Enable the driver to work in non-IRQ mode, i.e. there will not be any MSI-X
vectors associated with queues dedicated to polling. The IOC hardware is
single submission queue and multiple reply queue. However, using the shared
host tagset support it is possible to simulate multiple hardware queues.
When poll_queues are enabled through the module parameter, the driver will
allocate extra reply queues without an MSI-X association. All I/O
completion on these queues will be done through the iopoll interface.
Link: https://lore.kernel.org/r/20210727081212.2742-1-sreekanth.reddy@broadcom.com
Signed-off-by: Sreekanth Reddy <sreekanth.reddy@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
When running command pipelining for WRITE direction commands (e.g. tape
device write), userspace sends cmd completion to cmd ring before processing
write data. In that case userspace has to copy data before sending
completion, because cmd completion also implicitly releases the data buffer
in data area.
The new feature KEEP_BUF allows userspace to optionally keep the buffer
after completion by setting new bit TCMU_UFLAG_KEEP_BUF in
tcmu_cmd_entry_hdr->uflags. In that case buffer has to be released
explicitly by writing the cmd_id to new action item free_kept_buf.
All kept buffers are released during reset_ring and if userspace closes uio
device (tcmu_release).
Link: https://lore.kernel.org/r/20210713175021.20103-1-bostroesser@gmail.com
Reviewed-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Bodo Stroesser <bostroesser@gmail.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Clearing a unit attention synchronously from inside the UFS error handler
may trigger the following deadlock:
- ufshcd_err_handler() calls ufshcd_err_handling_unprepare() and the
latter function calls ufshcd_clear_ua_wluns().
- ufshcd_clear_ua_wluns() submits a REQUEST SENSE command and that command
activates the SCSI error handler.
- The SCSI error handler calls ufshcd_host_reset_and_restore().
- ufshcd_host_reset_and_restore() executes the following code:
ufshcd_schedule_eh_work(hba); flush_work(&hba->eh_work);
This sequence results in a deadlock (circular wait). Fix this by requesting
sense data asynchronously.
Link: https://lore.kernel.org/r/20210722033439.26550-16-bvanassche@acm.org
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Stanley Chu <stanley.chu@mediatek.com>
Cc: Can Guo <cang@codeaurora.org>
Cc: Asutosh Das <asutoshd@codeaurora.org>
Cc: Avri Altman <avri.altman@wdc.com>
Reviewed-by: Bean Huo <beanhuo@micron.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Make the following changes in ufshcd_abort():
- Return FAILED instead of SUCCESS if the abort handler notices that a
SCSI command has already been completed. Returning SUCCESS in this case
triggers a use-after-free and may trigger a kernel crash.
- Fix the code for aborting SCSI commands submitted to a WLUN.
The current approach for aborting SCSI commands that have been submitted to
a WLUN and that timed out is as follows:
- Report to the SCSI core that the command has completed successfully.
Let the block layer free any data buffers associated with the command.
- Mark the command as outstanding in 'outstanding_reqs'.
- If the block layer tries to reuse the tag associated with the aborted
command, busy-wait until the tag is freed.
This approach can result in:
- Memory corruption if the controller accesses the data buffer after the
block layer has freed the associated data buffers.
- A race condition if ufshcd_queuecommand() or ufshcd_exec_dev_cmd()
checks the bit that corresponds to an aborted command in
'outstanding_reqs' after it has been cleared and before it is reset.
- High energy consumption if ufshcd_queuecommand() repeatedly returns
SCSI_MLQUEUE_HOST_BUSY.
Fix this by reporting to the SCSI error handler that aborting a SCSI
command failed if the SCSI command was submitted to a WLUN.
Link: https://lore.kernel.org/r/20210722033439.26550-15-bvanassche@acm.org
Fixes: 7a7e66c65d ("scsi: ufs: Fix a race condition between ufshcd_abort() and eh_work()")
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Stanley Chu <stanley.chu@mediatek.com>
Cc: Can Guo <cang@codeaurora.org>
Cc: Asutosh Das <asutoshd@codeaurora.org>
Cc: Avri Altman <avri.altman@wdc.com>
Reviewed-by: Bean Huo <beanhuo@micron.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
From arch/arm/include/asm/io.h
#define __iowmb() wmb()
[ ... ]
#define writel(v,c) ({ __iowmb(); writel_relaxed(v,c); })
From Documentation/memory-barriers.txt: "Note that, when using writel(), a
prior wmb() is not needed to guarantee that the cache coherent memory
writes have completed before writing to the MMIO region."
In other words, calling wmb() before writel() is not necessary. Hence
remove the wmb() calls that precede a writel() call. Remove the wmb() calls
that precede a ufshcd_send_command() call since the latter function uses
writel(). Remove the wmb() call from ufshcd_wait_for_dev_cmd() since the
following chain of events guarantees that the CPU will see up-to-date LRB
values:
- UFS controller writes to host memory.
