This reverts commit b2d2440430.
It's true that creating rxe on top of 802.1q interfaces doesn't work.
Thus, commit fd49ddaf7e ("RDMA/rxe: prevent rxe creation on top of vlan
interface") was absolutely correct.
But b2d2440430 was incorrect assuming that with this change, RDMA and
VLAN don't work togehter at all. It just has to be set up
differently. Rather than creating rxe on top of the VLAN interface, rxe
must be created on top of the physical interface. RDMA then works just
fine through VLAN interfaces on top of that physical interface, via the
"upper device" logic.
This is hard to see in the rxe logic because it never talks about vlan,
but instead rxe carefully selects upper vlan netdevices when working with
packets which in turn imply certain vlan tagging. This is all done
correctly and interacts with the gid table with VLAN support the same as
real HW does.
b2d2440430 broke this setup deliberately and should thus be
reverted. Also, b2d2440430 removed rxe_dma_device(), so adapt the revert
to discard that hunk.
Fixes: b2d2440430 ("RDMA/rxe: Remove VLAN code leftovers from RXE")
Link: https://lore.kernel.org/r/20210120161913.7347-1-mwilck@suse.com
Signed-off-by: Martin Wilck <mwilck@suse.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
This moves siw and rxe to be virtual devices in the device tree:
lrwxrwxrwx 1 root root 0 Nov 6 13:55 /sys/class/infiniband/rxe0 -> ../../devices/virtual/infiniband/rxe0/
Previously they were trying to parent themselves to the physical device of
their attached netdev, which doesn't make alot of sense.
My hope is this will solve some weird syzkaller hits related to sysfs as
it could be possible that the parent of a netdev is another netdev, eg
under bonding or some other syzkaller found netdev configuration.
Nesting a ib_device under anything but a physical device is going to cause
inconsistencies in sysfs during destructions.
Link: https://lore.kernel.org/r/0-v1-dcbfc68c4b4a+d6-virtual_dev_jgg@nvidia.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Use the ib_dma_* helpers to skip the DMA translation instead. This
removes the last user if dma_virt_ops and keeps the weird layering
violation inside the RDMA core instead of burderning the DMA mapping
subsystems with it. This also means the software RDMA drivers now don't
have to mess with DMA parameters that are not relevant to them at all, and
that in the future we can use PCI P2P transfers even for software RDMA, as
there is no first fake layer of DMA mapping that the P2P DMA support.
Link: https://lore.kernel.org/r/20201106181941.1878556-8-hch@lst.de
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
dma_virt_ops requires that all pages have a kernel virtual address.
Introduce a INFINIBAND_VIRT_DMA Kconfig symbol that depends on !HIGHMEM
and make all three drivers depend on the new symbol.
Also remove the ARCH_DMA_ADDR_T_64BIT dependency, which has been obsolete
since commit 4965a68780 ("arch: define the ARCH_DMA_ADDR_T_64BIT config
symbol in lib/Kconfig")
Fixes: 551199aca1 ("lib/dma-virt: Add dma_virt_ops")
Link: https://lore.kernel.org/r/20201106181941.1878556-2-hch@lst.de
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The code which limited the number of unacknowledged PSNs was incorrect.
The PSNs are limited to 24 bits and wrap back to zero from 0x00ffffff.
The test was computing a 32 bit value which wraps at 32 bits so that
qp->req.psn can appear smaller than the limit when it is actually larger.
Replace '>' test with psn_compare which is used for other PSN comparisons
and correctly handles the 24 bit size.
Fixes: 8700e3e7c4 ("Soft RoCE driver")
Link: https://lore.kernel.org/r/20201013170741.3590-1-rpearson@hpe.com
Signed-off-by: Bob Pearson <rpearson@hpe.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Allowing userspace to invoke these commands is probably going to crash
these drivers as they are not tested and not expecting to use them on a
user object.
For example pvrdma touches cq->ring_state which is not initialized for
user QPs.
These commands are effected:
- IB_USER_VERBS_CMD_REQ_NOTIFY_CQ is ibv_cmd_req_notify_cq() in
rdma-core, only hfi1, ipath and rxe calls it.
- IB_USER_VERBS_CMD_POLL_CQ is ibv_cmd_poll_cq() in rdma-core, only
ipath and hfi1 calls it.
- IB_USER_VERBS_CMD_POST_SEND/RECV is ibv_cmd_post_send/recv() in
rdma-core, only ipath and hfi1 call them.
- IB_USER_VERBS_CMD_POST_SRQ_RECV is ibv_cmd_post_srq_recv() in
rdma-core, only ipath and hfi1 calls it.
- IB_USER_VERBS_CMD_PEEK_CQ isn't even implemented anywhere
- IB_USER_VERBS_CMD_CREATE/DESTROY_AH is ibv_cmd_create/destroy_ah() in
rdma-core, only bnxt_re, efa, hfi1, ipath, mlx5, orcrdma, and rxe call
it.
