Hannes Reinecke [Fri, 28 Feb 2020 07:53:12 +0000 (08:53 +0100)]
scsi: aacraid: move scsi_(block,unblock)_requests out of _aac_reset_adapter()
_aac_reset_adapter() only has one caller, and that one already calls
scsi_block_requests(). Move the calls out of _aac_reset_adapter() to avoid
calling scsi_block_requests() twice.
Link: https://lore.kernel.org/r/20200228075318.91255-8-hare@suse.de Reviewed-by: Christoph Hellwig <hch@lst.de> Acked-by: Balsundar P < Balsundar.P@microchip.com> Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Hannes Reinecke [Fri, 28 Feb 2020 07:53:11 +0000 (08:53 +0100)]
scsi: aacraid: replace aac_flush_ios() with midlayer helper
Use the midlayer helper scsi_host_complete_all_commands() to flush all
outstanding commands.
Link: https://lore.kernel.org/r/20200228075318.91255-7-hare@suse.de Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Acked-by: Balsundar P <balsundar.p@microchip.com> Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Hannes Reinecke [Fri, 28 Feb 2020 07:53:10 +0000 (08:53 +0100)]
scsi: aacraid: use scsi_host_complete_all_commands() to terminate outstanding commands
Use scsi_host_complete_all_commands() to terminate all outstanding commands
and change the command result for terminated commands to the more common
'DID_RESET' instead of 'QUEUE_FULL'.
Link: https://lore.kernel.org/r/20200228075318.91255-6-hare@suse.de Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Acked-by: Balsundar P <balsundar.p@microchip.com> Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Hannes Reinecke [Fri, 28 Feb 2020 07:53:09 +0000 (08:53 +0100)]
scsi: aacraid: Do not wait for outstanding write commands on synchronize_cache
There is no need to wait for outstanding write commands on synchronize
cache; the block layer is responsible for I/O scheduling, no need to
out-guess it in the driver layer.
Link: https://lore.kernel.org/r/20200228075318.91255-5-hare@suse.de Reviewed-by: Christoph Hellwig <hch@lst.de> Acked-by: Balsundar P <balsundar.b@microchip.com> Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Hannes Reinecke [Fri, 28 Feb 2020 07:53:08 +0000 (08:53 +0100)]
scsi: dpt_i2o: use scsi_host_complete_all_commands() to abort outstanding commands
Rather than traversing all outstanding commands manually, use the
scsi_host_complete_all_commands() helper to terminate all commands during
reset. With that we can drop the cmd_list usage from the midlayer.
Link: https://lore.kernel.org/r/20200228075318.91255-4-hare@suse.de Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Hannes Reinecke [Fri, 28 Feb 2020 07:53:06 +0000 (08:53 +0100)]
scsi: dpt_i2o: rename adpt_i2o_to_scsi() to adpt_i2o_scsi_complete()
Rename the badly named function into adpt_i2o_scsi_complete(), and make it
a void function as the return value is never used. This also fixes a
potential use-after-free as the return value might be evaluated from the
command result after the command has been freed.
Link: https://lore.kernel.org/r/20200228075318.91255-2-hare@suse.de Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Stanley Chu [Thu, 20 Feb 2020 13:48:48 +0000 (21:48 +0800)]
scsi: ufs: ufs-mediatek: add waiting time for reference clock
Some delays may be required either after gating or before ungating
reference clock for device according to vendor requirements.
Note that in UFS 3.0, the delay time after gating reference
clock can be defined by attribute bRefClkGatingWaitTime. Use the
formal value instead if it can be queried from device.
Can Guo [Mon, 24 Feb 2020 04:09:21 +0000 (20:09 -0800)]
scsi: ufs: Allow vendor device quirks to be applied early
Currently ufshcd_vops_apply_dev_quirks() comes after all UniPro parameters
have been tuned. Move it up so that vendors have a chance to apply device
quirks in advance.
Quinn Tran [Wed, 26 Feb 2020 22:40:21 +0000 (14:40 -0800)]
scsi: qla2xxx: Set Nport ID for N2N
When transitioning from loop to N2N, stale NPort ID is not
re-assigned. Stale ID can collide with remote device. This patch will
re-assign NPort ID on N2N is detected.
