Parav Pandit [Tue, 13 Mar 2018 14:06:13 +0000 (16:06 +0200)]
IB/core: Honor return status of ib_init_ah_from_mcmember()
The return status of ib_init_ah_from_mcmember() is ignored by
cma_ib_mc_handler(). Honor it and return error event if ah attribute
initialization failed.
Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Tue, 13 Mar 2018 14:06:12 +0000 (16:06 +0200)]
IB/{core, ipoib}: Simplify ib_find_gid() for unused ndev
ib_find_gid() is only used by IPoIB driver. For IB link layer, GID table
entries are not based on netdevice. Netdevice parameter is unused here.
Therefore, it is removed.
Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Reviewed-by: Mark Bloch <markb@mellanox.com> Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Parav Pandit [Tue, 13 Mar 2018 14:06:11 +0000 (16:06 +0200)]
IB/core: Fix comments of GID query functions
Exported symbol's comments should be with function definition and not in
the header file. Therefore comments of ib_find_cached_gid() and
ib_find_cached_gid_by_port() functions are moved closer to their
definitions.
The function name in then comment is different than the actual function
name, fix it to be same as ib_cache_gid_find_by_filter().
Also current comment section of ib_find_cached_gid_by_port() contains the
desciption of ib_find_cached_gid(), fix that as well.
Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Doug Ledford [Wed, 14 Mar 2018 22:49:12 +0000 (18:49 -0400)]
Merge branch 'k.o/wip/dl-for-rc' into k.o/wip/dl-for-next
Due to bug fixes found by the syzkaller bot and taken into the for-rc
branch after development for the 4.17 merge window had already started
being taken into the for-next branch, there were fairly non-trivial
merge issues that would need to be resolved between the for-rc branch
and the for-next branch. This merge resolves those conflicts and
provides a unified base upon which ongoing development for 4.17 can
be based.
Conflicts:
drivers/infiniband/hw/mlx5/main.c - Commit 42cea83f9524
(IB/mlx5: Fix cleanup order on unload) added to for-rc and
commit b5ca15ad7e61 (IB/mlx5: Add proper representors support)
add as part of the devel cycle both needed to modify the
init/de-init functions used by mlx5. To support the new
representors, the new functions added by the cleanup patch
needed to be made non-static, and the init/de-init list
added by the representors patch needed to be modified to
match the init/de-init list changes made by the cleanup
patch.
Updates:
drivers/infiniband/hw/mlx5/mlx5_ib.h - Update function
prototypes added by representors patch to reflect new function
names as changed by cleanup patch
drivers/infiniband/hw/mlx5/ib_rep.c - Update init/de-init
stage list to match new order from cleanup patch
Arnd Bergmann [Tue, 20 Feb 2018 20:56:27 +0000 (21:56 +0100)]
infiniband: bnxt_re: use BIT_ULL() for 64-bit bit masks
On 32-bit targets, we otherwise get a warning about an impossible constant
integer expression:
In file included from include/linux/kernel.h:11,
from include/linux/interrupt.h:6,
from drivers/infiniband/hw/bnxt_re/ib_verbs.c:39:
drivers/infiniband/hw/bnxt_re/ib_verbs.c: In function 'bnxt_re_query_device':
include/linux/bitops.h:7:24: error: left shift count >= width of type [-Werror=shift-count-overflow]
#define BIT(nr) (1UL << (nr))
^~
drivers/infiniband/hw/bnxt_re/bnxt_re.h:61:34: note: in expansion of macro 'BIT'
#define BNXT_RE_MAX_MR_SIZE_HIGH BIT(39)
^~~
drivers/infiniband/hw/bnxt_re/bnxt_re.h:62:30: note: in expansion of macro 'BNXT_RE_MAX_MR_SIZE_HIGH'
#define BNXT_RE_MAX_MR_SIZE BNXT_RE_MAX_MR_SIZE_HIGH
^~~~~~~~~~~~~~~~~~~~~~~~
drivers/infiniband/hw/bnxt_re/ib_verbs.c:149:25: note: in expansion of macro 'BNXT_RE_MAX_MR_SIZE'
ib_attr->max_mr_size = BNXT_RE_MAX_MR_SIZE;
^~~~~~~~~~~~~~~~~~~
Fixes: 872f3578241d ("RDMA/bnxt_re: Add support for MRs with Huge pages") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Arnd Bergmann [Tue, 20 Feb 2018 20:56:26 +0000 (21:56 +0100)]
infiniband: qplib_fp: fix pointer cast
Building for a 32-bit target results in a couple of warnings from casting
between a 32-bit pointer and a 64-bit integer:
