Geliang Tang [Mon, 1 Feb 2021 23:09:07 +0000 (15:09 -0800)]
mptcp: drop *_max fields in mptcp_pm_data
This patch drops the per-msk values add_addr_signal_max,
add_addr_accept_max, local_addr_max and subflows_max fields in struct
mptcp_pm_data, uses the pernet *_max values instead. And adds four new
helpers to get the pernet *_max values separately.
Co-developed-by: Matthieu Baerts <matthieu.baerts@tessares.net> Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net> Signed-off-by: Geliang Tang <geliangtang@gmail.com> Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Geliang Tang [Mon, 1 Feb 2021 23:09:06 +0000 (15:09 -0800)]
mptcp: use WRITE_ONCE for the pernet *_max
This patch uses WRITE_ONCE() for all the pernet add_addr_signal_max,
add_addr_accept_max, local_addr_max and subflows_max fields in struct
pm_nl_pernet to avoid concurrency issues.
Signed-off-by: Geliang Tang <geliangtang@gmail.com> Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
====================
Add notifications when route hardware flags change
Routes installed to the kernel can be programmed to capable devices, in
which case they are marked with one of two flags. RTM_F_OFFLOAD for
routes that offload traffic from the kernel and RTM_F_TRAP for routes
that trap packets to the kernel for processing (e.g., host routes).
These flags are of interest to routing daemons since they would like to
delay advertisement of routes until they are installed in hardware. This
allows them to avoid packet loss or misrouted packets. Currently,
routing daemons do not receive any notifications when these flags are
changed, requiring them to poll the kernel tables for changes which is
inefficient.
This series addresses the issue by having the kernel emit RTM_NEWROUTE
notifications whenever these flags change. The behavior is controlled by
two sysctls (net.ipv4.fib_notify_on_flag_change and
net.ipv6.fib_notify_on_flag_change) that default to 0 (no
notifications).
Note that even if route installation in hardware is improved to be more
synchronous, these notifications are still of interest. For example, a
multipath route can change from RTM_F_OFFLOAD to RTM_F_TRAP if its
neighbours become invalid. A routing daemon can choose to withdraw /
replace the route in that case. In addition, the deletion of a route
from the kernel can prompt the installation of an identical route
(already in kernel, with an higher metric) to hardware.
For testing purposes, netdevsim is aligned to simulate a "real" driver
that programs routes to hardware.
Series overview:
Patches #1-#2 align netdevsim to perform route programming in a
non-atomic context
Patches #3-#5 add sysctl to control IPv4 notifications
Patches #6-#8 add sysctl to control IPv6 notifications
Patch #9 extends existing fib tests to set sysctls before running tests
Patch #10 adds test for fib notifications over netdevsim
====================
Amit Cohen [Mon, 1 Feb 2021 19:47:57 +0000 (21:47 +0200)]
selftests: netdevsim: Add fib_notifications test
Add test to check fib notifications behavior.
The test checks route addition, route deletion and route replacement for
both IPv4 and IPv6.
When fib_notify_on_flag_change=0, expect single notification for route
addition/deletion/replacement.
When fib_notify_on_flag_change=1, expect:
- two notification for route addition/replacement, first without RTM_F_TRAP
and second with RTM_F_TRAP.
- single notification for route deletion.
$ ./fib_notifications.sh
TEST: IPv4 route addition [ OK ]
TEST: IPv4 route deletion [ OK ]
TEST: IPv4 route replacement [ OK ]
TEST: IPv6 route addition [ OK ]
TEST: IPv6 route deletion [ OK ]
TEST: IPv6 route replacement [ OK ]
Signed-off-by: Amit Cohen <amcohen@nvidia.com> Signed-off-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: David Ahern <dsahern@kernel.org> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Amit Cohen [Mon, 1 Feb 2021 19:47:56 +0000 (21:47 +0200)]
selftests: Extend fib tests to run with and without flags notifications
Run the test cases with both `fib_notify_on_flag_change` sysctls set to
'1', and then with both sysctls set to '0' to verify there are no
regressions in the test when notifications are added.
Signed-off-by: Amit Cohen <amcohen@nvidia.com> Signed-off-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Amit Cohen [Mon, 1 Feb 2021 19:47:55 +0000 (21:47 +0200)]
net: ipv6: Emit notification when fib hardware flags are changed
After installing a route to the kernel, user space receives an
acknowledgment, which means the route was installed in the kernel,
but not necessarily in hardware.
The asynchronous nature of route installation in hardware can lead
to a routing daemon advertising a route before it was actually installed in
hardware. This can result in packet loss or mis-routed packets until the
route is installed in hardware.