Arun Easi [Wed, 26 Feb 2020 22:40:20 +0000 (14:40 -0800)]
scsi: qla2xxx: Handle NVME status iocb correctly
Certain state flags bit combinations are not checked and not handled
correctly. Plus, do not log a normal underrun situation where there is
no frame drop.
Quinn Tran [Wed, 26 Feb 2020 22:40:18 +0000 (14:40 -0800)]
scsi: qla2xxx: Serialize fc_port alloc in N2N
For N2N, fc_port struct is created during report id acquisition. At
later time, the loop resync (fabric, n2n, loop) would trigger the rest
of the login using the created fc_port struct. The loop resync logic
can trigger another fc_port allocation if the 1st allocation was not
able to execute. This patch prevents the 2nd allocation trigger.
Quinn Tran [Wed, 26 Feb 2020 22:40:17 +0000 (14:40 -0800)]
scsi: qla2xxx: Fix NPIV instantiation after FW dump
NPIV re-enable code was blocked after FW had been initialized. The
blocking check was too broad. Reduce the check to make sure if FW is
initialized or not.
Quinn Tran [Wed, 26 Feb 2020 22:40:16 +0000 (14:40 -0800)]
scsi: qla2xxx: Fix RDP respond data format
RPD information failed to display by switch cli command. This is
caused by driver failure to properly format RDP response data with
data descriptor to allow switch to parse it correctly.
Quinn Tran [Wed, 26 Feb 2020 22:40:15 +0000 (14:40 -0800)]
scsi: qla2xxx: Force semaphore on flash validation failure
For single port 28XX adapter, the second core can still run in the
background. The flash semaphore can be held by the non-active core.
This patch tell MPI FW to check for this case and clear the semaphore
from the non-active core.
Andrew Vasquez [Wed, 26 Feb 2020 22:40:13 +0000 (14:40 -0800)]
scsi: qla2xxx: Update BPM enablement semantics.
commit e4e3a2ce9556 ("scsi: qla2xxx: Add ability to autodetect SFP
type") takes a heavy handed approach to BPM (Buffer Plus Management)
enablement:
1) During hardware initialization, if an LR-capable transceiver is
recognized, the driver schedules a disruptive post-initialization
chip-reset (ISP-ABORT) to allow the BPM settings to be sent to the
firmware. This chip-reset will result in (short-term) path-loss to
all fc-rports and their attached SCSI devices.
2) LR-detection is triggered during any link-up event, resulting in a
refresh and potential chip-reset
Based on firmware-team guidance, upon LR-capable transceiver
recognition, the driver's hardware initialization code will now
re-execute firmware with the new BPM settings, then continue on with
driver initialization. To address the second issue, the driver
performs LR-capable detection upon the driver receiving a
transceiver-insertion asynchronous event from firmware. No short-term
path loss is needed with this new semantic.
Andrew Vasquez [Wed, 26 Feb 2020 22:40:11 +0000 (14:40 -0800)]
scsi: qla2xxx: Use a dedicated interrupt handler for 'handshake-required' ISPs
There's no point checking flags.disable_msix_handshake in the
interrupt handler hot-path. Instead perform the check during
queue-pair instantiation and use the proper interrupt handler.
Bart Van Assche [Thu, 20 Feb 2020 04:34:40 +0000 (20:34 -0800)]
scsi: qla2xxx: Convert MAKE_HANDLE() from a define into an inline function
This patch allows sparse to verify the endianness of the arguments passed
to make_handle().
Link: https://lore.kernel.org/r/20200220043441.20504-5-bvanassche@acm.org Cc: Roman Bolshakov <r.bolshakov@yadro.com> Cc: Daniel Wagner <dwagner@suse.de> Cc: Martin Wilck <mwilck@suse.com> Cc: Quinn Tran <qutran@marvell.com> Reviewed-by: Daniel Wagner <dwagner@suse.de> Reviewed-by: Roman Bolshakov <r.bolshakov@yadro.com> Acked-by: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Bart Van Assche [Thu, 20 Feb 2020 04:34:38 +0000 (20:34 -0800)]
scsi: qla2xxx: Suppress endianness complaints in qla2x00_configure_local_loop()
Instead of changing endianness in-place, write the data in CPU endian
format in another buffer and copy that buffer back. This patch does not
change any functionality but silences some sparse endianness warnings.
