Like any other verbs objects, CQ shouldn't fail during destroy, but
mlx5_ib didn't follow this contract with mixed IB verbs objects with
DEVX. Such mix causes to the situation where FW and kernel are fully
interdependent on the reference counting of each side.
Kernel verbs and drivers that don't have DEVX flows shouldn't fail.
Fixes: e39afe3d6d ("RDMA: Convert CQ allocations to be under core responsibility")
Link: https://lore.kernel.org/r/20200907120921.476363-7-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
In similar way to other IB objects, restore the ability to return error on
SRQ destroy. Strictly speaking, this change is not necessary, and provided
here to ensure a symmetrical interface like other destroy functions.
Fixes: 68e326dea1 ("RDMA: Handle SRQ allocations by IB/core")
Link: https://lore.kernel.org/r/20200907120921.476363-5-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
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]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 7649773293 ("cxgb3/l2t: Fix undefined behaviour")
Link: https://lore.kernel.org/r/20200213010425.GA13068@embeddedor.com
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> # added a few more
Like all other destroy commands, .destroy_cq() call is not supposed
to fail. In all flows, the attempt to return earlier caused to memory
leaks.
This patch converts .destroy_cq() to do not return any errors.
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Acked-by: Gal Pressman <galpress@amazon.com>
Acked-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Add await in destroy_qp() so that all references to qp are dereferenced
and qp is freed in destroy_qp() itself. This ensures freeing of all QPs
before invocation of dealloc_ucontext(), which prevents loss of in use
qpids stored in the ucontext.
Signed-off-by: Nirranjan Kirubaharan <nirranjan@chelsio.com>
Reviewed-by: Potnuri Bharat Teja <bharat@chelsio.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Convert SRQ allocation from drivers to be in the IB/core
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Now when ib_udata is passed to all the driver's object create/destroy APIs
the ib_udata will carry the ib_ucontext for every user command. There is
no need to also pass the ib_ucontext via the functions prototypes.
Make ib_udata the only argument psssed.
Signed-off-by: Shamir Rabinovitch <shamir.rabinovitch@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
The uverbs_attr_bundle with the ucontext is sent down to the drivers ib_x
destroy path as ib_udata. The next patch will use the ib_udata to free the
drivers destroy path from the dependency in 'uobject->context' as we
already did for the create path.
Signed-off-by: Shamir Rabinovitch <shamir.rabinovitch@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Ucontext allocation and release aren't async events and don't need kref
accounting. The common layer of RDMA subsystem ensures that dealloc
ucontext will be called after all other objects are released.
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Tested-by: Raju Rangoju <rajur@chelsio.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
If TP fetches an SRQ buffer but ends up not using it before the connection
is aborted, then it passes the index of that SRQ buffer to the host in
ABORT_REQ_RSS or ABORT_RPL CPL message.
But, if the srqidx field is zero in the received ABORT_RPL or
ABORT_REQ_RSS CPL, then we need to read the tcb.rq_start field to see if
it really did have an RQE cached. This works around a case where HW does
not include the srqidx in the ABORT_RPL/ABORT_REQ_RSS CPL.
The final value of rq_start is the one present in TCB with the
TF_RX_PDU_OUT bit cleared. So, we need to read the TCB, examine the
TF_RX_PDU_OUT (bit 49 of t_flags) in order to determine if there's a rx
PDU feedback event pending.
Signed-off-by: Raju Rangoju <rajur@chelsio.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
This will allow FW to not send more data to TP (which would then need to
be buffered). Pass the negotiated TCP window scale to FW in the FLOWC WR.
Also refactor send_flowc() a bit to clean it up.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Potnuri Bharat Teja <bharat@chelsio.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
The rdma core is taking care of return the right error code when the
rdma device callbacks aren't supported.
Signed-off-by: Kamal Heib <kamalheib1@gmail.com>
Acked-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Since neither ib_post_send() nor ib_post_recv() modify the data structure
their second argument points at, declare that argument const. This change
makes it necessary to declare the 'bad_wr' argument const too and also to
modify all ULPs that call ib_post_send(), ib_post_recv() or
ib_post_srq_recv(). This patch does not change any functionality but makes
it possible for the compiler to verify whether the
ib_post_(send|recv|srq_recv) really do not modify the posted work request.