It is also possible for a route already installed in hardware to change
its action and therefore its flags. For example, a host route that is
trapping packets can be "promoted" to perform decapsulation following
the installation of an IPinIP/VXLAN tunnel.
Emit RTM_NEWROUTE notifications whenever RTM_F_OFFLOAD/RTM_F_TRAP flags
are changed. The aim is to provide an indication to user-space
(e.g., routing daemons) about the state of the route in hardware.
Introduce a sysctl that controls this behavior.
Keep the default value at 0 (i.e., do not emit notifications) for several
reasons:
- Multiple RTM_NEWROUTE notification per-route might confuse existing
routing daemons.
- Convergence reasons in routing daemons.
- The extra notifications will negatively impact the insertion rate.
- Not all users are interested in these notifications.
Move fib6_info_hw_flags_set() to C file because it is no longer a short
function.
Signed-off-by: Amit Cohen <amcohen@nvidia.com> Signed-off-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: David Ahern <dsahern@kernel.org> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Amit Cohen [Mon, 1 Feb 2021 19:47:52 +0000 (21:47 +0200)]
net: ipv4: Emit notification when fib hardware flags are changed
After installing a route to the kernel, user space receives an
acknowledgment, which means the route was installed in the kernel,
but not necessarily in hardware.
The asynchronous nature of route installation in hardware can lead to a
routing daemon advertising a route before it was actually installed in
hardware. This can result in packet loss or mis-routed packets until the
route is installed in hardware.
It is also possible for a route already installed in hardware to change
its action and therefore its flags. For example, a host route that is
trapping packets can be "promoted" to perform decapsulation following
the installation of an IPinIP/VXLAN tunnel.
Emit RTM_NEWROUTE notifications whenever RTM_F_OFFLOAD/RTM_F_TRAP flags
are changed. The aim is to provide an indication to user-space
(e.g., routing daemons) about the state of the route in hardware.
Introduce a sysctl that controls this behavior.
Keep the default value at 0 (i.e., do not emit notifications) for several
reasons:
- Multiple RTM_NEWROUTE notification per-route might confuse existing
routing daemons.
- Convergence reasons in routing daemons.
- The extra notifications will negatively impact the insertion rate.
- Not all users are interested in these notifications.
Signed-off-by: Amit Cohen <amcohen@nvidia.com> Acked-by: Roopa Prabhu <roopa@nvidia.com> Signed-off-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: David Ahern <dsahern@kernel.org> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Amit Cohen [Mon, 1 Feb 2021 19:47:49 +0000 (21:47 +0200)]
netdevsim: fib: Perform the route programming in a non-atomic context
Currently, netdevsim implements dummy FIB offload and marks notified
routes with RTM_F_TRAP flag. netdevsim does not defer route notifications
to a work queue because it does not need to program any hardware.
Given that netdevsim's purpose is to both give an example implementation
and allow developers to test their code, align netdevsim to a "real"
hardware device driver like mlxsw and have it also perform the route
"programming" in a non-atomic context.
It will be used to test route flags notifications which will be added in
the next patches.
The following changes are needed when route handling is performed in WQ:
- Handle the accounting in the main context, to be able to return an
error for adding route when all the routes are used.
For FIB_EVENT_ENTRY_REPLACE increase the counter before scheduling
the delayed work, and in case that this event replaces an existing route,
decrease the counter as part of the delayed work.
- For IPv6, cannot use fen6_info->rt->fib6_siblings list because it
might be changed during handling the delayed work.
Save an array with the nexthops as part of fib6_event struct, and take
a reference for each nexthop to prevent them from being freed while
event is queued.
- Change GFP_ATOMIC allocations to GFP_KERNEL.
- Use single work item that is handling a list of ordered routes.
Handling routes must be processed in the order they were submitted to
avoid logical errors that could lead to unexpected failures.
Signed-off-by: Amit Cohen <amcohen@nvidia.com> Signed-off-by: Ido Schimmel <idosch@nvidia.com> Acked-by: David Ahern <dsahern@kernel.org> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Amit Cohen [Mon, 1 Feb 2021 19:47:48 +0000 (21:47 +0200)]
netdevsim: fib: Convert the current occupancy to an atomic variable
When route is added/deleted, the appropriate counter is increased/decreased
to maintain number of routes.
User can limit the number of routes and then according to the appropriate
counter, adding more routes than the limitation is forbidden.
Currently, there is one lock which protects hashtable, list and accounting.
Handling the counters will be performed from both atomic context and
non-atomic context, while the hashtable and the list will be used only from
non-atomic context and therefore will be protected by a separate lock.