Link: https://lore.kernel.org/r/20200220043441.20504-3-bvanassche@acm.org Cc: Roman Bolshakov <r.bolshakov@yadro.com> Cc: Martin Wilck <mwilck@suse.com> Cc: Quinn Tran <qutran@marvell.com> Reviewed-by: Daniel Wagner <dwagner@suse.de> Reviewed-by: Roman Bolshakov <r.bolshakov@yadro.com> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Bart Van Assche [Thu, 20 Feb 2020 04:34:37 +0000 (20:34 -0800)]
scsi: qla2xxx: Simplify the code for aborting SCSI commands
Since the SCSI core does not reuse the tag of the SCSI command that is
being aborted by .eh_abort() before .eh_abort() has finished it is not
necessary to check from inside that callback whether or not the SCSI
command has already completed. Instead, rely on the firmware to return an
error code when attempting to abort a command that has already
completed. Additionally, rely on the firmware to return an error code when
attempting to abort an already aborted command.
In qla2x00_abort_srb(), use blk_mq_request_started() instead of
sp->completed and sp->aborted.
Link: https://lore.kernel.org/r/20200220043441.20504-2-bvanassche@acm.org Cc: Martin Wilck <mwilck@suse.com> Cc: Quinn Tran <qutran@marvell.com> Reviewed-by: Daniel Wagner <dwagner@suse.de> Reviewed-by: Roman Bolshakov <r.bolshakov@yadro.com> Acked-by: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Link: https://lore.kernel.org/r/20200227201148.13973-1-hmadhani@marvell.com Fixes: a31056ddc665 ("scsi: qla2xxx: Use endian macros to assign static fields in fwdump header") Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Himanshu Madhani <hmadhani@marvell.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Merlijn Wajer [Tue, 18 Feb 2020 14:39:17 +0000 (15:39 +0100)]
scsi: sr: get rid of sr global mutex
When replacing the Big Kernel Lock in commit 2a48fc0ab242 ("block:
autoconvert trivial BKL users to private mutex"), the lock was replaced
with a sr-wide lock.
This causes very poor performance when using multiple sr devices, as the sr
driver was not able to execute more than one command to one drive at any
given time, even when there were many CD drives available.
Replace the global mutex with per-sr-device mutex.
Someone tried this patch at the time, but it never made it upstream, due to
possible concerns with race conditions, but it's not clear the patch
actually caused those:
scsi: sr: remove references to BLK_DEV_SR_VENDOR, leave it enabled
This kernel configuration is basically enabling/disabling sr driver quirks
detection. While these quirks are for fairly rare devices (very old CD
burners, and a glucometer), the additional detection of these models is a
very minimal amount of code.
The logic behind the quirks is always built into the sr driver.
This also removes the config from all the defconfig files that are enabling
this already.
Hannes Reinecke [Thu, 13 Feb 2020 15:32:06 +0000 (16:32 +0100)]
scsi: ch: synchronize ch_probe() and ch_open()
The 'ch' device node is created before the configuration is being read in,
which leads to a race window when ch_open() is called before that.
To avoid any races we should be taking the device mutex during
ch_readconfig() and ch_init_elem(), and also during ch_open().
That ensures ch_probe is finished before ch_open() completes.
Hannes Reinecke [Thu, 13 Feb 2020 15:32:05 +0000 (16:32 +0100)]
scsi: ch: fixup refcounting imbalance for SCSI devices
The SCSI device is required to be present during ch_probe() and ch_open().
But the SCSI device itself is only checked during ch_open(), so it's anyones
guess if it had been present during ch_probe(). And consequently we can't
reliably detach it during ch_release(), as ch_remove() might have been
called first. So initialize the changer device during ch_probe(), and take
a reference to the SCSI device during both ch_probe() and ch_open().