drivers/infiniband/hw/bnxt_re/qplib_fp.c: In function 'bnxt_qplib_service_nq':
drivers/infiniband/hw/bnxt_re/qplib_fp.c:333:23: error: cast to pointer from integer of different size [-Werror=int-to-pointer-cast]
bnxt_qplib_arm_srq((struct bnxt_qplib_srq *)q_handle,
^
drivers/infiniband/hw/bnxt_re/qplib_fp.c:336:12: error: cast to pointer from integer of different size [-Werror=int-to-pointer-cast]
(struct bnxt_qplib_srq *)q_handle,
^
In file included from include/linux/byteorder/little_endian.h:5,
from arch/arm/include/uapi/asm/byteorder.h:22,
from include/asm-generic/bitops/le.h:6,
from arch/arm/include/asm/bitops.h:342,
from include/linux/bitops.h:38,
from include/linux/kernel.h:11,
from include/linux/interrupt.h:6,
from drivers/infiniband/hw/bnxt_re/qplib_fp.c:39:
drivers/infiniband/hw/bnxt_re/qplib_fp.c: In function 'bnxt_qplib_create_srq':
include/uapi/linux/byteorder/little_endian.h:31:43: error: cast from pointer to integer of different size [-Werror=pointer-to-int-cast]
#define __cpu_to_le64(x) ((__force __le64)(__u64)(x))
^
include/linux/byteorder/generic.h:86:21: note: in expansion of macro '__cpu_to_le64'
#define cpu_to_le64 __cpu_to_le64
^~~~~~~~~~~~~
drivers/infiniband/hw/bnxt_re/qplib_fp.c:569:19: note: in expansion of macro 'cpu_to_le64'
req.srq_handle = cpu_to_le64(srq);
Using a uintptr_t as an intermediate works on all architectures.
Fixes: 37cb11acf1f7 ("RDMA/bnxt_re: Add SRQ support for Broadcom adapters") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Andrew Morton [Tue, 13 Mar 2018 22:06:45 +0000 (15:06 -0700)]
drivers/infiniband/ulp/srpt/ib_srpt.c: fix build with gcc-4.4.4
gcc-4.4.4 has issues with initialization of anonymous unions:
drivers/infiniband/ulp/srpt/ib_srpt.c: In function 'srpt_zerolength_write':
drivers/infiniband/ulp/srpt/ib_srpt.c:854: error: unknown field 'wr_cqe' specified in initializer
drivers/infiniband/ulp/srpt/ib_srpt.c:854: warning: initialization makes integer from pointer without a cast
Work aound this.
Fixes: 2a78cb4db487 ("IB/srpt: Fix an out-of-bounds stack access in srpt_zerolength_write()") Cc: Bart Van Assche <bart.vanassche@wdc.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Jason Gunthorpe <jgg@mellanox.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
Andrew Morton [Tue, 13 Mar 2018 21:51:57 +0000 (14:51 -0700)]
drivers/infiniband/core/verbs.c: fix build with gcc-4.4.4
gcc-4.4.4 has issues with initialization of anonymous unions.
drivers/infiniband/core/verbs.c: In function '__ib_drain_sq':
drivers/infiniband/core/verbs.c:2204: error: unknown field 'wr_cqe' specified in initializer
drivers/infiniband/core/verbs.c:2204: warning: initialization makes integer from pointer without a cast
Work around this.
Fixes: a1ae7d0345edd5 ("RDMA/core: Avoid that ib_drain_qp() triggers an out-of-bounds stack access") Cc: Bart Van Assche <bart.vanassche@wdc.com> Cc: Steve Wise <swise@opengridcomputing.com> Cc: Sagi Grimberg <sagi@grimberg.me> Cc: Jason Gunthorpe <jgg@mellanox.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
Mark Bloch [Wed, 14 Mar 2018 07:14:15 +0000 (09:14 +0200)]
IB/mlx5: Fix cleanup order on unload
On load we create private CQ/QP/PD in order to be used by UMR, we create
those resources after we register ourself as an IB device, and we destroy
them after we unregister as an IB device. This was changed by commit 16c1975f1032 ("IB/mlx5: Create profile infrastructure to add and remove
stages") which moved the destruction before we unregistration. This
allowed to trigger an invalid memory access when unloading mlx5_ib while
there are open resources:
We restore the original behavior by breaking the UMR stage into two parts,
pre and post IB registration stages, this way we can restore the original
functionality and maintain clean separation of logic between stages.
Fixes: 16c1975f1032 ("IB/mlx5: Create profile infrastructure to add and remove stages") Signed-off-by: Mark Bloch <markb@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Martin Wilck [Wed, 14 Feb 2018 20:45:43 +0000 (21:45 +0100)]
rdma_rxe: make rxe work over 802.1q VLAN devices
This patch fixes RDMA/rxe over 802.1q VLAN devices.