Protect accounting by using an atomic variable, so lock is not needed.
v2:
* Use atomic64_sub() in nsim_nexthop_account()'s error path
Signed-off-by: Amit Cohen <amcohen@nvidia.com> Signed-off-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: David Ahern <dsahern@kernel.org> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
====================
net: ipa: don't disable NAPI in suspend
This is version 2 of a series that reworks the order in which things
happen during channel stop and suspend (and start and resume), in
order to address a hang that has been observed during suspend.
The introductory message on the first version of the series gave
some history which is omitted here.
The end result of this series is that we only enable NAPI and the
I/O completion interrupt on a channel when we start the channel for
the first time. And we only disable them when stopping the channel
"for good." In other words, NAPI and the completion interrupt
remain enabled while a channel is stopped for suspend.
One comment on version 1 of the series suggested *not* returning
early on success in a function, instead having both success and
error paths return from the same point at the end of the function
block. This has been addressed in this version.
In addition, this version consolidates things a little bit, but the
net result of the series is exactly the same as version 1 (with the
exception of the return fix mentioned above).
First, patch 6 in the first version was a small step to make patch 7
easier to understand. The two have been combined now.
Second, previous version moved (and for suspend/resume, eliminated)
I/O completion interrupt and NAPI disable/enable control in separate
steps (patches). Now both are moved around together in patch 5 and
6, which eliminates the need for the final (NAPI-only) patch.
I won't repeat the patch summaries provided in v1:
https://lore.kernel.org/netdev/20210129202019.2099259-1-elder@linaro.org/
Many thanks to Willem de Bruijn for his thoughtful input.
====================
Alex Elder [Mon, 1 Feb 2021 17:28:50 +0000 (11:28 -0600)]
net: ipa: expand last transaction check
Transactions to send data for a network device can be allocated at
any time up until the point the TX queue is stopped. It is possible
for ipa_start_xmit() to be called in one context just before a
the transmit queue is stopped in another.
Update gsi_channel_trans_last() so that for TX channels the
allocated and pending transaction lists are checked--in addition
to the completed and polled lists--to determine the "last"
transaction. This means any transaction that has been allocated
before the TX queue is stopped will be allowed to complete before
we conclude the channel is quiesced.
Rework the function a bit to use a list pointer and gotos.
Signed-off-by: Alex Elder <elder@linaro.org> Acked-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 1 Feb 2021 17:28:49 +0000 (11:28 -0600)]
net: ipa: don't disable interrupt on suspend
No completion interrupts will occur while an endpoint is suspended,
nor when a channel has been stopped for suspend. So there's no need
to disable the interrupt during suspend and re-enable it when
resuming. Without any interrupts occurring, there is no need to
disable/re-enable NAPI for channel suspend/resume either.
We'll only enable NAPI and the interrupt when we first start the
channel, and disable it again only when it's "really" stopped.
To accomplish this, move the enable/disable calls out of
__gsi_channel_start() and __gsi_channel_stop(), and into
gsi_channel_start() and gsi_channel_stop() instead.
Add a call to napi_synchronize() to gsi_channel_suspend(), to ensure
NAPI polling is done before moving on.
Signed-off-by: Alex Elder <elder@linaro.org> Acked-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 1 Feb 2021 17:28:48 +0000 (11:28 -0600)]
net: ipa: disable interrupt and NAPI after channel stop
Disable both the I/O completion interrupt and NAPI polling on a
channel *after* we successfully stop it rather than before. This
ensures a completion occurring just before the channel is stopped
gets processed.
Enable NAPI polling and the interrupt *before* starting a channel
rather than after, to be symmetric. A stopped channel won't
generate any completion interrupts anyway.
Enable NAPI before the interrupt and disable it afterward.
Signed-off-by: Alex Elder <elder@linaro.org> Acked-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 1 Feb 2021 17:28:47 +0000 (11:28 -0600)]
net: ipa: kill gsi_channel_freeze() and gsi_channel_thaw()
Open-code gsi_channel_freeze() and gsi_channel_thaw() in all callers
and get rid of these two functions. This is part of reworking the
sequence of things done during channel suspend/resume and start/stop.
Signed-off-by: Alex Elder <elder@linaro.org> Acked-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 1 Feb 2021 17:28:46 +0000 (11:28 -0600)]
net: ipa: introduce __gsi_channel_start()
Create a new function that does most of the work of starting a
channel. What's different is that it takes a flag indicating
whether the channel should really be started or not. Create
another new function __gsi_channel_stop() that behaves similarly.
IPA v3.5.1 implements suspend using a special SUSPEND endpoint
setting. If the endpoint is suspended when an I/O completes on the
underlying GSI channel, a SUSPEND interrupt is generated.
