HiP08 RoCE hardware lacks ability(a known hardware problem) to flush
outstanding WQEs if QP state gets into errored mode for some reason.  To
overcome this hardware problem and as a workaround, when QP is detected to
be in errored state during various legs like post send, post receive
etc[1], flush needs to be performed from the driver.

The earlier patch[1] sent to solve the hardware limitation explained in
the cover-letter had a bug in the software flushing leg. It acquired mutex
while modifying QP state to errored state and while conveying it to the
hardware using the mailbox. This caused leg to sleep while holding
spin-lock and caused crash.

Suggested Solution: we have proposed to defer the flushing of the QP in
the Errored state using the workqueue to get around with the limitation of
our hardware.

This patch specifically adds the calls to the flush handler from where
parts of the code like post_send/post_recv etc. when the QP state gets
into the errored mode.

[1] https://patchwork.kernel.org/patch/10534271/

Link: https://lore.kernel.org/r/1580983005-13899-3-git-send-email-liuyixian@huawei.comSigned-off-by: NYixian Liu <liuyixian@huawei.com>
Reviewed-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

HiP08 RoCE hardware lacks ability(a known hardware problem) to flush
outstanding WQEs if QP state gets into errored mode for some reason.  To
overcome this hardware problem and as a workaround, when QP is detected to
be in errored state during various legs like post send, post receive etc
[1], flush needs to be performed from the driver.

The earlier patch[1] sent to solve the hardware limitation explained in
the cover-letter had a bug in the software flushing leg. It acquired mutex
while modifying QP state to errored state and while conveying it to the
hardware using the mailbox. This caused leg to sleep while holding
spin-lock and caused crash.

Suggested Solution:
we have proposed to defer the flushing of the QP in the Errored state
using the workqueue to get around with the limitation of our hardware.

This patch adds the framework of the workqueue and the flush handler
function.

[1] https://patchwork.kernel.org/patch/10534271/

Link: https://lore.kernel.org/r/1580983005-13899-2-git-send-email-liuyixian@huawei.comSigned-off-by: NYixian Liu <liuyixian@huawei.com>
Reviewed-by: NSalil Mehta <salil.mehta@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Some magic numbers are hard to understand, so replace them with macros or
add some comments for them.

Link: https://lore.kernel.org/r/20200126145504.9700-1-liweihang@huawei.comSigned-off-by: NLang Cheng <chenglang@huawei.com>
Signed-off-by: NYixian Liu <liuyixian@huawei.com>
Signed-off-by: NWenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: NYixing Liu <liuyixing1@huawei.com>
Signed-off-by: NWeihang Li <liweihang@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

So far the assumption was that ib_umem_get() and ib_umem_odp_get()
are called from flows that start in UVERBS and therefore has a user
context. This assumption restricts flows that are initiated by ULPs
and need the service that ib_umem_get() provides.

This patch changes ib_umem_get() and ib_umem_odp_get() to get IB device
directly by relying on the fact that both UVERBS and ULPs sets that
field correctly.
Reviewed-by: NGuy Levi <guyle@mellanox.com>
Signed-off-by: NMoni Shoua <monis@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>

When hardware is in resetting stage, we may can't poll back all the
expected work completions as the hardware won't generate cqe anymore.

This patch allows the driver to compose the expected wc instead of the
hardware during resetting stage. Once the hardware finished resetting, we
can poll cq from hardware again.

Link: https://lore.kernel.org/r/1578572412-25756-1-git-send-email-liweihang@huawei.comSigned-off-by: NXi Wang <wangxi11@huawei.com>
Signed-off-by: NWeihang Li <liweihang@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

HNS_ROCE_ALIGN_UP can be replaced by round_up() which is defined in
kernel.h.

Link: https://lore.kernel.org/r/1578313276-29080-7-git-send-email-liweihang@huawei.comSigned-off-by: NWenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: NWeihang Li <liweihang@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

The argument is always ignored, so remove it.

Link: https://lore.kernel.org/r/20191113073214.9514-3-hch@lst.deSigned-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NJason Gunthorpe <jgg@mellanox.com>
Acked-by: NMichal Kalderon <michal.kalderon@marvell.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Modify some printings that is not in uniformed style, non-standard or with
spelling errors.