- UFS controller posts completion interrupt after the memory writes from
the previous step are visible to the CPU.
- complete(hba->dev_cmd.complete) is called from the UFS interrupt handler.
- The wait_for_completion(hba->dev_cmd.complete) call in
ufshcd_wait_for_dev_cmd() returns.
Link: https://lore.kernel.org/r/20210722033439.26550-10-bvanassche@acm.org
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Stanley Chu <stanley.chu@mediatek.com>
Cc: Can Guo <cang@codeaurora.org>
Cc: Asutosh Das <asutoshd@codeaurora.org>
Cc: Avri Altman <avri.altman@wdc.com>
Tested-by: Avri altman <avri.altman@wdc.com>
Reviewed-by: Avri Altman <avri.altman@wdc.com>
Reviewed-by: Daejun Park <daejun7.park@samsung.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
In order not to hang on to "cold" regions, we inactivate a region that has
had no READ access for a predefined amount of time - READ_TO_MS. For that
purpose monitor the active regions list, polling it on every
POLLING_INTERVAL_MS. On timeout expiry add the region to the
"to-be-inactivated" list unless it is clean and did not exhaust its
READ_TO_EXPIRIES - another parameter.
None of this applies to pinned regions.
Link: https://lore.kernel.org/r/20210712095039.8093-9-avri.altman@wdc.com
Reviewed-by: Daejun Park <daejun7.park@samsung.com>
Signed-off-by: Avri Altman <avri.altman@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
In host control mode, reads are the major source of activation trials.
Keep track of those reads counters, for both active as well inactive
regions.
We reset the read counter upon write - we are only interested in "clean"
reads.
Keep those counters normalized, as we are using those reads as a
comparative score, to make various decisions. If during consecutive
normalizations an active region has exhaust its reads - inactivate it.
While at it, protect the {active,inactive}_count stats by adding them into
the applicable handler.
Link: https://lore.kernel.org/r/20210712095039.8093-5-avri.altman@wdc.com
Reviewed-by: Daejun Park <daejun7.park@samsung.com>
Signed-off-by: Avri Altman <avri.altman@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
In device control mode, the device may recommend the host to either
activate or inactivate a region, and the host should follow. Meaning those
are not actually recommendations, but more of instructions.
Conversely, in host control mode, the recommendation protocol is slightly
changed:
a) The device may only recommend the host to update a subregion of an
already-active region. And,
b) The device may *not* recommend to inactivate a region.
Furthermore, in host control mode, the host may choose not to follow any of
the device's recommendations. However, in case of a recommendation to
update an active and clean subregion, it is better to follow those
recommendation because otherwise the host has no other way to know that
some internal relocation took place.
Link: https://lore.kernel.org/r/20210712095039.8093-3-avri.altman@wdc.com
Reviewed-by: Daejun Park <daejun7.park@samsung.com>
Signed-off-by: Avri Altman <avri.altman@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Implement L2P map management in HPB.
The HPB divides logical addresses into several regions. A region consists
of several sub-regions. The sub-region is a basic unit where L2P mapping is
managed. The driver loads L2P mapping data of each sub-region. The loaded
sub-region is called active-state. The HPB driver unloads L2P mapping data
as region unit. The unloaded region is called inactive-state.
Sub-region/region candidates to be loaded and unloaded are delivered from
the UFS device. The UFS device delivers the recommended active sub-region
and inactivate region to the driver using sense data. The HPB module
performs L2P mapping management on the host through the delivered
information.
A pinned region is a preset region on the UFS device that is always
in activate-state.
The data structures for map data requests and L2P mappings use the mempool
API, minimizing allocation overhead while avoiding static allocation.
The mininum size of the memory pool used in the HPB is implemented
as a module parameter so that it can be configurable by the user.
To guarantee a minimum memory pool size of 4MB: ufshpb_host_map_kbytes=4096.
The map_work manages active/inactive via 2 "to-do" lists:
- hpb->lh_inact_rgn: regions to be inactivated
- hpb->lh_act_srgn: subregions to be activated
These lists are maintained on I/O completion.
[mkp: switch to REQ_OP_DRV_*]
Link: https://lore.kernel.org/r/20210712085859epcms2p36e420f19564f6cd0c4a45d54949619eb@epcms2p3
Tested-by: Bean Huo <beanhuo@micron.com>
Tested-by: Can Guo <cang@codeaurora.org>
Tested-by: Stanley Chu <stanley.chu@mediatek.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Can Guo <cang@codeaurora.org>
Reviewed-by: Bean Huo <beanhuo@micron.com>
Reviewed-by: Stanley Chu <stanley.chu@mediatek.com>
Acked-by: Avri Altman <Avri.Altman@wdc.com>
Signed-off-by: Daejun Park <daejun7.park@samsung.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Implement Host Performance Buffer (HPB) initialization and add function
calls to UFS core driver.
