Introduce a new netlink command to allow rdma event monitoring.
The rdma events supported now are IB device
registration/unregistration and net device attachment/detachment.
Example output of rdma monitor and the commands which trigger
the events:
$ rdma monitor
$ rmmod mlx5_ib
[UNREGISTER] dev 1 rocep8s0f1
[UNREGISTER] dev 0 rocep8s0f0
$ modprobe mlx5_ib
[REGISTER] dev 2 mlx5_0
[NETDEV_ATTACH] dev 2 mlx5_0 port 1 netdev 4 eth2
[REGISTER] dev 3 mlx5_1
[NETDEV_ATTACH] dev 3 mlx5_1 port 1 netdev 5 eth3
$ devlink dev eswitch set pci/0000:08:00.0 mode switchdev
[UNREGISTER] dev 2 rocep8s0f0
[REGISTER] dev 4 mlx5_0
[NETDEV_ATTACH] dev 4 mlx5_0 port 30 netdev 4 eth2
$ echo 4 > /sys/class/net/eth2/device/sriov_numvfs
[NETDEV_ATTACH] dev 4 rdmap8s0f0 port 2 netdev 7 eth4
[NETDEV_ATTACH] dev 4 rdmap8s0f0 port 3 netdev 8 eth5
[NETDEV_ATTACH] dev 4 rdmap8s0f0 port 4 netdev 9 eth6
[NETDEV_ATTACH] dev 4 rdmap8s0f0 port 5 netdev 10 eth7
[REGISTER] dev 5 mlx5_0
[NETDEV_ATTACH] dev 5 mlx5_0 port 1 netdev 11 eth8
[REGISTER] dev 6 mlx5_0
[NETDEV_ATTACH] dev 6 mlx5_0 port 1 netdev 12 eth9
[REGISTER] dev 7 mlx5_0
[NETDEV_ATTACH] dev 7 mlx5_0 port 1 netdev 13 eth10
[REGISTER] dev 8 mlx5_0
[NETDEV_ATTACH] dev 8 mlx5_0 port 1 netdev 14 eth11
$ echo 0 > /sys/class/net/eth2/device/sriov_numvfs
[UNREGISTER] dev 5 rocep8s0f0v0
[UNREGISTER] dev 6 rocep8s0f0v1
[UNREGISTER] dev 7 rocep8s0f0v2
[UNREGISTER] dev 8 rocep8s0f0v3
[NETDEV_DETACH] dev 4 rdmap8s0f0 port 2
[NETDEV_DETACH] dev 4 rdmap8s0f0 port 3
[NETDEV_DETACH] dev 4 rdmap8s0f0 port 4
[NETDEV_DETACH] dev 4 rdmap8s0f0 port 5
Signed-off-by: Chiara Meiohas <cmeiohas@nvidia.com>
Signed-off-by: Michael Guralnik <michaelgur@nvidia.com>
Link: https://patch.msgid.link/20240909173025.30422-7-michaelgur@nvidia.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The IB layer provides a common interface to store and get net
devices associated to an IB device port (ib_device_set_netdev()
and ib_device_get_netdev()).
Previously, mlx5_ib stored and managed the associated net devices
internally.
Replace internal net device management in mlx5_ib with
ib_device_set_netdev() when attaching/detaching a net device and
ib_device_get_netdev() when retrieving the net device.
Export ib_device_get_netdev().
For mlx5 representors/PFs/VFs and lag creation we replace the netdev
assignments with the IB set/get netdev functions.
In active-backup mode lag the active slave net device is stored in the
lag itself. To assure the net device stored in a lag bond IB device is
the active slave we implement the following:
- mlx5_core: when modifying the slave of a bond we send the internal driver event
MLX5_DRIVER_EVENT_ACTIVE_BACKUP_LAG_CHANGE_LOWERSTATE.
- mlx5_ib: when catching the event call ib_device_set_netdev()
This patch also ensures the correct IB events are sent in switchdev lag.