Link: https://lore.kernel.org/r/10-v1-caa70ba3d1ab+1436e-ucmd_mask_jgg@nvidia.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Each driver should check that the QP attrs create_flags is supported.
Unfortuantely when create_flags was added to the QP attrs the drivers were
not updated. uverbs_ex_cmd_mask was used to block it - even though kernel
drivers use these flags too.
Check that flags is zero in all drivers that don't use it, remove
IB_USER_VERBS_EX_CMD_CREATE_QP from uverbs_ex_cmd_mask. Fix the error code
to be EOPNOTSUPP.
Link: https://lore.kernel.org/r/8-v1-caa70ba3d1ab+1436e-ucmd_mask_jgg@nvidia.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Each driver should check that the CQ attrs is supported. Unfortuantely
when flags was added to the CQ attrs the drivers were not updated,
uverbs_ex_cmd_mask was used to block it. This was missed when create CQ
was converted to ioctl, so non-zero flags could have been passed into
drivers.
Check that flags is zero in all drivers that don't use it, remove
IB_USER_VERBS_EX_CMD_CREATE_CQ from uverbs_ex_cmd_mask.
Fixes: 41b2a71fc8 ("IB/uverbs: Move ioctl path of create_cq and destroy_cq to a new file")
Link: https://lore.kernel.org/r/7-v1-caa70ba3d1ab+1436e-ucmd_mask_jgg@nvidia.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
These functions all depend on the driver providing a specific op:
- REREG_MR is rereg_user_mr(). bnxt_re set this without providing the op
- ATTACH/DEATCH_MCAST is attach_mcast()/detach_mcast(). usnic set this
without providing the op
- OPEN_QP doesn't involve the driver but requires a XRCD. qedr provides
xrcd but forgot to set it, usnic doesn't provide XRCD but set it anyhow.
- OPEN/CLOSE_XRCD are the ops alloc_xrcd()/dealloc_xrcd()
- CREATE_SRQ/DESTROY_SRQ are the ops create_srq()/destroy_srq()
- QUERY/MODIFY_SRQ is op query_srq()/modify_srq(). hns sets this but
sometimes supplies a NULL op.
- RESIZE_CQ is op resize_cq(). bnxt_re sets this boes doesn't supply an op
- ALLOC/DEALLOC_MW is alloc_mw()/dealloc_mw(). cxgb4 provided an
(now deleted) implementation but no userspace
All drivers were checked that no drivers provide the op without also
setting uverbs_cmd_mask so this should have no functional change.
Link: https://lore.kernel.org/r/4-v1-caa70ba3d1ab+1436e-ucmd_mask_jgg@nvidia.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The patch referenced below has a typo that results in using the wrong L2
header size for outbound traffic. (V4 <-> V6).
It also breaks kernel-side RC traffic because they use AVs that use
RDMA_NETWORK_XXX enums instead of RXE_NETWORK_TYPE_XXX enums. Fix this by
transcoding between these enum types.
Fixes: e0d696d201 ("RDMA/rxe: Move the definitions for rxe_av.network_type to uAPI")
Link: https://lore.kernel.org/r/20201016211343.22906-1-rpearson@hpe.com
Signed-off-by: Bob Pearson <rpearson@hpe.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
If skb_clone() is unable to allocate memory for a new sk_buff this is not
detected by the current code.
Check for a NULL return and continue. This is similar to other errors in
this loop over QPs attached to the multicast address and consistent with
the unreliable UD transport.
Fixes: e7ec96fc79 ("RDMA/rxe: Fix skb lifetime in rxe_rcv_mcast_pkt()")
Addresses-Coverity-ID: 1497804: Null pointer dereferences (NULL_RETURNS)
Link: https://lore.kernel.org/r/20201013184236.5231-1-rpearson@hpe.com
Signed-off-by: Bob Pearson <rpearson@hpe.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
RXE was wrongly using an internal kernel enum as part of its uAPI, split
this out into a dedicated uAPI enum just for RXE. It only uses the IPv4
and IPv6 values.
This was exposed by changing the internal kernel enum definition which
broke RXE.
Fixes: 1c15b4f2a4 ("RDMA/core: Modify enum ib_gid_type and enum rdma_network_type")
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Fix a bug in rxe_rcv() that causes all multicast packets to be
dropped. Currently rxe_match_dgid() is called for each packet to verify
that the destination IP address matches one of the entries in the port
source GID table. This is incorrect for IP multicast addresses since they
do not appear in the GID table.
Add code to detect multicast addresses.
Change function name to rxe_chk_dgid() which is clearer.