Without it, I observed the following behavior:
a) adding a VLAN device to RXE via rxe_net_add() creates a non-functional
RDMA device. This is caused by the logic in enum_all_gids_of_dev_cb() /
is_eth_port_of_netdev(), which only considers networks connected to
"upper devices" of the configured network device, resulting in an empty
set of gids for a VLAN interface that is an "upper device" itself.
Later attempts to connect via this rdma device fail in cma_acuire_dev()
because no gids can be resolved.
b) adding the master device of the VLAN device instead seems to work
initially, target addresses via VLAN devices are resolved successfully.
But the connection times out because no 802.1q VLAN headers are
inserted in the ethernet packets, which are therefore never received.
This happens because the RXE layer sends the packets via the master
device rather than the VLAN device.
The problem could be solved by changing either a) or b). My thinking was
that the logic in a) was created deliberately, thus I decided to work on
b). It turns out that the information about the VLAN interface for the gid
at hand is available in the AV information. My patch converts the RXE code
to use this netdev instead of rxe->ndev. With this change, RXE over vlan
works on my test system.
Signed-off-by: Martin Wilck <mwilck@suse.com> Reviewed-by: Moni Shoua <monis@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Leon Romanovsky [Tue, 13 Mar 2018 16:37:27 +0000 (18:37 +0200)]
RDMA/ucma: Don't allow join attempts for unsupported AF family
Users can provide garbage while calling to ucma_join_ip_multicast(),
it will indirectly cause to rdma_addr_size() return 0, making the
call to ucma_process_join(), which had the right checks, but it is
better to check the input as early as possible.
Fixes: 5bc2b7b397b0 ("RDMA/ucma: Allow user space to specify AF_IB when joining multicast") Reported-by: <syzbot+2287ac532caa81900a4e@syzkaller.appspotmail.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Reviewed-by: Sean Hefty <sean.hefty@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Arnd Bergmann [Tue, 13 Mar 2018 12:06:20 +0000 (13:06 +0100)]
RDMA/i40iw: include linux/irq.h
We get a build failure on ARM unless the header is included explicitly:
drivers/infiniband/hw/i40iw/i40iw_verbs.c: In function 'i40iw_get_vector_affinity':
drivers/infiniband/hw/i40iw/i40iw_verbs.c:2747:9: error: implicit declaration of function 'irq_get_affinity_mask'; did you mean 'irq_create_affinity_masks'? [-Werror=implicit-function-declaration]
return irq_get_affinity_mask(msix_vec->irq);
^~~~~~~~~~~~~~~~~~~~~
irq_create_affinity_masks
drivers/infiniband/hw/i40iw/i40iw_verbs.c:2747:9: error: returning 'int' from a function with return type 'const struct cpumask *' makes pointer from integer without a cast [-Werror=int-conversion]
return irq_get_affinity_mask(msix_vec->irq);
Ilya Lesokhin [Tue, 13 Mar 2018 13:18:48 +0000 (15:18 +0200)]
IB/mlx5: Maintain a single emergency page
The mlx5 driver needs to be able to issue invalidation to ODP MRs
even if it cannot allocate memory. To this end it preallocates
emergency pages to use when the situation arises.
This flow should be extremely rare enough, that we don't need
to worry about contention and therefore a single emergency page
is good enough.
Daniel Jurgens [Tue, 13 Mar 2018 13:18:47 +0000 (15:18 +0200)]
IB/mlx5: Only synchronize RCU once when removing mkeys
Instead synchronizing RCU in a loop when removing mkeys in a batch do it
once at the end before freeing them. The result is only waiting for one
RCU grace period instead of many serially.
Signed-off-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
Mark Bloch [Tue, 13 Mar 2018 13:18:46 +0000 (15:18 +0200)]
IB/mlx5: Expose more priorities for bypass namespace
BYPASS namespace is used by the RDMA side to insert flow rules into
the vport RX flow tables. Currently only 8 priorities are exposed,
increase this to 16 to allow more flexibility. This change will also
cause the BYPASS namespace to use 32 levels (as apposed to 16 today) of
flow tables, 16 levels for regular rules and 16 for don't trap rules.
Reviewed-by: Maor Gottlieb <maorg@mellanox.com> Signed-off-by: Mark Bloch <markb@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
Tatyana Nikolova [Mon, 12 Mar 2018 22:14:02 +0000 (17:14 -0500)]
RDMA/core: Do not use invalid destination in determining port reuse
cma_port_is_unique() allows local port reuse if the quad (source
address and port, destination address and port) for this connection
is unique. However, if the destination info is zero or unspecified, it
can't make a correct decision but still allows port reuse. For example,
sometimes rdma_bind_addr() is called with unspecified destination and
reusing the port can lead to creating a connection with a duplicate quad,
after the destination is resolved. The issue manifests when MPI scale-up
tests hang after the duplicate quad is used.
Set the destination address family and add checks for zero destination
address and port to prevent source port reuse based on invalid destination.