Link: https://lore.kernel.org/r/1572952082-6681-10-git-send-email-liweihang@hisilicon.comSigned-off-by: NWenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: NWeihang Li <liweihang@hisilicon.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Special QP have no differences with normal qp in data structure, so
definition of struct hns_roce_sqp should be removed and replaced by struct
hns_roce_qp.

Link: https://lore.kernel.org/r/1572952082-6681-3-git-send-email-liweihang@hisilicon.comSigned-off-by: NLang Cheng <chenglang@huawei.com>
Signed-off-by: NWeihang Li <liweihang@hisilicon.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

There is no need to define max_post in hns_roce_wq, as it does same thing
as wqe_cnt.

Link: https://lore.kernel.org/r/1572952082-6681-2-git-send-email-liweihang@hisilicon.comSigned-off-by: NYixian Liu <liuyixian@huawei.com>
Signed-off-by: NWeihang Li <liweihang@hisilicon.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

The "ucmd->log_sq_bb_count" variable is a user controlled variable in the
0-255 range.  If we shift more than then number of bits in an int then
it's undefined behavior (it shift wraps), and potentially the int could
become negative.

Fixes: 9a443537 ("IB/hns: Add driver files for hns RoCE driver")
Link: https://lore.kernel.org/r/20190608092514.GC28890@mwandaReported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Reviewed-by: NDan Carpenter <dan.carpenter@oracle.com>

HNS_ROCE_ALOGN_UP should be HNS_ROCE_ALIGN_UP, this patch fix it.
Signed-off-by: NYixing Liu <liuyixing1@huawei.com>
Signed-off-by: NWeihang Li <liweihang@hisilicon.com>
Link: https://lore.kernel.org/r/1567566885-23088-6-git-send-email-liweihang@hisilicon.comSigned-off-by: NDoug Ledford <dledford@redhat.com>

Here packages the codes of allocating and freeing rq inline buffer in
hns_roce_create_qp_common function in order to reduce the complexity.

Link: https://lore.kernel.org/r/1567068102-56919-3-git-send-email-liweihang@hisilicon.comSigned-off-by: NLijun Ou <oulijun@huawei.com>
Signed-off-by: NWeihang Li <liweihang@hisilicon.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Sparse is whining about the u32 and __le32 mixed usage in the driver.
The roce_set_field() is used to __le32 data of hardware only.
If a variable is not delivered to the hardware, the __le32 type and
related operations are not required.
Signed-off-by: NLang Cheng <chenglang@huawei.com>
Link: https://lore.kernel.org/r/1566393276-42555-6-git-send-email-oulijun@huawei.comSigned-off-by: NDoug Ledford <dledford@redhat.com>

This line was indented a bit too far.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Link: https://lore.kernel.org/r/20190816113907.GA30799@mwandaSigned-off-by: NDoug Ledford <dledford@redhat.com>

When create a qp and attached to srq, rq will no longer be used
and the members of rq will be set zero. As a result, the wrid
of rq will not be allocated and used.

Fixes: 926a01dc ("RDMA/hns: Add QP operations support for hip08 SoC")
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Link: https://lore.kernel.org/r/1565343666-73193-3-git-send-email-oulijun@huawei.comSigned-off-by: NDoug Ledford <dledford@redhat.com>

Here uses the new APIs instead of some dev print interfaces in
some functions.
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Link: https://lore.kernel.org/r/1565276034-97329-15-git-send-email-oulijun@huawei.comSigned-off-by: NDoug Ledford <dledford@redhat.com>

Current prompt message is uncorrect when destroying qp, add qpn
information when creating qp.
Signed-off-by: NYixian Liu <liuyixian@huawei.com>
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Link: https://lore.kernel.org/r/1565276034-97329-4-git-send-email-oulijun@huawei.comSigned-off-by: NDoug Ledford <dledford@redhat.com>

Here mainly packages some code into some new functions in order to
reduce code compelexity.
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Link: https://lore.kernel.org/r/1565276034-97329-3-git-send-email-oulijun@huawei.comSigned-off-by: NDoug Ledford <dledford@redhat.com>

It needs to check the sq size with integrity when configures
the relatived parameters of sq. Here moves the relatived code
into a special function.
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Link: https://lore.kernel.org/r/1565276034-97329-2-git-send-email-oulijun@huawei.comSigned-off-by: NDoug Ledford <dledford@redhat.com>

Moves the related lines of setting extended sge size into a separate
function as well as remove the unused variables.