To make this possible, only one cast had to be introduce that casts away
constness, namely in rpcrdma_post_recvs(). The only way I can think of to
avoid that cast is to introduce an additional loop in that function or to
change the data type of bad_wr from struct ib_recv_wr ** into int
(an index that refers to an element in the work request list). However,
both approaches would require even more extensive changes than this
patch.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
This patch implements the srq specific verbs such as create/destroy/modify
and post_srq_recv. And adds srq specific structures and defines to t4.h
and uapi.
Also updates the cq poll logic to deal with completions that are
associated with the SRQ's.
This patch also handles kernel mode SRQ_LIMIT events as well as flushed
SRQ buffers
Signed-off-by: Raju Rangoju <rajur@chelsio.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
This patch adds kernel mode t4_srq structures and support functions,
uapi structures and defines, as well as firmware work request structures.
Signed-off-by: Raju Rangoju <rajur@chelsio.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
This patch adds support for iw_cxb4 to extend cqes from existing 32Byte
size to 64Byte.
Also includes adds backward compatibility support (for 32Byte) to work
with older libraries.
Signed-off-by: Raju Rangoju <rajur@chelsio.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Several items of conflict have arisen between the RDMA stack's for-rc
branch and upcoming for-next work:
9fd4350ba8 ("IB/rxe: avoid double kfree_skb") directly conflicts with
2e47350789 ("IB/rxe: optimize the function duplicate_request")
Patches already submitted by Intel for the hfi1 driver will fail to
apply cleanly without this merge
Other people on the mailing list have notified that their upcoming
patches also fail to apply cleanly without this merge
Signed-off-by: Doug Ledford <dledford@redhat.com>
Provide a cxgb4-specific function to fill in qp state details.
This allows dumping important c4iw_qp state useful for debugging.
Included in the dump are the t4_sq, t4_rq structs, plus a dump
of the t4_swsqe and t4swrqe descriptors for the first and last
pending entries.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
When a CQ is shared by multiple QPs, c4iw_flush_hw_cq() needs to acquire
corresponding QP lock before moving the CQEs into its corresponding SW
queue and accessing the SQ contents for completing a WR.
Ignore CQEs if corresponding QP is already flushed.
Cc: stable@vger.kernel.org
Signed-off-by: Potnuri Bharat Teja <bharat@chelsio.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
The c4iw_rdev_close() logic was not releasing all the hw
resources (PBL and RQT memory) during the device removal
event (driver unload / system reboot). This can cause panic
in gen_pool_destroy().
The module remove function will wait for all the hw
resources to be released during the device removal event.
Fixes c12a67fe(iw_cxgb4: free EQ queue memory on last deref)
Signed-off-by: Raju Rangoju <rajur@chelsio.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Cc: stable@vger.kernel.org
Signed-off-by: Doug Ledford <dledford@redhat.com>
Patches for 4.16 that are dependent on patches sent to 4.15-rc.
These are small clean ups for the vmw_pvrdma and i40iw drivers.
* 'from-rc' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma.git:
RDMA/vmw_pvrdma: Remove usage of BIT() from UAPI header
RDMA/vmw_pvrdma: Use refcount_t instead of atomic_t
RDMA/vmw_pvrdma: Use more specific sizeof in kcalloc
RDMA/vmw_pvrdma: Clarify QP and CQ is_kernel logic
RDMA/vmw_pvrdma: Add UAR SRQ macros in ABI header file
i40iw: Change accelerated flag to bool
The flush/drain logic was not retaining the original wr opcode in
its completion. This can cause problems if the application uses
the completion opcode to make decisions.
Use bit 10 of the CQE header word to indicate the CQE is a special
drain completion, and save the original WR opcode in the cqe header
opcode field.
Fixes: 4fe7c2962e ("iw_cxgb4: refactor sq/rq drain logic")
Cc: stable@vger.kernel.org
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
The debugfs file prints the difference between host timestamps as a
seconds/nanoseconds tuple, along with a 64-bit nanoseconds hardware
timestamp. The host time is read using getnstimeofday() which is
deprecated because of the y2038 overflow, and it suffers from time jumps
during settimeofday() and leap seconds.
Converting to ktime_get_ts64() would solve those two, but I'm going
a little further here by changing to ktime_get() and printing 64-bit
nanoseconds on both host and hw timestamps. This simplifies the code
further and makes the output easier to understand.
The format of the debugfs file obviously changes here, but this should
only be read by humans and not scripts, so I assume it's fine.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Locking sequence of iw_cxgb4 and RoCE drivers in ib_register_device() is
slightly different and this leads to possible circular dependency locking
warning when both the devices are brought up.