While at it, when in multiport eswitch mode, only a single IB device is
created for all ports. The said IB device will receive all netdev events
of its VFs once loaded, thus to avoid overwriting the mapping of PF IB
device to PF netdev, ignore NETDEV_REGISTER events if the ib device has
already been mapped to a netdev.
Signed-off-by: Chiara Meiohas <cmeiohas@nvidia.com>
Signed-off-by: Michael Guralnik <michaelgur@nvidia.com>
Link: https://patch.msgid.link/20240909173025.30422-6-michaelgur@nvidia.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The caller of ib_device_get_netdev() relies on its result to accurately
match a given netdev with the ib device associated netdev.
ib_device_get_netdev returns NULL when the IB device associated
netdev is unregistering, preventing the caller of matching netdevs properly.
Thus, remove this optimization and return the netdev even if
it is undergoing unregistration, allowing matching by the caller.
This change ensures proper netdev matching and reference count handling
by the caller of ib_device_get_netdev/ib_device_set_netdev API.
Signed-off-by: Maher Sanalla <msanalla@nvidia.com>
Signed-off-by: Chiara Meiohas <cmeiohas@nvidia.com>
Signed-off-by: Michael Guralnik <michaelgur@nvidia.com>
Link: https://patch.msgid.link/20240909173025.30422-5-michaelgur@nvidia.com
Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
If a netdev has already been assigned, ib_device_set_netdev needs to
release the reference on the older netdev but it is mistakenly being
called for the new netdev. Fix it and in the process use netdev_put
to be symmetrical with the netdev_hold.
Fixes: 09f530f0c6 ("RDMA: Add netdevice_tracker to ib_device_set_netdev()")
Signed-off-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20240710203310.19317-1-dsahern@kernel.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
This patch adds 2 APIs, as well as driver operations to support adding
and deleting an IB sub device, which provides part of functionalities
of it's parent.
A sub device has a type; for a sub device with type "SMI", it provides
the smi capability through umad for its parent, meaning uverb is not
supported.
A sub device cannot live without a parent. So when a parent is
released, all it's sub devices are released as well.
Signed-off-by: Mark Zhang <markzhang@nvidia.com>
Link: https://lore.kernel.org/r/44253f7508b21eb2caefea3980c2bc072869116c.1718553901.git.leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Coccinelle reports a warning
WARNING: NULL check before dev_{put, hold} functions is not needed
The reason is the call netdev_{put, hold} of dev_{put,hold} will check NULL
There is no need to check before using dev_{put, hold}
Signed-off-by: Jules Irenge <jbi.octave@gmail.com>
Link: https://lore.kernel.org/r/ZjF1Eedxwhn4JSkz@octinomon.home
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The mad_client will be initialized in enable_device_and_get(), while the
devices_rwsem will be downgraded to a read semaphore. There is a window
that leads to the failed initialization for cm_client, since it can not
get matched mad port from ib_mad_port_list, and the matched mad port will
be added to the list after that.
mad_client | cm_client
------------------|--------------------------------------------------------
ib_register_device|
enable_device_and_get
down_write(&devices_rwsem)
xa_set_mark(&devices, DEVICE_REGISTERED)
downgrade_write(&devices_rwsem)
|
|ib_cm_init
|ib_register_client(&cm_client)
|down_read(&devices_rwsem)
|xa_for_each_marked (&devices, DEVICE_REGISTERED)
|add_client_context
|cm_add_one
|ib_register_mad_agent
|ib_get_mad_port
|__ib_get_mad_port
|list_for_each_entry(entry, &ib_mad_port_list, port_list)
|return NULL
|up_read(&devices_rwsem)
|
add_client_context|
ib_mad_init_device|
ib_mad_port_open |
list_add_tail(&port_priv->port_list, &ib_mad_port_list)
up_read(&devices_rwsem)
|
Fix it by using down_write(&devices_rwsem) in ib_register_client().