Link: https://lore.kernel.org/r/1562593285-8037-5-git-send-email-oulijun@huawei.comSigned-off-by: NLijun Ou <oulijun@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

When smmu is enable, if execute the perftest command and then use 'kill
-9' to exit, follow this operation repeatedly, the kernel will have a high
probability to print the following smmu event:

  arm-smmu-v3 arm-smmu-v3.1.auto: event 0x10 received:
  arm-smmu-v3 arm-smmu-v3.1.auto:  0x00007d0000000010
  arm-smmu-v3 arm-smmu-v3.1.auto:  0x0000020900000080
  arm-smmu-v3 arm-smmu-v3.1.auto:  0x00000000f47cf000
  arm-smmu-v3 arm-smmu-v3.1.auto:  0x00000000f47cf000

This is because the hw will periodically refresh the qpc cache until the
next reset.

This patch fixed it by removing the action that release qpc memory in the
'hns_roce_qp_free' function.

Fixes: 9a443537 ("IB/hns: Add driver files for hns RoCE driver")
Signed-off-by: NXi Wang <wangxi11@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

The buffer size of qp which used to allocate qp buffer space for storing
sqwqe and rqwqe will be the length of buffer space. The kernel driver will
use the buffer address and the same size to get the user memory. The same
size named buff_size of qp. According the algorithm of calculating, The
size of the two is not equal when users set the max sge of sq.

Fixes: b28ca7cc ("RDMA/hns: Limit extend sq sge num")
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Hip06 reserve 12 qps, Hip08 reserve 8 qps. When the QP is released, the
chip model is not judged, and the Hip08 cannot release the qpn 8~12
Signed-off-by: NLang Cheng <chenglang@huawei.com>
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

There is a spelling mistake in an dev_err message. Fix it.
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

This patch removes the hns-roce.ko for cleanup all the exported symbols in
common part.
Signed-off-by: NXi Wang <wangxi11@huawei.com>
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Update ib_umem_release() to behave similarly to kfree() and allow
submitting NULL pointer as safe input to this function.

Fixes: a52c8e24 ("RDMA: Clean destroy CQ in drivers do not return errors")
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

hip08 can support up to 32768 wqes in one qp. currently if the wqe num
is larger than 16384, the driver will lead a calltrace as follows.

[21361.393725] Call trace:
[21361.398605]  hns_roce_v2_modify_qp+0xbcc/0x1360 [hns_roce_hw_v2]
[21361.410627]  hns_roce_modify_qp+0x1d8/0x2f8 [hns_roce]
[21361.420906]  _ib_modify_qp+0x70/0x118
[21361.428222]  ib_modify_qp+0x14/0x1c
[21361.435193]  rt_ktest_modify_qp+0xb8/0x650 [rdma_test]
[21361.445472]  exec_modify_qp_cmd+0x110/0x4d8 [rdma_test]
[21361.455924]  rt_ktest_dispatch_cmd_3+0xa94/0x2edc [rdma_test]
[21361.467422]  rt_ktest_dispatch_cmd_2+0x9c/0x108 [rdma_test]
[21361.478570]  rt_ktest_dispatch_cmd+0x138/0x904 [rdma_test]
[21361.489545]  rt_ktest_dev_write+0x328/0x4b0 [rdma_test]
[21361.499998]  __vfs_write+0x38/0x15c
[21361.506966]  vfs_write+0xa8/0x1a0
[21361.513586]  ksys_write+0x50/0xb0
[21361.520206]  sys_write+0xc/0x14
[21361.526479]  el0_svc_naked+0x30/0x34
[21361.533622] Code: 1ac10841 d37d7c22 0b000021 d37df021 (f86268c0)
[21361.545815] ---[ end trace e2a1feb2c3d7f13c ]---

When the wqe num is larger than 16384, hns_roce_table_find will return an
invalid mtt, this will lead an kernel paging requet error if the driver try
to access it. It's the mtt design defect which can't support up to the max
wqe num of hip08.