Fixes: 19b752a19dce ("IB/cma: Allow port reuse for rdma_id") Reviewed-by: Sean Hefty <sean.hefty@intel.com> Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com> Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Bart Van Assche [Mon, 12 Mar 2018 20:55:55 +0000 (13:55 -0700)]
IB/srp: Fix IPv6 address parsing
Split IPv6 addresses at the colon that separates the IPv6 address
and the port number instead of at a colon in the middle of the IPv6
address. Check whether the IPv6 address is surrounded with square
brackets.
Leon Romanovsky [Mon, 12 Mar 2018 19:26:37 +0000 (21:26 +0200)]
RDMA/mlx5: Fix crash while accessing garbage pointer and freed memory
The failure in rereg_mr flow caused to set garbage value (error value)
into mr->umem pointer. This pointer is accessed at the release stage
and it causes to the following crash.
There is not enough to simply change umem to point to NULL, because the
MR struct is needed to be accessed during MR deregistration phase, so
delay kfree too.
Leon Romanovsky [Sun, 11 Mar 2018 11:51:35 +0000 (13:51 +0200)]
RDMA/verbs: Simplify modify QP check
All callers to ib_modify_qp_is_ok() provides enum ib_qp_state
makes the checks of out-of-scope redundant. Let's remove them
together with updating function signature to return boolean result.
Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Leon Romanovsky [Sun, 11 Mar 2018 11:51:33 +0000 (13:51 +0200)]
RDMA/uverbs: Ensure validity of current QP state value
The QP state is internal enum which is checked at the driver
level by calling to ib_modify_qp_is_ok(). Move this check closer
to user and leave kernel users to be checked by compiler.
Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Leon Romanovsky [Sun, 11 Mar 2018 11:51:32 +0000 (13:51 +0200)]
RDMA/mlx5: Fix NULL dereference while accessing XRC_TGT QPs
mlx5 modify_qp() relies on FW that the error will be thrown if wrong
state is supplied. The missing check in FW causes the following crash
while using XRC_TGT QPs.
Memory state around the buggy address: ffff880066b99180: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ffff880066b99200: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>ffff880066b99280: 00 00 fc fc fc fc fc fc fc fc fc fc fc fc fc fc
^ ffff880066b99300: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc ffff880066b99380: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
Cc: syzkaller <syzkaller@googlegroups.com> Fixes: 0fb2ed66a14c ("IB/mlx5: Add create and destroy functionality for Raw Packet QP") Signed-off-by: Boris Pismenny <borisp@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
Doug Ledford [Tue, 13 Mar 2018 19:49:34 +0000 (15:49 -0400)]
Merge tag 'mlx5-updates-2018-02-28-2' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux into k.o/wip/dl-for-next
mlx5-updates-2018-02-28-2 (IPSec-2)
This series follows our previous one to lay out the foundations for IPSec
in user-space and extend current kernel netdev IPSec support. As noted in
our previous pull request cover letter "mlx5-updates-2018-02-28-1 (IPSec-1)",
the IPSec mechanism will be supported through our flow steering mechanism.
Therefore, we need to change the initialization order. Furthermore, IPsec
is also supported in both egress and ingress. Since our current flow
steering is egress only, we add an empty (only implemented through FPGA
steering ops) egress namespace to handle that case. We also implement
the required flow steering callbacks and logic in our FPGA driver.
We extend the FPGA support for ESN and modifying a xfrm too. Therefore, we
add support for some new FPGA command interface that supports them. The
other required bits are added too. The new features and requirements are
advertised via cap bits.
Last but not least, we revise our driver's accel_esp API. This API will be
shared between our netdev and IB driver, so we need to have all the required
functionality from both worlds.
Doug Ledford [Fri, 9 Mar 2018 23:07:46 +0000 (18:07 -0500)]
Revert "RDMA/mlx5: Fix integer overflow while resizing CQ"
The original commit of this patch has a munged log message that is
missing several of the tags the original author intended to be on the
patch. This was due to patchworks misinterpreting a cut-n-paste
separator line as an end of message line and munging the mbox that was
used to import the patch:
Steve Wise [Thu, 1 Mar 2018 21:58:20 +0000 (13:58 -0800)]
mlx4_ib: zero out struct ib_pd when allocating
Zero out the fields of the struct ib_pd for user mode pds so that
users querying pds via nldev will not get garbage. For simplicity,
use kzalloc() to allocate the mlx4_ib_pd struct.
Signed-off-by: Steve Wise <swise@opengridcomputing.com> Reviewed-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Steve Wise [Thu, 1 Mar 2018 21:57:44 +0000 (13:57 -0800)]
RDMA/nldev: provide detailed CM_ID information
Implement RDMA nldev netlink interface to get detailed CM_ID information.
