There is much common code and functionality between the HW versions to set
the PHY linkrate.

As such, this patch factors out the common code into a generic function
hisi_sas_phy_set_linkrate().
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Currently we lock the DQ to protect whole delivery process.  So this
stops us building slots for the same queue in parallel, and can affect
performance.

To optimise it, only lock the DQ during special periods, specifically
when allocating a slot from the DQ and when delivering a slot to the HW.

This approach is now safe, thanks to the previous patches to ensure that
we always deliver a slot to the HW once allocated.
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Since the task prep functions now should not fail, adjust the return
types to void.

In addition, some checks in the task prep functions are relocated to the
main module; this is specifically the check for the number of elements
in an sg list exceeded the HW SGE limit.
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Currently we use DQ lock to protect delivery of DQ entry one by one.

To optimise to allow more than one slot to be built for a single DQ in
parallel, we need to remove the DQ lock when preparing slots, prior to
delivery.

To achieve this, we rearrange the slot build order to ensure that once
we allocate a slot for a task, we do cannot fail to deliver the task.

In this patch, we rearrange the slot building for SMP tasks to ensure
that sg mapping part (which can fail) happens before we allocate the
slot in the DQ.
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

The patch does some code cleanup and fixes some small bugs:

- Correct return status of phy_up_v3_hw() and phy_bcast_v3_hw()
- Add static for function phy_get_max_linkrate_v3_hw()
- Change exception return status when no reset method
- Change magic value to ts->stat in slot_complete_vx_hw()
- Remove unnecessary check for dev_is_sata()
- Fix some issues of alignment and indents (Authored by Xiaofei Tan in
  another patch, but added here to be practical)
Signed-off-by: NXiaofei Tan <tanxiaofei@huawei.com>
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

The structure element hisi_sas_devices.running_req to count how many
commands are active is in effect only ever written in the code, so remove
it.
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

It is not right to set the register PROG_PHY_LINK_RATE while PHY is still
enabled. So if we want to change PHY linkrate, we need to disable PHY before
setting the register PROG_PHY_LINK_RATE, and then start-up PHY. This patch
is to fix this issue.
Signed-off-by: NXiaofei Tan <tanxiaofei@huawei.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

In sysfs, there are two files about minimum linkrate, and also two files for
maximum linkrate. Take maximum linkrate example, maximum_linkrate_hw is
read-only and indicated by the register HARD_PHY_LINKRATE, and
maximum_linkrate is read-write and corresponding to the register
PROG_PHY_LINK_RATE.

But in the function phy_up_v*_hw(), we get *_linkrate value from
HARD_PHY_LINKRATE. It is not right. This patch is to fix this issue.

Unreferenced PHY-interrupt enum is also removed for v3 hw.
Signed-off-by: NXiaofei Tan <tanxiaofei@huawei.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Use an general way to do delay work for a PHY. Then it will be easier to add
new delayed work for a PHY in future.
Signed-off-by: NXiaofei Tan <tanxiaofei@huawei.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

In certain scenarios we may just want to clear the ITCT for a device, and not
free other resources like the SATA bitmap using in v2 hw.

To facilitate this, this patch relocates the code of clearing ITCT from
free_device() to a new hw interface clear_itct().  Then for some hw, we should
not realise free_device() if there's nothing left to do for it.

[mkp: typo]
Signed-off-by: NXiaofei Tan <tanxiaofei@huawei.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

In preparation for unconditionally passing the struct timer_list pointer to
all timer callbacks, switch to using the new timer_setup() and from_timer()
to pass the timer pointer explicitly. This requires adding a pointer to
hold the timer's target task, as there isn't a link back from slow_task.

Cc: John Garry <john.garry@huawei.com>
Cc: "James E.J. Bottomley" <jejb@linux.vnet.ibm.com>
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>
Cc: Jack Wang <jinpu.wang@profitbricks.com>
Cc: lindar_liu@usish.com
Cc: Jens Axboe <axboe@fb.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Benjamin Block <bblock@linux.vnet.ibm.com>
Cc: Baoyou Xie <baoyou.xie@linaro.org>
Cc: Wei Yongjun <weiyongjun1@huawei.com>
Cc: linux-scsi@vger.kernel.org
Signed-off-by: NKees Cook <keescook@chromium.org>
Reviewed-by: NMartin K. Petersen <martin.petersen@oracle.com>
Acked-by: John Garry <john.garry@huawei.com> # for hisi_sas part
Tested-by: John Garry <john.garry@huawei.com> # basic sanity test for hisi_sas
Reviewed-by: NJack Wang <jinpu.wang@profitbricks.com>

When a PHY_FUNC_LINK_RESET is issued, we need to fill the transport
identify_frame to SAS controller before the PHYs are enabled.