NAND flash-based storage devices, including UFS, have mechanisms to
translate logical addresses of I/O requests to the corresponding physical
addresses of the flash storage. In UFS, logical-to-physical-address (L2P)
map data, which is required to identify the physical address for the
requested I/Os, can only be partially stored in SRAM from NAND flash. Due
to this partial loading, accessing the flash address area, where the L2P
information for that address is not loaded in the SRAM, can result in
serious performance degradation.
The basic concept of HPB is to cache L2P mapping entries in host system
memory so that both physical block address (PBA) and logical block address
(LBA) can be delivered in HPB read command. The HPB read command allows to
read data faster than a regular read command in UFS since it provides the
physical address (HPB Entry) of the desired logical block in addition to
its logical address. The UFS device can access the physical block in NAND
directly without searching and uploading L2P mapping table. This improves
read performance because the NAND read operation for uploading L2P mapping
table is removed.
In HPB initialization, the host checks if the UFS device supports HPB
feature and retrieves related device capabilities. Then, HPB parameters are
configured in the device.
Total start-up time of popular applications was measured and the difference
observed between HPB being enabled and disabled. Popular applications are
12 game apps and 24 non-game apps. Each test cycle consists of running 36
applications in sequence. We repeated the cycle for observing performance
improvement by L2P mapping cache hit in HPB.
The following is the test environment:
- kernel version: 4.4.0
- RAM: 8GB
- UFS 2.1 (64GB)
Results:
+-------+----------+----------+-------+
| cycle | baseline | with HPB | diff |
+-------+----------+----------+-------+
| 1 | 272.4 | 264.9 | -7.5 |
| 2 | 250.4 | 248.2 | -2.2 |
| 3 | 226.2 | 215.6 | -10.6 |
| 4 | 230.6 | 214.8 | -15.8 |
| 5 | 232.0 | 218.1 | -13.9 |
| 6 | 231.9 | 212.6 | -19.3 |
+-------+----------+----------+-------+
We also measured HPB performance using iozone:
$ iozone -r 4k -+n -i2 -ecI -t 16 -l 16 -u 16 -s $IO_RANGE/16 -F \
mnt/tmp_1 mnt/tmp_2 mnt/tmp_3 mnt/tmp_4 mnt/tmp_5 mnt/tmp_6 mnt/tmp_7 \
mnt/tmp_8 mnt/tmp_9 mnt/tmp_10 mnt/tmp_11 mnt/tmp_12 mnt/tmp_13 \
mnt/tmp_14 mnt/tmp_15 mnt/tmp_16
Results:
+----------+--------+---------+
| IO range | HPB on | HPB off |
+----------+--------+---------+
| 1 GB | 294.8 | 300.87 |
| 4 GB | 293.51 | 179.35 |
| 8 GB | 294.85 | 162.52 |
| 16 GB | 293.45 | 156.26 |
| 32 GB | 277.4 | 153.25 |
+----------+--------+---------+
Link: https://lore.kernel.org/r/20210712085830epcms2p8c1288b7f7a81b044158a18232617b572@epcms2p8
Reported-by: kernel test robot <lkp@intel.com>
Tested-by: Bean Huo <beanhuo@micron.com>
Tested-by: Can Guo <cang@codeaurora.org>
Tested-by: Stanley Chu <stanley.chu@mediatek.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Can Guo <cang@codeaurora.org>
Reviewed-by: Bean Huo <beanhuo@micron.com>
Reviewed-by: Stanley Chu <stanley.chu@mediatek.com>
Acked-by: Avri Altman <Avri.Altman@wdc.com>
Signed-off-by: Daejun Park <daejun7.park@samsung.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The macros cpu_to_le16() and cpu_to_le32() have special cases for
constants. Their __constant_<foo> versions are not required.
On little endian systems, both cpu_to_le16() and __constant_cpu_to_le16()
expand to the same expression. Same is the case with cpu_to_le32().
On big endian systems, cpu_to_le16() expands to __swab16() which has a
__builtin_constant_p check. Similarly, cpu_to_le32() expands to __swab32().
Consequently these macros can be safely used with constants, and hence all
those uses are converted. This was discovered as a part of a checkpatch
evaluation, looking at all reports of WARNING:CONSTANT_CONVERSION error
type.
Link: https://lore.kernel.org/r/20210716112852.24598-1-dwaipayanray1@gmail.com
Signed-off-by: Dwaipayan Ray <dwaipayanray1@gmail.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Existing blogic_msg() invocations do not appear to overrun its internal
buffer of a fixed length of 100, which would cause stack corruption, but
it's easy to miss with possible further updates and a fix is cheap in
performance terms, so limit the output produced into the buffer by using
vscnprintf() rather than vsprintf().
Link: https://lore.kernel.org/r/alpine.DEB.2.21.2104201939390.44318@angie.orcam.me.uk
Acked-by: Khalid Aziz <khalid@gonehiking.org>
Signed-off-by: Maciej W. Rozycki <macro@orcam.me.uk>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