Because cm_id's are attached to rdma devices in various work queue
contexts, the pid and task information at restrak_add() time is sometimes
not useful. For example, an nvme/f host connection cm_id ends up being
bound to a device in a work queue context and the resulting pid at attach
time no longer exists after connection setup. So instead we mark all
cm_id's created via the rdma_ucm as "user", and all others as "kernel".
This required tweaking the restrack code a little. It also required
wrapping some rdma_cm functions to allow passing the module name string.
Signed-off-by: Steve Wise <swise@opengridcomputing.com> Reviewed-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Steve Wise [Thu, 1 Mar 2018 21:57:29 +0000 (13:57 -0800)]
RDMA/nldev: common resource dumpit function
Create a common dumpit function that can be used by all common resource
types. This reduces code replication and simplifies the code as we add
more resource types.
Signed-off-by: Steve Wise <swise@opengridcomputing.com> Reviewed-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
The kernel compiled with CONFIG_REFCOUNT_FULL produces the following
error. The reason to it that initial value of refcount_t is supposed
to be more than 0, change it.
Aviad Yehezkel [Thu, 18 Jan 2018 14:02:17 +0000 (16:02 +0200)]
net/mlx5: IPSec, Add support for ESN
Currently ESN is not supported with IPSec device offload.
This patch adds ESN support to IPsec device offload.
Implementing new xfrm device operation to synchronize offloading device
ESN with xfrm received SN. New QP command to update SA state at the
following:
^ - marks where QP command invoked to update the SA ESN state
machine.
| - marks the start of the ESN scope (0-2^32-1). At this point move SA
ESN overlap bit to zero and increment ESN.
* - marks the middle of the ESN scope (2^31). At this point move SA
ESN overlap bit to one.
Aviad Yehezkel [Sun, 18 Feb 2018 13:07:20 +0000 (15:07 +0200)]
net/mlx5: Add flow-steering commands for FPGA IPSec implementation
In order to add a context to the FPGA, we need to get both the software
transform context (which includes the keys, etc) and the
source/destination IPs (which are included in the steering
rule). Therefore, we register new set of firmware like commands for
the FPGA. Each time a rule is added, the steering core infrastructure
calls the FPGA command layer. If the rule is intended for the FPGA,
it combines the IPs information with the software transformation
context and creates the respective hardware transform.
Afterwards, it calls the standard steering command layer.
Aviad Yehezkel [Thu, 18 Jan 2018 11:05:48 +0000 (13:05 +0200)]
net/mlx5: Refactor accel IPSec code
The current code has one layer that executed FPGA commands and
the Ethernet part directly used this code. Since downstream patches
introduces support for IPSec in mlx5_ib, we need to provide some
abstractions. This patch refactors the accel code into one layer
that creates a software IPSec transformation and another one which
creates the actual hardware context.
The internal command implementation is now hidden in the FPGA
core layer. The code also adds the ability to share FPGA hardware
contexts. If two contexts are the same, only a reference count
is taken.
Aviad Yehezkel [Tue, 16 Jan 2018 14:12:22 +0000 (16:12 +0200)]
net/mlx5: IPSec, Add command V2 support
This patch adds V2 command support.
New fpga devices support extended features (udp encap, esn etc...), this
features require new hardware sadb format therefore we have a new version
of commands to manipulate it.
Yossi Kuperman [Sun, 22 Oct 2017 16:45:45 +0000 (19:45 +0300)]
net/mlx5e: IPSec, Add support for ESP trailer removal by hardware
Current hardware decrypts and authenticates incoming ESP packets.
Subsequently, the software extracts the nexthdr field, truncates the
trailer and adjusts csum accordingly.
With this patch and a capable device, the trailer is being removed
by the hardware and the nexthdr field is conveyed via PET. This way
we avoid both the need to access the trailer (cache miss) and to
compute its relative checksum, which significantly improve
the performance.
Experiment shows that trailer removal improves the performance by
2Gbps, (netperf). Both forwarding and host-to-host configurations.
Yossi Kuperman [Sun, 22 Oct 2017 16:43:58 +0000 (19:43 +0300)]
net/mlx5: IPSec, Generalize sandbox QP commands
The current code assume only SA QP commands.
Refactor in order to pave the way for new QP commands:
1. Generic cmd response format.
2. SA cmd checks are in dedicated functions.
3. Aligned debug prints.
Doug Ledford [Wed, 7 Mar 2018 20:40:29 +0000 (15:40 -0500)]
Merge tag 'mlx5-updates-2018-02-28-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux into k.o/wip/dl-for-next
mlx5-updates-2018-02-28-1 (IPSec-1)
This series consists of some fixes and refactors for the mlx5 drivers,
especially around the FPGA and flow steering. Most of them are trivial
fixes and are the foundation of allowing IPSec acceleration from user-space.