Fixes: d0899892ed ("RDMA/device: Provide APIs from the core code to help unregistration")
Link: https://lore.kernel.org/r/20240203035313.98991-1-lishifeng@sangfor.com.cn
Suggested-by: Jason Gunthorpe <jgg@ziepe.ca>
Signed-off-by: Shifeng Li <lishifeng@sangfor.com.cn>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Add support to dump SRQ resource in raw format. It enable drivers to
return the entire device specific SRQ context without setting each
field separately.
Example:
$ rdma res show srq -r
dev hns3 149000...
$ rdma res show srq -j -r
[{"ifindex":0,"ifname":"hns3","data":[149,0,0,...]}]
Signed-off-by: wenglianfa <wenglianfa@huawei.com>
Link: https://lore.kernel.org/r/20230918131110.3987498-3-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
If, for any reason, the open-coded arithmetic causes a wraparound,
the protection that `struct_size()` provides against potential integer
overflows is defeated. Fix this by hardening calls to `struct_size()`
with `size_add()`, `size_sub()` and `size_mul()`.
Fixes: 467f432a52 ("RDMA/core: Split port and device counter sysfs attributes")
Fixes: a4676388e2 ("RDMA/core: Simplify how the gid_attrs sysfs is created")
Fixes: e9dd5daf88 ("IB/umad: Refactor code to use cdev_device_add()")
Fixes: 324e227ea7 ("RDMA/device: Add ib_device_get_by_netdev()")
Fixes: 5aad26a7ea ("IB/core: Use struct_size() in kzalloc()")
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Link: https://lore.kernel.org/r/ZQdt4NsJFwwOYxUR@work
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Pull driver core updates from Greg KH:
"Here is the set of driver core and kernfs changes for 6.2-rc1.
The "big" change in here is the addition of a new macro,
container_of_const() that will preserve the "const-ness" of a pointer
passed into it.
The "problem" of the current container_of() macro is that if you pass
in a "const *", out of it can comes a non-const pointer unless you
specifically ask for it. For many usages, we want to preserve the
"const" attribute by using the same call. For a specific example, this
series changes the kobj_to_dev() macro to use it, allowing it to be
used no matter what the const value is. This prevents every subsystem
from having to declare 2 different individual macros (i.e.
kobj_const_to_dev() and kobj_to_dev()) and having the compiler enforce
the const value at build time, which having 2 macros would not do
either.
The driver for all of this have been discussions with the Rust kernel
developers as to how to properly mark driver core, and kobject,
objects as being "non-mutable". The changes to the kobject and driver
core in this pull request are the result of that, as there are lots of
paths where kobjects and device pointers are not modified at all, so
marking them as "const" allows the compiler to enforce this.
So, a nice side affect of the Rust development effort has been already
to clean up the driver core code to be more obvious about object
rules.
All of this has been bike-shedded in quite a lot of detail on lkml
with different names and implementations resulting in the tiny version
we have in here, much better than my original proposal. Lots of
subsystem maintainers have acked the changes as well.
Other than this change, included in here are smaller stuff like:
- kernfs fixes and updates to handle lock contention better
- vmlinux.lds.h fixes and updates
- sysfs and debugfs documentation updates
- device property updates
All of these have been in the linux-next tree for quite a while with
no problems"
* tag 'driver-core-6.2-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core: (58 commits)
device property: Fix documentation for fwnode_get_next_parent()
firmware_loader: fix up to_fw_sysfs() to preserve const
usb.h: take advantage of container_of_const()
device.h: move kobj_to_dev() to use container_of_const()
container_of: add container_of_const() that preserves const-ness of the pointer
driver core: fix up missed drivers/s390/char/hmcdrv_dev.c class.devnode() conversion.
driver core: fix up missed scsi/cxlflash class.devnode() conversion.
driver core: fix up some missing class.devnode() conversions.
driver core: make struct class.devnode() take a const *
driver core: make struct class.dev_uevent() take a const *
cacheinfo: Remove of_node_put() for fw_token
device property: Add a blank line in Kconfig of tests
device property: Rename goto label to be more precise
device property: Move PROPERTY_ENTRY_BOOL() a bit down
device property: Get rid of __PROPERTY_ENTRY_ARRAY_EL*SIZE*()
kernfs: fix all kernel-doc warnings and multiple typos
driver core: pass a const * into of_device_uevent()
kobject: kset_uevent_ops: make name() callback take a const *
kobject: kset_uevent_ops: make filter() callback take a const *
kobject: make kobject_namespace take a const *
...