Joe Carnuccio [Wed, 12 Feb 2020 21:44:25 +0000 (13:44 -0800)]
scsi: qla2xxx: Handle cases for limiting RDP response payload length
This patch reduces RDP response payload length, if requesting port is a
domain controller (sid 0xfffc01) and fw is earlier than 8.09.00 and fw is
not 8.05.65 then limit the RDP response payload length to maximum of 256
bytes by terminating the response just before the optical element
descriptor.
Joe Carnuccio [Wed, 12 Feb 2020 21:44:19 +0000 (13:44 -0800)]
scsi: qla2xxx: Add ql2xrdpenable module parameter for RDP
This patch provides separate module parameter ql2xrdpenable to turn on/off
RDP capability in the driver. However, if ql2xsmartsan parameter is
enabled, it will also turn on ql2xfdmienable parameter since it is required
for RDP to work.
Joe Carnuccio [Wed, 12 Feb 2020 21:44:18 +0000 (13:44 -0800)]
scsi: qla2xxx: Add vendor extended RDP additions and amendments
This patch adds RDP command support in the driver. With the help of new
ql2xsmartsan parameter, driver will use PUREX IOCB mode to send RDP command
to switch and will be able to receive various diagnostic data.
Joe Carnuccio [Wed, 12 Feb 2020 21:44:17 +0000 (13:44 -0800)]
scsi: qla2xxx: Add changes in preparation for vendor extended FDMI/RDP
This patch prepares code for implementing Vendor specific extended FDMI/RDP
commands. It also addes support for MBC_GET_PORT_DATABASE and
MBC_GET_RNID_PARAMS commands.
Joe Carnuccio [Wed, 12 Feb 2020 21:44:12 +0000 (13:44 -0800)]
scsi: qla2xxx: Add beacon LED config sysfs interface
This patch provides an interface to do the following (using MBC 0x3B):
- Displays (in hex) the LED config words for all three LEDs.
- Programs the config words for one LED or for all three LEDs.
The sysfs node defined is named beacon_config.
First, to allow driver to gain LED control, do this:
# echo 1 > /sys/class/scsi_host/host#/beacon
Then, to display config words for all three LEDs do this:
# cat /sys/class/scsi_host/host#/beacon_config
To set config words for all three LEDs do this:
# echo 3 xxxx yyyy zzzz > /sys/class/scsi_host/host#/beacon_config
Or, to set config word for a specific single LED n do this:
# echo n xxxx > /sys/class/scsi_host/host#/beacon_config
where n is the LED number (0, 1, 2)
Finally, to restore LED control back to firmware, do this:
# echo 0 > /sys/class/scsi_host/host#/beacon
David Disseldorp [Fri, 21 Feb 2020 09:44:01 +0000 (10:44 +0100)]
scsi: target: use an enum to track emulate_ua_intlck_ctrl
The emulate_ua_intlck_ctrl device attribute accepts values of 0, 1 or 2 via
ConfigFS, which map to unit attention interlocks control codes in the MODE
SENSE control Mode Page. Use an enum to track these values so that it's
clear that, unlike the remaining emulate_X attributes,
emulate_ua_intlck_ctrl isn't boolean.
Link: https://marc.info/?l=target-devel&m=158227825428798 Suggested-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: David Disseldorp <ddiss@suse.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
David Disseldorp [Wed, 19 Feb 2020 13:01:36 +0000 (14:01 +0100)]
scsi: target: convert boolean se_dev_attrib types to bool
This should harden us against configfs API regressions similar to the one
fixed by the previous commit.
Link: https://marc.info/?l=target-devel&m=158211731505174 Signed-off-by: David Disseldorp <ddiss@suse.de> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The LIO unmap_zeroes_data device attribute is mapped to the LBPRZ flag in
the READ CAPACITY(16) and Thin Provisioning VPD INQUIRY responses.
The unmap_zeroes_data attribute is exposed via configfs, where any write
value is correctly validated via strtobool(). However, when initialised via
target_configure_unmap_from_queue() it takes the value of the device's
max_write_zeroes_sectors queue limit, which is non-boolean.
