomas Henzl <thenzl@redhat.com>
Acked-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Checkpatch emits a warning when using symbolic permissions. Use octal
permissions instead. No functional change.
Signed-off-by: Tomas Henzl <thenzl@redhat.com>
Acked-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Support is easier with all driver parameters visible in sysfs.
Signed-off-by: Tomas Henzl <thenzl@redhat.com>
Acked-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Fixes gcc '-Wunused-but-set-variable' warnings:
drivers/scsi/megaraid/megaraid_sas_base.c: In function megasas_fw_crash_buffer_show:
drivers/scsi/megaraid/megaraid_sas_base.c:3138:16: warning: variable buff_addr set but not used [-Wunused-but-set-variable]
drivers/scsi/megaraid/megaraid_sas_base.c: In function megasas_get_pd_list:
drivers/scsi/megaraid/megaraid_sas_base.c:4426:13: warning: variable ci_h set but not used [-Wunused-but-set-variable]
'buff_addr' is never used since inroduction in commit fc62b3fc90
("megaraid_sas : Firmware crash dump feature support")
'ci_h' is not used since commit 9b3d028f34 ("scsi: megaraid_sas:
Pre-allocate frequently used DMA buffers")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Fixes gcc '-Wunused-but-set-variable' warning:
drivers/scsi/megaraid/megaraid_sas_base.c: In function megasas_create_frame_pool:
drivers/scsi/megaraid/megaraid_sas_base.c:4124:6: warning: variable sge_sz set but not used [-Wunused-but-set-variable]
It's not used any more since commit 200aed582d ("megaraid_sas: endianness
related bug fixes and code optimization")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Fixes gcc '-Wunused-but-set-variable' warnings:
drivers/scsi/megaraid/megaraid_sas_base.c: In function megasas_suspend:
drivers/scsi/megaraid/megaraid_sas_base.c:7269:20: warning: variable host set but not used [-Wunused-but-set-variable]
drivers/scsi/megaraid/megaraid_sas_base.c: In function megasas_aen_polling:
drivers/scsi/megaraid/megaraid_sas_base.c:8397:15: warning: variable wait_time set but not used [-Wunused-but-set-variable]
'host' never used since introduction in commit 31ea708897 ("[SCSI]
megaraid_sas: add hibernation support")
'wait_time' never used since commit 11c71cb4ab ("megaraid_sas: Do
not allow PCI access during OCR")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Fixes gcc '-Wunused-but-set-variable' warning:
drivers/scsi/megaraid/megaraid_sas_base.c: In function megasas_transition_to_ready:
drivers/scsi/megaraid/megaraid_sas_base.c:3900:6: warning: variable cur_state set but not used [-Wunused-but-set-variable]
Never used since commit 7218df69e3 ("[SCSI] megaraid_sas: use the
firmware boot timeout when waiting for commands")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Sumit Saxena <sumit.saxena@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>
Print FW supported MSI-X vector count only if FW supports
MSI-X.
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>
Add debug prints related to device list being returned by firmware. The a
debug flag to activate these prints.
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>
Add prints in resume/suspend path to help in debugging hibernation
issues. The print gives an indication when the driver entry points are
called.
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>
Add a print to dump the interrupt status in system log for debugging.
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>
When driver detects a firmware fault during load, dump additional
information on fault code and subcode that will help in debugging.
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>
Add a sysfs interface to get the raid map index that is being used by
driver.
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>
Add prints for BAR address information during driver load. This helps in
debugging issues with BAR address changing during OS boot.
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>
When controller fails to transition to READY state during driver probe,
dump the system interface register set. This will give snapshot of the
firmware status for debugging driver load issues.
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>
Add a sysfs interface to dump the controller's system interface registers.
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>
Add option to format the buffer that is being dumped. Currently, the IO
frame and chain frame dumped in the syslog is getting split across multiple
lines based on the formatting. Fix this by using KERN_CONT in printk.
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>
Add prints to identify the internal DCMD opcode that has timed out.
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>
This patch enhances the existing debug prints in reset and task management
path.
These debug prints in adapter reset path helps with debugging issues
related to IO timeouts that are seen frequently in the field. Add
additional debug prints to dump the pending command frames before
initiating an adapter reset. Also, print FastPath IOs that are
outstanding.
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>
Driver will use "reply descriptor post queues" in round robin fashion when
the combined MSI-X mode is not enabled. With this IO completions are
distributed and load balanced across all the available reply descriptor
post queues equally.
This is enabled only if combined MSI-X mode is not enabled in firmware.
This improves performance and also fixes soft lockups.
When load balancing is enabled, IRQ affinity from driver needs to be
disabled.
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>
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>
While an online controller reset(OCR) is in progress, there is short
duration where all access to controller's PCI config space from the host
needs to be blocked. This is due to a hardware limitation of MegaRAID
controllers.
With this patch, driver will block all access to controller's config space
from userland applications by calling pci_cfg_access_lock() while OCR is in
progress and unlocking after controller comes back to ready state.
Added helper function which locks the config space before initiating OCR
and wait for controller to become READY.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
No functional change. This patch reworks code around controller reset path
which gets rid of a couple of goto labels. This is in preparation for the
next patch which adds PCI config space access locking while controller
reset is in progress.
Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