This will cause an informative backtrace to print if the user of
ib_device_set_netdev() isn't careful about tearing down the ibdevice
before its the netdevice parent is destroyed. Such as like this:
unregister_netdevice: waiting for vlan0 to become free. Usage count = 2
leaked reference.
ib_device_set_netdev+0x266/0x730
siw_newlink+0x4e0/0xfd0
nldev_newlink+0x35c/0x5c0
rdma_nl_rcv_msg+0x36d/0x690
rdma_nl_rcv+0x2ee/0x430
netlink_unicast+0x543/0x7f0
netlink_sendmsg+0x918/0xe20
sock_sendmsg+0xcf/0x120
____sys_sendmsg+0x70d/0x8b0
___sys_sendmsg+0x11d/0x1b0
__sys_sendmsg+0xfa/0x1d0
do_syscall_64+0x35/0xb0
entry_SYSCALL_64_after_hwframe+0x63/0xcd
This will help debug the issues syzkaller is seeing.
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/0-v1-a7c81b3842ce+e5-netdev_tracker_jgg@nvidia.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The callbacks in struct class namespace() and get_ownership() do not
modify the struct device passed to them, so mark the pointer as constant
and fix up all callbacks in the kernel to have the correct function
signature.
This helps make it more obvious what calls and callbacks do, and do not,
modify structures passed to them.
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Link: https://lore.kernel.org/r/20221001165426.2690912-1-gregkh@linuxfoundation.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
KASAN reported a null-ptr-deref error:
KASAN: null-ptr-deref in range [0x0000000000000118-0x000000000000011f]
CPU: 1 PID: 379
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996)
RIP: 0010:destroy_workqueue+0x2f/0x740
RSP: 0018:ffff888016137df8 EFLAGS: 00000202
...
Call Trace:
ib_core_cleanup+0xa/0xa1 [ib_core]
__do_sys_delete_module.constprop.0+0x34f/0x5b0
do_syscall_64+0x3a/0x90
entry_SYSCALL_64_after_hwframe+0x63/0xcd
RIP: 0033:0x7fa1a0d221b7
...
It is because the fail of roce_gid_mgmt_init() is ignored:
ib_core_init()
roce_gid_mgmt_init()
gid_cache_wq = alloc_ordered_workqueue # fail
...
ib_core_cleanup()
roce_gid_mgmt_cleanup()
destroy_workqueue(gid_cache_wq)
# destroy an unallocated wq
Fix this by catching the fail of roce_gid_mgmt_init() in ib_core_init().
Fixes: 03db3a2d81 ("IB/core: Add RoCE GID table management")
Signed-off-by: Chen Zhongjin <chenzhongjin@huawei.com>
Link: https://lore.kernel.org/r/20221025024146.109137-1-chenzhongjin@huawei.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Currently, ib_find_gid() will stop searching after encountering the first
empty GID table entry. This behavior is wrong since neither IB nor RoCE
spec enforce tightly packed GID tables.
For example, when a valid GID entry exists at index N, and if a GID entry
is empty at index N-1, ib_find_gid() will fail to find the valid entry.
Fix it by making ib_find_gid() continue searching even after encountering
missing entries.
Fixes: 5eb620c81c ("IB/core: Add helpers for uncached GID and P_Key searches")
Link: https://lore.kernel.org/r/e55d331b96cecfc2cf19803d16e7109ea966882d.1639055490.git.leonro@nvidia.com
Signed-off-by: Avihai Horon <avihaih@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
An optional counter is a driver-specific counter that may be dynamically
enabled/disabled. This enhancement allows drivers to expose counters
which are, for example, mutually exclusive and cannot be enabled at the
same time, counters that might degrades performance, optional debug
counters, etc.