Without this, we may find that if a PHY which belonged to a wideport
before the reset may generate a new port id.
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Currently we allocate 3 sets of DMA memories from separate pools for
each slot. This is inefficient in terms of memory usage
(buffers are less than 1 page in size, so we lose due to alignment),
and also time spent in doing 3 allocations + de-allocations per slot,
instead of 1.

To optimise, combine the 3 DMA buffers into a single buffer from a
single pool.
Signed-off-by: NXiaofei Tan <tanxiaofei@huawei.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Since hip08 SAS controller is based on pci device, add hisi_hba.pci_dev
for hip08 (will be v3), and also rename hisi_hba.pdev to .platform_dev
for clarity.

In addition, for common code which wants to reference the controller
device struct, add hisi_hba.dev, and change the common code to use it.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Currently hisi_hba.lock is locked to deliver and receive a command
to/from any hw queue. This causes much contention at high data-rates.

To boost performance, lock on a per queue basis for sending and
receiving commands to/from hw.

Certain critical regions still need to be locked in the delivery and
completion stages with hisi_hba.lock.

New element hisi_sas_device.dq is added to store the delivery queue for
a device, so it does not need to be needlessly re-calculated for every
task.
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Some more locking needs to be added/modified for when
read-modify-writing sas_task.task_state_flags.

Note: since we can attempt to grab this lock in interrupt
      context we should use irq variant of spin_lock.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Currently when a root PHY is deformed from a asd_sas_port we try to
release the slots in the LLDD, and fail.

Regardless, it is not right to release this early.

This patch removes the deformed function. As it was before, port
deformation is still done in hisi_sas_phy_down().

It would be nice to actually remove the hisi_sas_port_{de}formed() pair,
however we cannot as we need to know the asd_sas_port index libsas has
associated with an asd_sas_phy.

The hw does actually generate a port id for a PHY, but this seems to a
random number, so ignored for this purpose.

This patch also changes the code to link slots to the hisi_sas_device,
and not hisi_sas_port.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Relocate the PHY init code from LLDD hw init path to
hisi_sas_scan_start().
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Introduce function to get hisi_sas_port from asd_sas_port.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

There is a bug in the current driver in that certain hisi_hba and port
structure elements which we access when servicing the CQ interrupt do
not use thread-safe accesses; these include hisi_sas_port linked-list of
active slots (hisi_sas_port.entry), bitmap of currently allocated IPTT
(in hisi_hba.slot_index_tags), and completion queue read pointer.

As a solution, lock these elements with the hisi_hba.lock.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NXiang Chen <chenxiang66@hisilicon.com>
Reviewed-by: NZhangfei Gao <zhangfei.gao@linaro.org>
Tested-by: NHanjun Guo <hanjun.guo@linaro.org>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Add the function to set PHY min and max linkrate through
sysfs interface.
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NZhangfei Gao <zhangfei.gao@linaro.org>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Sometimes the value of hisi_sas_device.running_req
would go negative unless we have the check for
running_req >= 0 before trying to decrement.

This is because using running_req is not thread-safe.

As such, the value for running_req may be actually incorrect,
so use atomic64_t instead.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NXiang Chen <chenxiang66@hisilicon.com>
Reviewed-by: NZhangfei Gao <zhangfei.gao@linaro.org>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Currently slots are allocated from queues in a round-robin fashion.
This causes a problem for internal commands in device mode. For this
mode, we should ensure that the internal abort command is the last
command seen in the host for that device. We can only ensure this when
we place the internal abort command after the preceding commands for
device that in the same queue, as there is no order in which the host
will select a queue to execute the next command.

This queue restriction makes supporting scsi mq more tricky in
the future, but should not be a blocker.

Note: Even though v1 hw does not support internal abort, the
      allocation method is chosen to be the same for consistency.
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NZhangfei Gao <zhangfei.gao@linaro.org>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Function config_id_frame_v1_hw() is called twice for each PHY during
initialisation, which is unneeded.