A non-boolean value can be read from configfs, but attempting to write the
same value back results in -EINVAL, causing problems for configuration
utilities such as targetcli.
Link: https://marc.info/?l=target-devel&m=158213354011309 Fixes: 2237498f0b5c ("target/iblock: Convert WRITE_SAME to blkdev_issue_zeroout") Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: David Disseldorp <ddiss@suse.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
scsi: advansys: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language extension
to the C90 standard, but the preferred mechanism to declare variable-length
types such as these ones is a flexible array member[1][2], introduced in
C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by this
change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
James Smart [Mon, 10 Feb 2020 17:31:55 +0000 (09:31 -0800)]
scsi: lpfc: add RDF registration and Link Integrity FPIN logging
This patch modifies lpfc to register for Link Integrity events via the use
of an RDF ELS and to perform Link Integrity FPIN logging.
Specifically, the driver was modified to:
- Format and issue the RDF ELS immediately following SCR registration.
This registers the ability of the driver to receive FPIN ELS.
- Adds decoding of the FPIN els into the received descriptors, with
logging of the Link Integrity event information. After decoding, the ELS
is delivered to the scsi fc transport to be delivered to any user-space
applications.
- To aid in logging, simple helpers were added to create enum to name
string lookup functions that utilize the initialization helpers from the
fc_els.h header.
- Note: base header definitions for the ELS's don't populate the
descriptor payloads. As such, lpfc creates it's own version of the
structures, using the base definitions (mostly headers) and additionally
declaring the descriptors that will complete the population of the ELS.
James Smart [Mon, 10 Feb 2020 17:31:54 +0000 (09:31 -0800)]
scsi: fc: Update Descriptor definition and add RDF and Link Integrity FPINs
Update the FC headers for the RDF ELS and populate out the FPIN ELS and the
Link integrity FPIN payload.
RDF is used to register for diagnostic events.
FPIN is how the fabric reports a diagnostic event.
Specifically, this patch:
- Adds the formal definition of TLV descriptors that are now used in a lot
of the FC spec. The simplistic fc_fn_desc structure, basically no more
than the tlv definition, is removed.
- Small tlv helper functions are added as defines.
- The list of known Descriptor tags (identifying the TLV) is expanded and
a name initializer introduced.
- The LSRI descriptor, returned in many new ELS response payloads is
added.
- The RDF ELS code is added, and the RDF request response structures
added.
- The FPIN els definition is corrected.
- A full definition of a Link Integrity Notification descriptor is added,
[mkp: rolled in kbuild warning fix]
Link: https://lore.kernel.org/r/20200210173155.547-2-jsmart2021@gmail.com Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: James Smart <jsmart2021@gmail.com> Reviewed-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Can Guo [Tue, 11 Feb 2020 03:40:50 +0000 (19:40 -0800)]
scsi: ufs-qcom: Delay specific time before gate ref clk
After enter hibern8, as UFS JEDEC ver 3.0 requires, a specific gating wait
time is required before disable the device reference clock. If it is not
specified, use the old delay.
Can Guo [Tue, 11 Feb 2020 03:40:49 +0000 (19:40 -0800)]
scsi: ufs: Add dev ref clock gating wait time support
In UFS version 3.0, a newly added attribute bRefClkGatingWaitTime defines
the minimum time for which the reference clock is required by device during
transition to LS-MODE or HIBERN8 state. Make this change to reflect the new
requirement by adding delays before turning off the clock.
Link: https://lore.kernel.org/r/1581392451-28743-7-git-send-email-cang@codeaurora.org Reviewed-by: Asutosh Das <asutoshd@codeaurora.org> Reviewed-by: Bean Huo <beanhuo@micron.com> Reviewed-by: Stanley Chu <stanley.chu@mediatek.com> Signed-off-by: Can Guo <cang@codeaurora.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Can Guo [Tue, 11 Feb 2020 03:40:48 +0000 (19:40 -0800)]
scsi: ufs: Fix ufshcd_hold() caused scheduling while atomic
The async version of ufshcd_hold(async == true), which is only called in
queuecommand path as for now, is expected to work in atomic context, thus
it should not sleep or schedule out. When it runs into the condition that
clocks are ON but link is still in hibern8 state, it should bail out
without flushing the clock ungate work.