Here is the locking sequence upto ib_register_device():
iw_cxgb4: rtnl_mutex(net stack) --> uld_mutex --> device_mutex
RoCE drivers: device_mutex --> rtnl_mutex
Here is the possibility of cross locking:
CPU #0 (iw_cxgb4) CPU #1 (RoCE drivers)
-> on interface up cxgb4_up()
executed with rtnl_mutex held
-> hold uld_mutex and try
registering ib device
-> In ib_register_device() hold
device_mutex
-> hold device mutex in
ib_register_device
-> try acquiring rtnl_mutex in
ib_enum_roce_netdev()
Current patch schedules the ib_register_device() functionality of
iw_cxgb4 to a workqueue to prevent the possible cross-locking.
Also rename the labels in c4iw_reister_device().
Signed-off-by: Potnuri Bharat Teja <bharat@chelsio.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
iw_cxgb4 has many BUG_ON()s that were left over from various enhancemnets
made over the years. Almost all of them should just be removed. Some,
however indicate a ULP usage error and can be handled w/o bringing down
the system.
If the condition cannot happen with correctly implemented cxgb4 sw/fw,
then remove the BUG_ON.
If the condition indicates a misbehaving ULP (like CQ overflows), add
proper recovery logic.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Conflicts:
drivers/infiniband/hw/cxgb4/cm.c
drivers/infiniband/hw/qib/qib_driver.c
drivers/infiniband/hw/qib/qib_mad.c
There were minor fixups needed in these files. Just minor context diffs
due to patches from independent sources touching the same basic area.
Signed-off-by: Doug Ledford <dledford@redhat.com>
For messages sent from the host to fw that solicit a reply from fw,
the c4iw_wr_wait struct pointer is passed in the host->fw message, and
included in the fw->host fw6_msg reply. This allows the sender to wait
until the reply is received, and the code processing the ingress reply
to wake up the sender.
If c4iw_wait_for_reply() times out, however, we need to keep the
c4iw_wr_wait object around in case the reply eventually does arrive.
Otherwise we have touch-after-free bugs in the wake_up paths.
This was hit due to a bad kernel driver that blocked ingress processing
of cxgb4 for a long time, causing iw_cxgb4 timeouts, but eventually
resuming ingress processing and thus hitting the touch-after-free bug.
So I want to fix iw_cxgb4 such that we'll at least keep the wait object
around until the reply comes. If it never comes we leak a small amount of
memory, but if it does come late, we won't potentially crash the system.
So add a kref struct in the c4iw_wr_wait struct, and take a reference
before sending a message to FW that will generate a FW6 reply. And remove
the reference (and potentially free the wait object) when the reply
is processed.
The ep code also uses the wr_wait for non FW6 CPL messages and doesn't
embed the c4iw_wr_wait object in the message sent to firmware. So for
those cases we add c4iw_wake_up_noref().
The mr/mw, cq, and qp object create/destroy paths do need this reference
logic. For these paths, c4iw_ref_send_wait() is introduced to take the
wr_wait reference, send the msg to fw, and then wait for the reply.
So going forward, iw_cxgb4 either uses c4iw_ofld_send(),
c4iw_wait_for_reply() and c4iw_wake_up_noref() like is done in the some
of the endpoint logic, or c4iw_ref_send_wait() and c4iw_wake_up_deref()
(formerly c4iw_wake_up()) when sending messages with the c4iw_wr_wait
object pointer embedded in the message and resulting FW6 reply.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Remove the embedded c4iw_wr_wait object in preparation for correctly
handling timeouts.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Remove the local stack allocated c4iw_wr_wait object in preparation for
correctly handling timeouts.
Also cleaned up some error path unwind logic to make it more readable.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Remove the local stack allocated c4iw_wr_wait object in preparation for
correctly handling timeouts.
Also cleaned up some error path unwind logic to make it more readable.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Remove the local stack allocated c4iw_wr_wait object in preparation for
correctly handling timeouts.
Also refactored some code to simplify it and make errpath unwinding
more readable.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Error logs of iw_cxgb4 needs to be printed by default. This patch
changes the necessary pr_debug() to appropriate pr_<log level>.
Signed-off-by: Potnuri Bharat Teja <bharat@chelsio.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
pr_debug() can be enabled to print function names, So removing the
unwanted __func__ parameters from debug logs.
Realign function parameters.
Signed-off-by: Potnuri Bharat Teja <bharat@chelsio.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Use a more typical logging style.
Miscellanea:
o Obsolete the c4iw_debug module parameter
o Coalesce formats
o Realign arguments
Signed-off-by: Joe Perches <joe@perches.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