So remove init_id_frame_v1_hw(), which only calls
config_id_frame_v1_hw().

We will keep the call to config_id_frame_v1_hw() in start_phy_v1_hw()
since it will be used for PHY reset functions.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

The endianness for the SAS address in the TX_ID_DWORD registers is set
incorrectly.  We see errors like this in the boot log for v2 hw (which
would have the same issue as v1 hw):

[    7.583284] sas: target proto 0x0 at 50000d1108e7923f:0x1f not handled

This is due to the host SAS addr not matching the PHY SAS addr in the
expander host-attached phy discovery responses.

To fix, we byte swap the SAS addr from BE to LE (which is the endianness
of the SAS controller).
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NXiang Chen <chenxiang66@hisilicon.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Optimise by saving an avoidable read in the get_free_slot function.  The
delivery queue write pointer will only be updated by software, so don't
bother re-reading what was already written in the previous call to
start_delivery function.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Optimise by saving an avoidable read in the cq interrupt.  The queue
read pointer will only be updated by software, so don't bother
re-reading what was already written in the previous interrupt.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

When TRANS_TX_CREDIT_TIMEOUT_ERR or
TRANS_TX_CLOSE_NORMAL_ERR error occur in
a slot, the command should be re-attempted.
This error is equivalent to meaning that the queue
is full in the sdev (and not the host).
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Add support in v1 hw driver for ACPI.

A check on whether an ACPI handle is available for the device is used to
decide on whether to use ACPI reset handler or syscon for hw reset.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NZhangfei Gao <zhangfei.gao@linaro.org>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Since the error record structure is different for v2 hw, make
hisi_sas_err_record opaque and add hisi_sas_err_record_v1.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Since v2 hardware permits different numbers of commands to v1, set this
as configurable in hisi_sas_hw.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Completion header bit CMPLT_HDR_RSPNS_XFRD flags whether the response
frame is received into host memory, and not whether the response frame
has an error.  As such, change the decision on whether a slot has an
error.  Also redundant check on CMPLT_HDR_CMD_CMPLT_MSK is removed.

Fixes: 27a3f229 ("hisi_sas: Add cq interrupt handler")
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Tested-by: NRicardo Salveti <ricardo.salveti@linaro.org>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

By reading in itct.qw0 into a 32b variable the top 32 bits were being
lost.  In practice this was OK as they were zeroes.

Fixes: 27a3f229 ("hisi_sas: Add cq interrupt")
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NMatthew R. Ochs <mrochs@linux.vnet.ibm.com>
Reviewed-by: NShane Seymour <shane.seymour@hpe.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

We were doing a arithmetic comparison instead of logical shift by
accident.  Mis-programming the itct did not seem to make a difference to
operation.

Fixes: abda97c2 ("hisi_sas: Add dev_found")
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NMatthew R. Ochs <mrochs@linux.vnet.ibm.com>
Reviewed-by: NShane Seymour <shane.seymour@hpe.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

The mask fields are for quad-words, so add ULL suffix.  Also
unreferenced ITCT_HDR_BREAK_REPLY and ITCT_HDR_MAX_BURST are removed.

Fixes: 50af155b ("hisi_sas: Add v1 hardware reg")
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NShane Seymour <shane.seymour@hpe.com>
Reviewed-by: NMatthew R. Ochs <mrochs@linux.vnet.ibm.com>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

It is preferred that drivers use platform_get_irq() instead of
irq_of_parse_and_map(), so replace.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Acked-by: NRob Herring <robh@kernel.org>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Originally the driver would use of_irq_count to calculate how much
memory is required for storing the interrupt names, since the number of
interrupt sources for the controller is variable.  Since of_irq_count
cannot be used by the driver, use fixed names.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Signed-off-by: NZhangfei Gao <zhangfei.gao@linaro.org>
Acked-by: NRob Herring <robh@kernel.org>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Add handlers for fatal interrupts.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NArnd Bergmann <arnd@arndb.de>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Add method for lldd_control_phy. Currently link rate control and spinup
hold is unsupported.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NArnd Bergmann <arnd@arndb.de>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Add support for smp function, which allows devices attached by expander
to be controlled.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Reviewed-by: NArnd Bergmann <arnd@arndb.de>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>