Fixes: f2a785ac2312 ("scsi: ufshcd: Fix race between clk scaling and ungate work") Link: https://lore.kernel.org/r/1581392451-28743-6-git-send-email-cang@codeaurora.org Reviewed-by: Hongwu Su <hongwus@codeaurora.org> Reviewed-by: Asutosh Das <asutoshd@codeaurora.org> Reviewed-by: Bean Huo <beanhuo@micron.com> Reviewed-by: Stanley Chu <stanley.chu@mediatek.com> Signed-off-by: Can Guo <cang@codeaurora.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Can Guo [Tue, 11 Feb 2020 03:40:46 +0000 (19:40 -0800)]
scsi: ufs-qcom: Adjust bus bandwidth voting and unvoting
The bus bandwidth voting is required to be done before the bus clocks are
enabled, and the unvoting is required to be done only after the bus clocks
are disabled.
Asutosh Das [Tue, 11 Feb 2020 03:40:45 +0000 (19:40 -0800)]
scsi: ufs: set load before setting voltage in regulators
This sequence change is required to avoid dips in voltage during boot-up.
Apparently, this dip is caused because in the original sequence, the
regulators are initialized in lpm mode. And then when the load is set to
high, and more current is drawn, than is allowed in lpm, the dip is seen.
Link: https://lore.kernel.org/r/1581392451-28743-3-git-send-email-cang@codeaurora.org Reviewed-by: Hongwu Su <hongwus@codeaurora.org> Reviewed-by: Bean Huo <beanhuo@micron.com> Reviewed-by: Stanley Chu <stanley.chu@mediatek.com> Signed-off-by: Asutosh Das <asutoshd@codeaurora.org> Signed-off-by: Can Guo <cang@codeaurora.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Sayali Lokhande [Tue, 11 Feb 2020 03:40:44 +0000 (19:40 -0800)]
scsi: ufs: Flush exception event before suspend
Exception event can be raised by the device when system suspend is in
progress. This will result in unclocked register access in exception event
handler as clocks will be turned off during suspend. This change makes sure
to flush exception event handler work in suspend before disabling clocks to
avoid unclocked register access issue.
Link: https://lore.kernel.org/r/1581392451-28743-2-git-send-email-cang@codeaurora.org Reviewed-by: Bean Huo <beanhuo@micron.com> Signed-off-by: Sayali Lokhande <sayalil@codeaurora.org> Signed-off-by: Asutosh Das <asutoshd@codeaurora.org> Signed-off-by: Can Guo <cang@codeaurora.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Can Guo [Wed, 12 Feb 2020 05:38:29 +0000 (21:38 -0800)]
scsi: ufs: Select INITIAL ADAPT type for HS Gear4
ADAPT is added specifically for HS Gear4 mode only. Select INITIAL ADAPT
before power mode change to G4 and select NO ADAPT before switch to non-G4
modes.
Stanley Chu [Fri, 7 Feb 2020 07:03:57 +0000 (15:03 +0800)]
scsi: ufs: introduce common function to disable host TX LCC
Many vendors would like to disable host TX LCC during initialization
flow. Introduce a common function for all users to make drivers easier to
read and maintained. This patch does not change any functionality.
Link: https://lore.kernel.org/r/20200207070357.17169-3-stanley.chu@mediatek.com Reviewed-by: Can Guo <cang@codeaurora.org> Reviewed-by: Avri Altman <avri.altman@wdc.com> Reviewed-by: Asutosh Das <asutoshd@codeaurora.org> Signed-off-by: Stanley Chu <stanley.chu@mediatek.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
MediaTek UFS host requires TX LCC to be disabled on both host and device
sides. This can be done by disabling host's local TX LCC before link
startup. Correct TX LCC disabling timing in MediaTek UFS driver.
James Smart [Tue, 28 Jan 2020 00:23:09 +0000 (16:23 -0800)]
scsi: lpfc: Clean up hba max_lun_queue_depth checks
The current code does some odd +1 over maximum xri count checks and
requires that the lun_queue_count can't be bigger than maximum xri count
divided by 8. These items are bogus.