Optional counters are marked with IB_STAT_FLAG_OPTIONAL flag. They are not
exported in sysfs, and must be at the end of all stats, otherwise the
attr->show() in sysfs would get wrong indexes for hwcounters that are
behind optional counters.
Link: https://lore.kernel.org/r/20211008122439.166063-7-markzhang@nvidia.com
Signed-off-by: Aharon Landau <aharonl@nvidia.com>
Signed-off-by: Neta Ostrovsky <netao@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
ib_query_port() calls device->ops.query_port() to get the port
attributes. The method of querying is device driver specific. The same
function calls device->ops.query_gid() to get the GID and extract the
subnet_prefix (gid_prefix).
The GID and subnet_prefix are stored in a cache. But they do not get
read from the cache if the device is an Infiniband device. The
following change takes advantage of the cached subnet_prefix.
Testing with RDBMS has shown a significant improvement in performance
with this change.
Link: https://lore.kernel.org/r/20210712122625.1147-4-anand.a.khoje@oracle.com
Signed-off-by: Anand Khoje <anand.a.khoje@oracle.com>
Signed-off-by: Haakon Bugge <haakon.bugge@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Now that the port_groups data is being destroyed and managed by the core
code this restriction is no longer needed. All the ib_port_attrs are
compatible with the core's sysfs lifecycle.
When the main device is destroyed and moved to another namespace the
driver's port sysfs can be created/destroyed as well due to it now being a
simple attribute list.
Link: https://lore.kernel.org/r/afd8b676eace2821692d44489ff71856277c48d1.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
init_port was only being used to register sysfs attributes against the
port kobject. Now that all users are creating static attribute_group's we
can simply set the attribute_group list in the ops and the core code can
just handle it directly.
This makes all the sysfs management quite straightforward and prevents any
driver from abusing the naked port kobject in future because no driver
code can access it.
Link: https://lore.kernel.org/r/114f68f3d921460eafe14cea5a80ca65d81729c3.1623427137.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Current code uses many different types when dealing with a port of a RDMA
device: u8, unsigned int and u32. Switch to u32 to clean up the logic.
This allows us to make (at least) the core view consistent and use the
same type. Unfortunately not all places can be converted. Many uverbs
functions expect port to be u8 so keep those places in order not to break
UAPIs. HW/Spec defined values must also not be changed.
With the switch to u32 we now can support devices with more than 255
ports. U32_MAX is reserved to make control logic a bit easier to deal
with. As a device with U32_MAX ports probably isn't going to happen any
time soon this seems like a non issue.
When a device with more than 255 ports is created uverbs will report the
RDMA device as having 255 ports as this is the max currently supported.
The verbs interface is not changed yet because the IBTA spec limits the
port size in too many places to be u8 and all applications that relies in
verbs won't be able to cope with this change. At this stage, we are
extending the interfaces that are using vendor channel solely
Once the limitation is lifted mlx5 in switchdev mode will be able to have
thousands of SFs created by the device. As the only instance of an RDMA
device that reports more than 255 ports will be a representor device and
it exposes itself as a RAW Ethernet only device CM/MAD/IPoIB and other
ULPs aren't effected by this change and their sysfs/interfaces that are
exposes to userspace can remain unchanged.
While here cleanup some alignment issues and remove unneeded sanity
checks (mainly in rdmavt),
Link: https://lore.kernel.org/r/20210301070420.439400-1-leon@kernel.org
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
drivers/infiniband/core/device.c:859: warning: Function parameter or member 'dev' not described in 'ib_port_immutable_read'
drivers/infiniband/core/device.c:859: warning: Function parameter or member 'port' not described in 'ib_port_immutable_read'
Fixes: 7416790e22 ("RDMA/core: Introduce and use API to read port immutable data")
Link: https://lore.kernel.org/r/20210210151421.1108809-1-leon@kernel.org
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Currently mlx5 driver caches port GID table length for 2 ports. It is
also cached by IB core as port immutable data.