We use flow steering abstraction in order to accelerate IPSec packets.
When a user creates a steering rule, [s]he states that we'll carry an
encrypt/decrypt flow action (using a specific configuration) for every
packet which conforms to a certain match. Since currently offloading these
packets is done via FPGA, we'll add another set of flow steering ops.
These ops will execute the required FPGA commands and then call the
standard steering ops.
In order to achieve this, we need that the commands will get all the
required information. Therefore, we pass the fte object and embed the
flow_action struct inside the fte. In addition, we add the shim layer
that will later be used for alternating between the standard and the
FPGA steering commands.
Some fixes, like " net/mlx5e: Wait for FPGA command responses with a timeout"
are very relevant for user-space applications, as these applications could
be killed, but we still want to wait for the FPGA and update the kernel's
database.
Zhu Yanjun [Tue, 27 Feb 2018 11:04:32 +0000 (06:04 -0500)]
IB/rxe: remove unnecessary skb_clone
In send_atomic_ack function, it is not necessary to make a
skb_clone. To gain better performance (high throughput and
low latency), this skb_clone is removed.
The following tests are made.
server client
--------- ---------
|1.1.1.1|<----rxe-channel--->|1.1.1.2|
--------- ---------
On server: rping -s -a 1.1.1.1 -v -C 1000 -S 512
On client: rping -c -a 1.1.1.1 -v -C 1000 -S 512
The kernel config CONFIG_DEBUG_KMEMLEAK is enabled on both server
and client.
This test runs for several hours. There is no memory leak and the whole
system can work well.
Based on the above network, the following tests are made.
Bart Van Assche [Fri, 2 Mar 2018 21:14:15 +0000 (13:14 -0800)]
IB/srpt: Add RDMA/CM support
Add a parameter for configuring the port on which the ib_srpt driver
listens for incoming RDMA/CM connections, namely
/sys/kernel/config/target/srpt/discovery_auth/rdma_cm_port. The default
value for this parameter is 0 which means "do not listen for incoming
RDMA/CM connections". Add RDMA/CM support to all code that handles
connection state changes. Modify srpt_init_nodeacl() such that ACLs can
be configured for IPv4 and IPv6 addresses.
Note: incoming connection requests are only accepted for ports that
have been enabled. See also the "if (!sport->enabled)" code in the
connection request handler. See also the following configfs attribute:
/sys/kernel/config/target/srpt/$port/$port/enable.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Leon Romanovsky [Wed, 7 Mar 2018 12:49:09 +0000 (14:49 +0200)]
RDMA/ucma: Limit possible option size
Users of ucma are supposed to provide size of option level,
in most paths it is supposed to be equal to u8 or u16, but
it is not the case for the IB path record, where it can be
multiple of struct ib_path_rec_data.
This patch takes simplest possible approach and prevents providing
values more than possible to allocate.
Reported-by: syzbot+a38b0e9f694c379ca7ce@syzkaller.appspotmail.com Fixes: 7ce86409adcd ("RDMA/ucma: Allow user space to set service type") Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
Parav Pandit [Wed, 7 Mar 2018 06:07:41 +0000 (08:07 +0200)]
IB/core: Fix possible crash to access NULL netdev
resolved_dev returned might be NULL as ifindex is transient number.
Ignoring NULL check of resolved_dev might crash the kernel.
Therefore perform NULL check before accessing resolved_dev.
Additionally rdma_resolve_ip_route() invokes addr_resolve() which
performs check and address translation for loopback ifindex.
Therefore, checking it again in rdma_resolve_ip_route() is not helpful.
Therefore, the code is simplified to avoid IFF_LOOPBACK check.
Fixes: 200298326b27 ("IB/core: Validate route when we init ah") Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Doug Ledford <dledford@redhat.com>
Aviad Yehezkel [Sun, 18 Feb 2018 13:00:54 +0000 (15:00 +0200)]
net/mlx5: Flow steering cmd interface should get the fte when deleting
Previously, deleting a flow steering entry only got the index.
Since the FPGA implementation of FTE's deletion might need to dig
inside the FTE itself, we would like to get the FTE's context.
Changing the interface to pass the FTE context.
Aviad Yehezkel [Sun, 18 Feb 2018 11:17:17 +0000 (13:17 +0200)]
net/mlx5: Add empty egress namespace to flow steering core
Currently, we don't support egress flow steering namespace in mlx5
flow steering core implementation. However, when we want to encrypt
a packet, we model it as a flow steering rule in the egress path.
To overcome this, we add an empty egress namespace to flow steering.
This namespace is initialized only when ipsec support exists.
In the future, this will grow to a full blown full steering
implementation, resembling the ingress path.