Clean the code up to cap lun_queue_count to maximum xri count.
James Smart [Tue, 28 Jan 2020 00:23:08 +0000 (16:23 -0800)]
scsi: lpfc: Remove handler for obsolete ELS - Read Port Status (RPS)
There was report of an odd "Fix me..." log message, which was tracked down
to the lpfc_els_rcv_rps() routine. This was in handling of a very old and
obsolete ELS - Read Port Status. The RPS ELS was defined in FC-LS-1, but
deprecated in FC-LS-2, and removed from all later FC-LS revisions. It was
replaced by the Read Diagnostic Parameters (RDP) ELS and the Link Error
Status Block descriptor.
There should be no support for the RSP ELS. Remove support from driver.
James Smart [Tue, 28 Jan 2020 00:23:07 +0000 (16:23 -0800)]
scsi: lpfc: Fix coverity errors in fmdi attribute handling
Coverity reported a memory corruption error for the fdmi attributes
routines:
CID 15768 [Memory Corruption] Out-of-bounds access on FDMI
Sloppy coding of the fmdi structures. In both the lpfc_fdmi_attr_def and
lpfc_fdmi_reg_port_list structures, a field was placed at the start of
payload that may have variable content. The field was given an arbitrary
type (uint32_t). The code then uses the field name to derive an address,
which it used in things such as memset and memcpy. The memset sizes or
memcpy lengths were larger than the arbitrary type, thus coverity reported
an error.
Fix by replacing the arbitrary fields with the real field structures
describing the payload.
James Smart [Tue, 28 Jan 2020 00:23:06 +0000 (16:23 -0800)]
scsi: lpfc: Fix compiler warning on frame size
The following error is see from the compiler:
drivers/scsi/lpfc/lpfc_init.c: In function
‘lpfc_cpuhp_get_eq’: drivers/scsi/lpfc/lpfc_init.c:12660:1:
error: the frame size of 1032 bytes is larger than 1024 bytes
[-Werror=frame-larger-than=]
The issue is due to allocating a cpumask on the stack.
Fix by converting to a dynamical allocation of the cpu mask.
James Smart [Tue, 28 Jan 2020 00:23:05 +0000 (16:23 -0800)]
scsi: lpfc: Fix release of hwq to clear the eq relationship
When performing reset testing, the eq's list for related hwqs was getting
corrupted. In cases where there is not a 1:1 eq to hwq, the eq is
shared. The eq maintains a list of hwqs utilizing it in case of cpu
offlining and polling. During the reset, the hwqs are being torn down so
they can be recreated. The recreation was getting confused by seeing a
non-null eq assignment on the eq and the eq list became corrupt.
Correct by clearing the hdwq eq assignment when the hwq is cleaned up.
James Smart [Tue, 28 Jan 2020 00:23:03 +0000 (16:23 -0800)]
scsi: lpfc: Fix broken Credit Recovery after driver load
When driver is set to enable bb credit recovery, the switch displayed the
setting as inactive. If the link bounces, it switches to Active.
During link up processing, the driver currently does a MBX_READ_SPARAM
followed by a MBX_CONFIG_LINK. These mbox commands are queued to be
executed, one at a time and the completion is processed by the worker
thread. Since the MBX_READ_SPARAM is done BEFORE the MBX_CONFIG_LINK, the
BB_SC_N bit is never set the the returned values. BB Credit recovery status
only gets set after the driver requests the feature in CONFIG_LINK, which
is done after the link up. Thus the ordering of READ_SPARAM needs to follow
the CONFIG_LINK.
Fix by reordering so that READ_SPARAM is done after CONFIG_LINK. Added a
HBA_DEFER_FLOGI flag so that any FLOGI handling waits until after the
READ_SPARAM is done so that the proper BB credit value is set in the FLOGI
payload.