Link: https://lore.kernel.org/r/20201008212753.265249-1-rpearson@hpe.com
Signed-off-by: Bob Pearson <rpearson@hpe.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The changes referenced below replaced sbk_clone)_ by taking additional
references, passing the skb along and then freeing the skb. This
deleted the packets before they could be processed and additionally
passed bad data in each packet. Since pkt is stored in skb->cb
changing pkt->qp changed it for all the packets.
Replace skb_get() by sbk_clone() in rxe_rcv_mcast_pkt() for cases where
multiple QPs are receiving multicast packets on the same address.
Delete kfree_skb() because the packets need to live until they have been
processed by each QP. They are freed later.
Fixes: 86af617641 ("IB/rxe: remove unnecessary skb_clone")
Fixes: fe896ceb57 ("IB/rxe: replace refcount_inc with skb_get")
Link: https://lore.kernel.org/r/20201008203651.256958-1-rpearson@hpe.com
Signed-off-by: Bob Pearson <rpearson@hpe.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Struct rxe_mem had pd, lkey and rkey values both in itself and in the
struct ib_mr which is also included in rxe_mem.
Delete these entries and replace references with the ones in ibmr.Add
mr_pd, mr_lkey and mr_rkey macros which extract these values from mr.
Link: https://lore.kernel.org/r/20201008212818.265303-1-rpearson@hpe.com
Signed-off-by: Bob Pearson <rpearson@hpe.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Leon Romanovsky says:
====================
IBTA declares speed as 16 bits, but kernel stores it in u8. This series
fixes in-kernel declaration while keeping external interface intact.
====================
Based on the mlx5-next branch at
git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux
due to dependencies.
* branch 'mlx5_active_speed':
RDMA: Fix link active_speed size
RDMA/mlx5: Delete duplicated mlx5_ptys_width enum
net/mlx5: Refactor query port speed functions
Pull rdma fixes from Jason Gunthorpe:
"A number of driver bug fixes and a few recent regressions:
- Several bug fixes for bnxt_re. Crashing, incorrect data reported,
and corruption on new HW
- Memory leak and crash in rxe
- Fix sysfs corruption in rxe if the netdev name is too long
- Fix a crash on error unwind in the new cq_pool code
- Fix kobject panics in rtrs by working device lifetime properly
- Fix a data corruption bug in iser target related to misaligned
buffers"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
IB/isert: Fix unaligned immediate-data handling
RDMA/rtrs-srv: Set .release function for rtrs srv device during device init
RDMA/bnxt_re: Remove set but not used variable 'qplib_ctx'
RDMA/core: Fix reported speed and width
RDMA/core: Fix unsafe linked list traversal after failing to allocate CQ
RDMA/bnxt_re: Remove the qp from list only if the qp destroy succeeds
RDMA/bnxt_re: Fix driver crash on unaligned PSN entry address
RDMA/bnxt_re: Restrict the max_gids to 256
RDMA/bnxt_re: Static NQ depth allocation
RDMA/bnxt_re: Fix the qp table indexing
RDMA/bnxt_re: Do not report transparent vlan from QP1
RDMA/mlx4: Read pkey table length instead of hardcoded value
RDMA/rxe: Fix panic when calling kmem_cache_create()
RDMA/rxe: Fix memleak in rxe_mem_init_user
RDMA/rxe: Fix the parent sysfs read when the interface has 15 chars
RDMA/rtrs-srv: Replace device_register with device_initialize and device_add
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>
Like any other IB verbs objects, AH are refcounted by ib_core. The release
of those objects are controlled by ib_core with promise that AH destroy
can't fail.
Being SW object for now, this change makes dealloc_ah() to behave like any
other destroy IB flows.
Fixes: d345691471 ("RDMA: Handle AH allocations by IB/core")
Link: https://lore.kernel.org/r/20200907120921.476363-3-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The IB verbs objects are counted by the kernel and ib_core ensures that
deallocate PD will success so it will be called once all other objects
that depends on PD will be released. This is achieved by managing various
reference counters on such objects.
The mlx5 driver didn't follow this standard flow when allowed DEVX objects
that are not managed by ib_core to be interleaved with the ones under
ib_core responsibility.
In such interleaved scenarios deallocate command can fail and ib_core will
leave uobject in internal DB and attempt to clean it later to free
resources anyway.
This change partially restores returned value from dealloc_pd() for all
drivers, but keeping in mind that non-DEVX devices and kernel verbs paths
shouldn't fail.
Fixes: 21a428a019 ("RDMA: Handle PD allocations by IB/core")
Link: https://lore.kernel.org/r/20200907120921.476363-2-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Change rxe pools to use kzalloc instead of kmem_cache to allocate memory
for rxe objects. The pools are not really necessary and they trigger
hardened user copy warnings as the ioctl framework copies the QP number
directly to userspace.