Matan Barak [Sun, 20 Aug 2017 12:46:51 +0000 (15:46 +0300)]
net/mlx5: Add shim layer between fs and cmd
The shim layer allows each namespace to define possibly different
functionality for add/delete/update commands. The shim layer
introduced here, will be used to support flow steering with the FPGA.
Matan Barak [Wed, 16 Aug 2017 06:43:48 +0000 (09:43 +0300)]
{net,IB}/mlx5: Add has_tag to mlx5_flow_act
The has_tag member will indicate whether a tag action was specified
in flow specification.
A flow tag 0 = MLX5_FS_DEFAULT_FLOW_TAG is assumed a valid flow tag
that is currently used by mlx5 RDMA driver, whereas in HW flow_tag = 0
means that the user doesn't care about flow_tag. HW always provide
a flow_tag = 0 if all flow tags requested on a specific flow are 0.
So we need a way (in the driver) to differentiate between a user really
requesting flow_tag = 0 and a user who does not care, in order to be
able to report conflicting flow tags on a specific flow.
Boris Pismenny [Wed, 16 Aug 2017 06:33:30 +0000 (09:33 +0300)]
IB/mlx5: Pass mlx5_flow_act struct instead of multiple arguments
Group and pass all function arguments of parse_flow_attr call in one
common struct mlx5_flow_act.
This patch passes all the action arguments of parse_flow_attr in one common
struct mlx5_flow_act. It allows us to scale the number of actions without adding
new arguments to the function.
Signed-off-by: Matan Barak <matanb@mellanox.com> Signed-off-by: Boris Pismenny <borisp@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com> Acked-by: Jason Gunthorpe <jgg@mellanox.com>
Matan Barak [Sun, 19 Nov 2017 15:51:13 +0000 (15:51 +0000)]
net/mlx5: FPGA and IPSec initialization to be before flow steering
Some flow steering namespace initialization (i.e. egress namespace)
might depend on FPGA capabilities. Changing the initialization order
such that the FPGA will be initialized before flow steering.
Flow steering fs cmds initialization might depend on
IPSec capabilities. Changing the initialization order such
that the IPSec will be initialized before flow steering as well.
The issue happens if RQ poll_cq or SQ poll_cq or Async error event tries to
put the error QP in flush list. Since SQ and RQ of each error qp are added
to two different flush list, we need to protect it using locks of
corresponding CQs. Difference in order of acquiring the lock in
SQ poll_cq and RQ poll_cq can cause a hard lockup.
Revisits the locking strategy and removes the usage of qplib_cq.hwq.lock.
Instead of this lock, introduces qplib_cq.flush_lock to handle
addition/deletion of QPs in flush list. Also, always invoke the flush_lock
in order (SQ CQ lock first and then RQ CQ lock) to avoid any potential
deadlock.
Other than the poll_cq context, the movement of QP to/from flush list can
be done in modify_qp context or from an async error event from HW.
Synchronize these operations using the bnxt_re verbs layer CQ locks.
To achieve this, adds a call back to the HW abstraction layer(qplib) to
bnxt_re ib_verbs layer in case of async error event. Also, removes the
buddy cq functions as it is no longer required.
Max Gurtovoy [Mon, 5 Mar 2018 18:09:48 +0000 (20:09 +0200)]
RDMA/core: Reduce poll batch for direct cq polling
Fix warning limit for kernel stack consumption:
drivers/infiniband/core/cq.c: In function 'ib_process_cq_direct':
drivers/infiniband/core/cq.c:78:1: error: the frame size of 1032 bytes
is larger than 1024 bytes [-Werror=frame-larger-than=]
Using smaller ib_wc array on the stack brings us comfortably below that
limit again.
Fixes: 246d8b184c10 ("IB/cq: Don't force IB_POLL_DIRECT poll context for ib_process_cq_direct") Reported-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Sergey Gorenko <sergeygo@mellanox.com> Signed-off-by: Max Gurtovoy <maxg@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com> Acked-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Dan Carpenter [Tue, 6 Mar 2018 10:00:31 +0000 (13:00 +0300)]
IB/mlx5: Fix an error code in __mlx5_ib_modify_qp()
"err" is either zero or possibly uninitialized here. It should be
-EINVAL.
Fixes: 427c1e7bcd7e ("{IB, net}/mlx5: Move the modify QP operation table to mlx5_ib") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Acked-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Mark Bloch [Mon, 5 Mar 2018 18:09:47 +0000 (20:09 +0200)]
IB/mlx5: When not in dual port RoCE mode, use provided port as native
The series that introduced dual port RoCE mode assumed that we don't have
a dual port HCA that use the mlx5 driver, this is not the case for
Connect-IB HCAs. This reasoning led to assigning 1 as the native port
index which causes issue when the second port is used.
For example query_pkey() when called on the second port will return values
of the first port. Make sure that we assign the right port index as the
native port index.