Fixes: 6bfb16208298 ("scsi: lpfc: Fix configuration of BB credit recovery in service parameters") Cc: <stable@vger.kernel.org> # v5.4+ Link: https://lore.kernel.org/r/20200128002312.16346-4-jsmart2021@gmail.com Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: James Smart <jsmart2021@gmail.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
James Smart [Tue, 28 Jan 2020 00:23:01 +0000 (16:23 -0800)]
scsi: lpfc: Fix RQ buffer leakage when no IOCBs available
The driver is occasionally seeing the following SLI Port error, requiring
reset and reinit:
Port Status Event: ... error 1=0x52004a01, error 2=0x218
The failure means an RQ timeout. That is, the adapter had received
asynchronous receive frames, ran out of buffer slots to place the frames,
and the driver did not replenish the buffer slots before a timeout
occurred. The driver should not be so slow in replenishing buffers that a
timeout can occur.
When the driver received all the frames of a sequence, it allocates an IOCB
to put the frames in. In a situation where there was no IOCB available for
the frame of a sequence, the RQ buffer corresponding to the first frame of
the sequence was not returned to the FW. Eventually, with enough traffic
encountering the situation, the timeout occurred.
Fix by releasing the buffer back to firmware whenever there is no IOCB for
the first frame.
Ming Lei [Sun, 19 Jan 2020 07:14:28 +0000 (15:14 +0800)]
scsi: core: remove .for_blk_mq
Not in use anymore. Remove the flag.
Cc: Sathya Prakash <sathya.prakash@broadcom.com> Cc: Chaitra P B <chaitra.basappa@broadcom.com> Cc: Suganath Prabu Subramani <suganath-prabu.subramani@broadcom.com> Cc: Kashyap Desai <kashyap.desai@broadcom.com> Cc: Sumit Saxena <sumit.saxena@broadcom.com> Cc: Shivasharan S <shivasharan.srikanteshwara@broadcom.com> Cc: Ewan D. Milne <emilne@redhat.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.de> Cc: Bart Van Assche <bart.vanassche@wdc.com> Link: https://lore.kernel.org/r/20200119071432.18558-3-ming.lei@redhat.com Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Ming Lei <ming.lei@redhat.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Stanley Chu [Wed, 29 Jan 2020 10:52:51 +0000 (18:52 +0800)]
scsi: ufs: ufs-mediatek: gate ref-clk during Auto-Hibern8
In current UFS driver design, hba->uic_link_state will not be changed after
link enters Hibern8 state by Auto-Hibern8 mechanism. In this case,
reference clock gating will be skipped unless special handling is
implemented in vendor's callbacks.
Support reference clock gating during Auto-Hibern8 period in MediaTek
Chipsets: If link state is already in Hibern8 while Auto-Hibern8 feature is
enabled, gate reference clock in setup_clocks callback.
Stanley Chu [Wed, 29 Jan 2020 10:52:50 +0000 (18:52 +0800)]
scsi: ufs: fix Auto-Hibern8 error detection
Auto-Hibern8 may be disabled by some vendors or sysfs in runtime even if
Auto-Hibern8 capability is supported by host. If Auto-Hibern8 capability is
supported by host but not actually enabled, Auto-Hibern8 error shall not
happen.
To fix this, provide a way to detect if Auto-Hibern8 is actually enabled
first, and bypass Auto-Hibern8 disabling case in
ufshcd_is_auto_hibern8_error().
Fixes: 821744403913 ("scsi: ufs: Add error-handling of Auto-Hibernate") Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20200129105251.12466-4-stanley.chu@mediatek.com Reviewed-by: Bean Huo <beanhuo@micron.com> Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com> Reviewed-by: Asutosh Das <asutoshd@codeaurora.org> Reviewed-by: Can Guo <cang@codeaurora.org> Signed-off-by: Stanley Chu <stanley.chu@mediatek.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Stanley Chu [Wed, 29 Jan 2020 10:52:49 +0000 (18:52 +0800)]
scsi: ufs: ufs-mediatek: support linkoff state during suspend
If system suspend or runtime suspend mode is configured as linkoff state,
phy can be powered off and reference clock can be gated in MediaTek
Chipsets.
At the same time, remove redundant reference clock control in suspend and
resume callbacks because such control can be well-handled in setup_clocks
callback.