This patch fixs it by replacing mtt with mtr for wqe.

Fixes: 926a01dc ("RDMA/hns: Add QP operations support for hip08 SoC")
Signed-off-by: NXi Wang <wangxi11@huawei.com>
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

When the maximum send wr delivered by the user is zero, the qp does not
have a sq.

When allocating the sq db buffer to store the user sq pi pointer and map
it to the kernel mode, max_send_wr is used as the trigger condition, while
the kernel does not consider the max_send_wr trigger condition when
mapmping db. It will cause sq record doorbell map fail and create qp fail.

The failed print information as follows:

 hns3 0000:7d:00.1: Send cmd: tail - 418, opcode - 0x8504, flag - 0x0011, retval - 0x0000
 hns3 0000:7d:00.1: Send cmd: 0xe59dc000 0x00000000 0x00000000 0x00000000 0x00000116 0x0000ffff
 hns3 0000:7d:00.1: sq record doorbell map failed!
 hns3 0000:7d:00.1: Create RC QP failed

Fixes: 0425e3e6 ("RDMA/hns: Support flush cqe for hip08 in kernel space")
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

The method of hem free for SCC context is different from qp context.

In the current version, if free SCC hem during the execution of qp free,
there may be smmu error as below:

 arm-smmu-v3 arm-smmu-v3.1.auto: event 0x10 received:
 arm-smmu-v3 arm-smmu-v3.1.auto:  0x00007d0000000010
 arm-smmu-v3 arm-smmu-v3.1.auto:  0x000012000000017c
 arm-smmu-v3 arm-smmu-v3.1.auto:  0x00000000000009e0
 arm-smmu-v3 arm-smmu-v3.1.auto:  0x0000000000000000

As SCC context is still used by hardware after qp free, we can solve this
problem by removing SCC hem free from hns_roce_qp_free.

Fixes: 6a157f7d ("RDMA/hns: Add SCC context allocation support for hip08")
Signed-off-by: NYangyang Li <liyangyang20@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Also fully initialise the qp before storing it in the XArray.
Signed-off-by: NMatthew Wilcox <willy@infradead.org>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Now when we have the udata passed to all the ib_xxx object creation APIs
and the additional macro 'rdma_udata_to_drv_context' to get the
ib_ucontext from ib_udata stored in uverbs_attr_bundle, we can finally
start to remove the dependency of the drivers in the
ib_xxx->uobject->context.
Signed-off-by: NShamir Rabinovitch <shamir.rabinovitch@oracle.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

The memory allocated for wrid should be initialized to zero.
Signed-off-by: NYixian Liu <liuyixian@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Use the for_each_sg_dma_page iterator variant to walk the umem DMA-mapped
SGL and get the page DMA address. This avoids the extra loop to iterate
pages in the SGE when for_each_sg iterator is used.

Additionally, purge umem->page_shift usage in the driver as its only
relevant for ODP MRs. Use system page size and shift instead.
Signed-off-by: NShiraz, Saleem <shiraz.saleem@intel.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

This patch adds SCC context clear support for DCQCN in kernel space
driver.
Signed-off-by: NYangyang Li <liyangyang20@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

This patch adds SCC context allocation and initialization support for
DCQCN in kernel space driver.
Signed-off-by: NYangyang Li <liyangyang20@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

When flush cqe, it needs to get the pointer of rq and sq from db address
space of user and update it into qp context by modified qp. if rq does not
exist, it will not get the value from db address space of user.
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

ib_umem_get() can only be called in a method callback, which always has a
udata parameter. This allows ib_umem_get() to derive the ucontext pointer
directly from the udata without requiring the drivers to find it in some
way or another.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Signed-off-by: NShamir Rabinovitch <shamir.rabinovitch@oracle.com>

In some application scenario, the user could not have receive queue when
run rdma write or read operation.
Signed-off-by: NLijun Ou <oulijun@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Drivers should be using udata to determine if a method is invoked from
user space or kernel space. A pd does not necessarily say a different
objects is kernel or user.

Transforming the tests to use udata eliminates a large number of uobject
references from the drivers.
Signed-off-by: NShamir Rabinovitch <shamir.rabinovitch@oracle.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>