Also the general project to move object alloation to the core code will
eventually clean these out anyhow.
Link: https://lore.kernel.org/r/20200827163535.2632-1-rpearson@hpe.com
Signed-off-by: Bob Pearson <rpearson@hpe.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Creating rxe device on top of vlan interface will create a non-functional
device that has an empty gids table and can't be used for rdma cm
communication.
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, and attempts to
connect via this rdma device fail in cm_init_av_for_response because no
gids can be resolved.
Apparently, this behavior was implemented to fit the HW-RoCE devices that
create RoCE device per port, therefore RXE must behave the same like
HW-RoCE devices and create rxe device per real device only.
In order to communicate via a vlan interface, the user must use the gid
index of the vlan address instead of creating rxe over vlan.
Link: https://lore.kernel.org/r/20200811150415.3693-1-goody698@gmail.com
Signed-off-by: Mohammad Heib <goody698@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
'parent' sysfs reads will yield '\0' bytes when the interface name has 15
chars, and there will no "\n" output.
To reproduce, create one interface with 15 chars:
[root@test ~]# ip a s enp0s29u1u7u3c2
2: enp0s29u1u7u3c2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UNKNOWN group default qlen 1000
link/ether 02:21:28:57:47:17 brd ff:ff:ff:ff:ff:ff
inet6 fe80::ac41:338f:5bcd:c222/64 scope link noprefixroute
valid_lft forever preferred_lft forever
[root@test ~]# modprobe rdma_rxe
[root@test ~]# echo enp0s29u1u7u3c2 > /sys/module/rdma_rxe/parameters/add
[root@test ~]# cat /sys/class/infiniband/rxe0/parent
enp0s29u1u7u3c2[root@test ~]#
[root@test ~]# f="/sys/class/infiniband/rxe0/parent"
[root@test ~]# echo "$(<"$f")"
-bash: warning: command substitution: ignored null byte in input
enp0s29u1u7u3c2
Use scnprintf and PAGE_SIZE to fill the sysfs output buffer.
Cc: stable@vger.kernel.org
Fixes: 8700e3e7c4 ("Soft RoCE driver")
Link: https://lore.kernel.org/r/20200820153646.31316-1-yi.zhang@redhat.com
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Yi Zhang <yi.zhang@redhat.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The RoCE spec requires RoCE devices to support only the default pkey.
However the rxe driver maintains a 64 enties pkey table and uses only the
first entry. Remove the pkey table and hard code a table of length one
hard wired with the default pkey. Replace all checks of the pkey_table
with a comparison to the default_pkey instead.
Link: https://lore.kernel.org/r/20200721101618.686110-1-kamalheib1@gmail.com
Signed-off-by: Kamal Heib <kamalheib1@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
rxe_post_send_kernel() iterates over linked list of wr's, until the
wr->next ptr is NULL. However if we've got an interrupt after last wr is
posted, control may be returned to the code after send completion callback
is executed and wr memory is freed.
As a result, wr->next pointer may contain incorrect value leading to
panic. Store the wr->next on the stack before posting it.
Fixes: 8700e3e7c4 ("Soft RoCE driver")
Link: https://lore.kernel.org/r/20200716190340.23453-1-m.malygin@yadro.com
Signed-off-by: Mikhail Malygin <m.malygin@yadro.com>
Signed-off-by: Sergey Kojushev <s.kojushev@yadro.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
In the loopback tests, the following call trace occurs.
Call Trace:
__rxe_do_task+0x1a/0x30 [rdma_rxe]
rxe_qp_destroy+0x61/0xa0 [rdma_rxe]
rxe_destroy_qp+0x20/0x60 [rdma_rxe]
ib_destroy_qp_user+0xcc/0x220 [ib_core]
uverbs_free_qp+0x3c/0xc0 [ib_uverbs]
destroy_hw_idr_uobject+0x24/0x70 [ib_uverbs]
uverbs_destroy_uobject+0x43/0x1b0 [ib_uverbs]
uobj_destroy+0x41/0x70 [ib_uverbs]
__uobj_get_destroy+0x39/0x70 [ib_uverbs]
ib_uverbs_destroy_qp+0x88/0xc0 [ib_uverbs]
ib_uverbs_handler_UVERBS_METHOD_INVOKE_WRITE+0xb9/0xf0 [ib_uverbs]
ib_uverbs_cmd_verbs+0xb16/0xc30 [ib_uverbs]
The root cause is that the actual RDMA connection is not created in the
loopback tests and the rxe_match_dgid will fail randomly.
To fix this call trace which appear in the loopback tests, skip check of
the dgid.
Fixes: 8700e3e7c4 ("Soft RoCE driver")
Link: https://lore.kernel.org/r/20200630123605.446959-1-leon@kernel.org
Signed-off-by: Zhu Yanjun <yanjunz@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