When mlx5 representor ports are present, which are usually more than 2,
invalid access to port_caps array can happen while validating the GID
table length which is only for 2 ports.
To avoid this, take help of the IB cores port immutable data by exposing
an API to read the port immutable fields.
Remove mlx5 driver's internal cache, thereby reduce code and data.
Link: https://lore.kernel.org/r/20210203130133.4057329-5-leon@kernel.org
Signed-off-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Fixes the following W=1 kernel build warning(s):
drivers/infiniband/core/device.c:1896: warning: Function parameter or member 'ibdev' not described in 'ib_get_client_nl_info'
drivers/infiniband/core/device.c:1896: warning: Function parameter or member 'client_name' not described in 'ib_get_client_nl_info'
drivers/infiniband/core/device.c:1896: warning: Function parameter or member 'res' not described in 'ib_get_client_nl_info'
drivers/infiniband/core/device.c:2328: warning: Function parameter or member 'nldev_cb' not described in 'ib_enum_all_devs'
drivers/infiniband/core/device.c:2328: warning: Function parameter or member 'skb' not described in 'ib_enum_all_devs'
Link: https://lore.kernel.org/r/20210118223929.512175-3-lee.jones@linaro.org
Cc: Doug Ledford <dledford@redhat.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: linux-rdma@vger.kernel.org
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
In procedure ib_register_device, procedure kobject_uevent is called
(advertising that the device is ready for userspace usage) even when
device_enable_and_get() returned an error.
As a result, various RDMA modules attempted to register for the device
even while the device driver was preparing to unregister the device.
Fix this by advertising the device availability only after enabling the
device succeeds.
Fixes: e7a5b4aafd ("RDMA/device: Don't fire uevent before device is fully initialized")
Link: https://lore.kernel.org/r/20201208073545.9723-3-leon@kernel.org
Suggested-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The CQ pool mechanism had two problems:
1. The CQ pool lists were uninitialized in the device registration error
flow. As a result, all the list pointers remained NULL. This caused
the kernel to crash (in procedure ib_cq_pool_destroy) when that error
flow was taken (and unregister called). The stack trace snippet:
BUG: kernel NULL pointer dereference, address: 0000000000000000
#PF: supervisor read access in kernel mode
#PF: error_code(0×0000) ? not-present page
PGD 0 P4D 0
Oops: 0000 [#1] SMP PTI
. . .
RIP: 0010:ib_cq_pool_destroy+0x1b/0×70 [ib_core]
. . .
Call Trace:
disable_device+0x9f/0×130 [ib_core]
__ib_unregister_device+0x35/0×90 [ib_core]
ib_register_device+0x529/0×610 [ib_core]
__mlx5_ib_add+0x3a/0×70 [mlx5_ib]
mlx5_add_device+0x87/0×1c0 [mlx5_core]
mlx5_register_interface+0x74/0xc0 [mlx5_core]
do_one_initcall+0x4b/0×1f4
do_init_module+0x5a/0×223
load_module+0x1938/0×1d40
2. At device unregister, when cleaning up the cq pool, the cq's in the
pool lists were freed, but the cq entries were left in the list.
The fix for the first issue is to initialize the cq pool lists when the
ib_device structure is allocated for a new device (in procedure
_ib_alloc_device).
The fix for the second problem is to delete cq entries from the pool lists
when cleaning up the cq pool.
In addition, procedure ib_cq_pool_destroy() is renamed to the more
appropriate name ib_cq_pool_cleanup().
Fixes: 4aa1615268 ("RDMA/core: Fix ordering of CQ pool destruction")
Link: https://lore.kernel.org/r/20201208073545.9723-2-leon@kernel.org
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Leon Romanovsky <leonro@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>
No driver sets it, and the core code sets a maximum mask, simply remove
it.
Disabled operations are now handled either by having a NULL ops pointer,
or by having the common driver callbacks check for unsupported extended
attributes.
Link: https://lore.kernel.org/r/9-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>