Fixes: 32f69e4be269 ("{net, IB}/mlx5: Manage port association for multiport RoCE") Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Signed-off-by: Mark Bloch <markb@mellanox.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Jack M [Mon, 5 Mar 2018 18:09:46 +0000 (20:09 +0200)]
IB/mlx4: Include GID type when deleting GIDs from HW table under RoCE
The commit cited below added a gid_type field (RoCEv1 or RoCEv2)
to GID properties.
When adding GIDs, this gid_type field was copied over to the
hardware gid table. However, when deleting GIDs, the gid_type field
was not copied over to the hardware gid table.
As a result, when running RoCEv2, all RoCEv2 gids in the
hardware gid table were set to type RoCEv1 when any gid was deleted.
This problem would persist until the next gid was added (which would again
restore the gid_type field for all the gids in the hardware gid table).
Fix this by copying over the gid_type field to the hardware gid table
when deleting gids, so that the gid_type of all remaining gids is
preserved when a gid is deleted.
Fixes: b699a859d17b ("IB/mlx4: Add gid_type to GID properties") Reviewed-by: Moni Shoua <monis@mellanox.com> Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
When using IPv4 addresses in RoCEv2, the GID format for the mapped
IPv4 address should be: ::ffff:<4-byte IPv4 address>.
In the cited commit, IPv4 mapped IPV6 addresses had the 3 upper dwords
zeroed out by memset, which resulted in deleting the ffff field.
However, since procedure ipv6_addr_v4mapped() already verifies that the
gid has format ::ffff:<ipv4 address>, no change is needed for the gid,
and the memset can simply be removed.
Fixes: 7e57b85c444c ("IB/mlx4: Add support for setting RoCEv2 gids in hardware") Reviewed-by: Moni Shoua <monis@mellanox.com> Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
RDMA/qedr: Fix kernel panic when running fio over NFSoRDMA
Race in qedr_poll_cq, lastest_cqe wasn't protected by lock,
leading to a case where two context's accessing poll_cq at
the same time lead to one of them having a pointer to an old
latest_cqe and reading an invalid cqe element
Signed-off-by: Amit Radzi <Amit.Radzi@cavium.com> Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Fix iWARP connect and listen to use the mapped port for
ipv4 and ipv6. Without this fixed, running on a server
that has iwpmd enabled will not use the correct port
Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
RDMA/bnxt_re/qplib_sp: Use true and false for boolean values
Assign true or false to boolean variables instead of an integer value.
This issue was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> Acked-by: Selvin Xavier <selvin.xavier@broadcom.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Sergey Gorenko [Mon, 5 Mar 2018 18:15:56 +0000 (20:15 +0200)]
IB/srp: Use the IB_DEVICE_SG_GAPS_REG HCA feature if supported
If a HCA supports the SG_GAPS_REG feature then fewer memory regions
are required per command. This patch reduces the number of memory
regions that is allocated per SRP session.
Signed-off-by: Sergey Gorenko <sergeygo@mellanox.com> Reviewed-by: Max Gurtovoy <maxg@mellanox.com> Tested-by: Laurence Oberman <loberman@redhat.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Acked-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Bart Van Assche [Mon, 5 Mar 2018 17:58:33 +0000 (09:58 -0800)]
IB/hfi1: Add a missing rcu_read_unlock()
This patch avoids that sparse reports the following:
drivers/infiniband/hw/hfi1/driver.c:251:13: warning: context imbalance in 'rcv_hdrerr' - different lock contexts for basic block
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com> Cc: Mike Marciniszyn <mike.marciniszyn@intel.com> Cc: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Shiraz Saleem [Fri, 2 Mar 2018 21:17:14 +0000 (15:17 -0600)]
i40iw: Implement get_vector_affinity API
Storage ULPs (like NVMEoF) benefit from exposing affinity mapping
per completion vector to find the optimal multi-queue affinity
assignments. The ULPs call the verbs API ib_get_vector_affinity
introduced in commit c66cd353bbe ("RDMA/core: expose affinity mappings per
completion vector") to get the underlying devices affinity mappings.
Add support in driver to expose the affinity masks per MSI-X
completion vector.
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Shiraz Saleem [Fri, 2 Mar 2018 21:17:13 +0000 (15:17 -0600)]
i40iw: Improve CM node lookup time on connection setup
Currently all CM nodes involved in a connection are
maintained in a connected_node list per dev. During
connection setup, we need to search this every time
we receive a packet on the iWARP LAN Queue (ILQ) and
this can be pretty inefficient for large number of
connections.
Fix this by organizing the CM nodes in two lists -
accelerated list and non-accelerated list. The search
on ILQ receive would be limited to only non accelerated
nodes. When a node moves to RTS, it is added to the
accelerated list.
Benchmarking ucmatose 16k connections shows a 20%
improvement in test completion time.
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>