linux-stable.git/net/core/devlink.c, branch v5.4.26

devlink: validate length of region addr/len

2020-03-18T06:17:44+00:00

[ Upstream commit ff3b63b8c299b73ac599b120653b47e275407656 ]

DEVLINK_ATTR_REGION_CHUNK_ADDR and DEVLINK_ATTR_REGION_CHUNK_LEN
lack entries in the netlink policy. Corresponding nla_get_u64()s
may read beyond the end of the message.

Fixes: 4e54795a27f5 ("devlink: Add support for region snapshot read command")
Signed-off-by: Jakub Kicinski 
Reviewed-by: Jiri Pirko 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

devlink: validate length of param values

2020-03-18T06:17:43+00:00

[ Upstream commit 8750939b6ad86abc3f53ec8a9683a1cded4a5654 ]

DEVLINK_ATTR_PARAM_VALUE_DATA may have different types
so it's not checked by the normal netlink policy. Make
sure the attribute length is what we expect.

Fixes: e3b7ca18ad7b ("devlink: Add param set command")
Signed-off-by: Jakub Kicinski 
Reviewed-by: Jiri Pirko 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

devlink: report 0 after hitting end in region read

2020-02-11T12:35:48+00:00

[ Upstream commit d5b90e99e1d51b7b5d2b74fbc4c2db236a510913 ]

commit fdd41ec21e15 ("devlink: Return right error code in case of errors
for region read") modified the region read code to report errors
properly in unexpected cases.

In the case where the start_offset and ret_offset match, it unilaterally
converted this into an error. This causes an issue for the "dump"
version of the command. In this case, the devlink region dump will
always report an invalid argument:

000000000000ffd0 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
000000000000ffe0 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
devlink answers: Invalid argument
000000000000fff0 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff

This occurs because the expected flow for the dump is to return 0 after
there is no further data.

The simplest fix would be to stop converting the error code to -EINVAL
if start_offset == ret_offset. However, avoid unnecessary work by
checking for when start_offset is larger than the region size and
returning 0 upfront.

Fixes: fdd41ec21e15 ("devlink: Return right error code in case of errors for region read")
Signed-off-by: Jacob Keller 
Acked-by: Jiri Pirko 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

devlink: Wait longer before warning about unset port type

2020-01-23T07:22:54+00:00

commit 4c582234ab3948d08a24c82eb1e00436aabacbc6 upstream.

The commit cited below causes devlink to emit a warning if a type was
not set on a devlink port for longer than 30 seconds to "prevent
misbehavior of drivers". This proved to be problematic when
unregistering the backing netdev. The flow is always:

devlink_port_type_clear()	// schedules the warning
unregister_netdev()		// blocking
devlink_port_unregister()	// cancels the warning

The call to unregister_netdev() can block for long periods of time for
various reasons: RTNL lock is contended, large amounts of configuration
to unroll following dismantle of the netdev, etc. This results in
devlink emitting a warning despite the driver behaving correctly.

In emulated environments (of future hardware) which are usually very
slow, the warning can also be emitted during port creation as more than
30 seconds can pass between the time the devlink port is registered and
when its type is set.

In addition, syzbot has hit this warning [1] 1974 times since 07/11/19
without being able to produce a reproducer. Probably because
reproduction depends on the load or other bugs (e.g., RTNL not being
released).

To prevent bogus warnings, increase the timeout to 1 hour.

[1] https://syzkaller.appspot.com/bug?id=e99b59e9c024a666c9f7450dc162a4b74d09d9cb

Fixes: 136bf27fc0e9 ("devlink: add warning in case driver does not set port type")
Signed-off-by: Ido Schimmel 
Reported-by: syzbot+b0a18ed7b08b735d2f41@syzkaller.appspotmail.com
Reported-by: Alex Veber 
Tested-by: Alex Veber 
Acked-by: Jiri Pirko 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

devlink: Add method for time-stamp on reporter's dump

2019-11-12T00:04:21+00:00

When setting the dump's time-stamp, use ktime_get_real in addition to
jiffies. This simplifies the user space implementation and bypasses
some inconsistent behavior with translating jiffies to current time.
The time taken is transformed into nsec, to comply with y2038 issue.

Fixes: c8e1da0bf923 ("devlink: Add health report functionality")
Signed-off-by: Aya Levin 
Acked-by: Jiri Pirko 
Acked-by: Arnd Bergmann 
Signed-off-by: David S. Miller

devlink: disallow reload operation during device cleanup

2019-11-10T03:38:36+00:00

There is a race between driver code that does setup/cleanup of device
and devlink reload operation that in some drivers works with the same
code. Use after free could we easily obtained by running:

while true; do
        echo "0000:00:10.0" >/sys/bus/pci/drivers/mlxsw_spectrum2/bind
        devlink dev reload pci/0000:00:10.0 &
        echo "0000:00:10.0" >/sys/bus/pci/drivers/mlxsw_spectrum2/unbind
done

Fix this by enabling reload only after setup of device is complete and
disabling it at the beginning of the cleanup process.

Reported-by: Ido Schimmel 
Fixes: 2d8dc5bbf4e7 ("devlink: Add support for reload")
Signed-off-by: Jiri Pirko 
Signed-off-by: David S. Miller

devlink: Fix error handling in param and info_get dumpit cb

2019-10-01T17:08:30+00:00

If any of the param or info_get op returns error, dumpit cb is
skipping to dump remaining params or info_get ops for all the
drivers.

Fix to not return if any of the param/info_get op returns error
as not supported and continue to dump remaining information.

v2: Modify the patch to return error, except for params/info_get
op that return -EOPNOTSUPP as suggested by Andrew Lunn. Also, modify
commit message to reflect the same.

Cc: Andrew Lunn 
Cc: Jiri Pirko 
Cc: Michael Chan 
Signed-off-by: Vasundhara Volam 
Acked-by: Jiri Pirko 
Reviewed-by: Andrew Lunn 
Signed-off-by: David S. Miller

net: devlink: move reload fail indication to devlink core and expose to user

2019-09-13T20:11:14+00:00

Currently the fact that devlink reload failed is stored in drivers.
Move this flag into devlink core. Also, expose it to the user.

Signed-off-by: Jiri Pirko 
Signed-off-by: David S. Miller

net: devlink: split reload op into two

2019-09-13T20:11:14+00:00

In order to properly implement failure indication during reload,
split the reload op into two ops, one for down phase and one for
up phase.

Signed-off-by: Jiri Pirko 
Signed-off-by: David S. Miller

devlink: add 'reset_dev_on_drv_probe' param

2019-09-10T16:29:26+00:00

Add the 'reset_dev_on_drv_probe' devlink parameter, controlling the
device reset policy on driver probe.

This parameter is useful in conjunction with the existing
'fw_load_policy' parameter.

Signed-off-by: Dirk van der Merwe 
Signed-off-by: Jakub Kicinski 
Signed-off-by: Simon Horman 
Signed-off-by: David S. Miller