Newer versions of IPA do not implement the SUSPEND endpoint mode.
Instead, endpoint suspend is implemented by simply stopping the
underlying GSI channel. In this case, a completing I/O on a
*stopped* channel causes the SUSPEND interrupt condition.
These new functions put all activity related to starting or stopping
a channel (including "thawing/freezing" the channel) in one place,
whether or not the channel is actually started or stopped.
Signed-off-by: Alex Elder <elder@linaro.org> Acked-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Marco Elver [Mon, 1 Feb 2021 16:04:20 +0000 (17:04 +0100)]
net: fix up truesize of cloned skb in skb_prepare_for_shift()
Avoid the assumption that ksize(kmalloc(S)) == ksize(kmalloc(S)): when
cloning an skb, save and restore truesize after pskb_expand_head(). This
can occur if the allocator decides to service an allocation of the same
size differently (e.g. use a different size class, or pass the
allocation on to KFENCE).
Because truesize is used for bookkeeping (such as sk_wmem_queued), a
modified truesize of a cloned skb may result in corrupt bookkeeping and
relevant warnings (such as in sk_stream_kill_queues()).
Davide Caratti [Mon, 1 Feb 2021 13:05:26 +0000 (14:05 +0100)]
mptcp: fix length of MP_PRIO suboption
With version 0 of the protocol it was legal to encode the 'Subflow Id' in
the MP_PRIO suboption, to specify which subflow would change its 'Backup'
flag. This has been removed from v1 specification: thus, according to RFC
8684 ยง3.3.8, the resulting 'Length' for MP_PRIO changed from 4 to 3 byte.
Current Linux generates / parses MP_PRIO according to the old spec, using
'Length' equal to 4, and hardcoding 1 as 'Subflow Id'; RFC compliance can
improve if we change 'Length' in other to become 3, leaving a 'Nop' after
the MP_PRIO suboption. In this way the kernel will emit and accept *only*
MP_PRIO suboptions that are compliant to version 1 of the MPTCP protocol.
unpatched 5.11-rc kernel:
[root@bottarga ~]# tcpdump -tnnr unpatched.pcap | grep prio
reading from file unpatched.pcap, link-type LINUX_SLL (Linux cooked v1)
dropped privs to tcpdump
IP 10.0.3.2.48433 > 10.0.1.1.10006: Flags [.], ack 1, win 502, options [nop,nop,TS val 4032325513 ecr 1876514270,mptcp prio non-backup id 1,mptcp dss ack 14084896651682217737], length 0
Changes since v2:
- when accounting for option space, don't increment 'TCPOLEN_MPTCP_PRIO'
and use 'TCPOLEN_MPTCP_PRIO_ALIGN' instead, thanks to Matthieu Baerts.
Changes since v1:
- refactor patch to avoid using 'TCPOLEN_MPTCP_PRIO' with its old value,
thanks to Geliang Tang.
Fixes: 641df33b6ffc ("mptcp: add the outgoing MP_PRIO support") Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Reviewed-by: Matthieu Baerts <matthieu.baerts@tessares.net> Signed-off-by: Davide Caratti <dcaratti@redhat.com> Reviewed-by: Matteo Croce <mcroce@linux.microsoft.com> Link: https://lore.kernel.org/r/846cdd41e6ad6ec88ef23fee1552ab39c2f5a3d1.1612184361.git.dcaratti@redhat.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This updates the remaining callers of tasklet_init() in drivers/net
to the new API introduced in
commit ad117fab14f5 ("tasklet: Introduce new initialization API")
All changes are done by coccinelle using the following semantic patch.
Coccinelle needs a little help parsing drivers/net/arcnet/arcnet.c
@ match @
type T;
T *container;
identifier tasklet;
identifier callback;
@@
tasklet_init(&container->tasklet, callback, (unsigned long)container);
@ patch1 depends on match @
type match.T;
identifier match.tasklet;
identifier match.callback;
identifier data;
identifier container;
@@
-void callback(unsigned long data)
+void callback(struct tasklet_struct *t)
{
...
- T *container = (T *)data;
+ T *container = from_tasklet(container, t, tasklet);
...
}
@ patch2 depends on match @
type match.T;
identifier match.tasklet;
identifier match.callback;
identifier data;
identifier container;
@@
-void callback(unsigned long data)
+void callback(struct tasklet_struct *t)
{
...
- T *container;
+ T *container = from_tasklet(container, t, tasklet);
...
- container = (T *)data;
...
}
Linus Torvalds [Tue, 2 Feb 2021 18:26:09 +0000 (10:26 -0800)]
Merge tag 'net-5.11-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Pull networking fixes from Jakub Kicinski:
"Networking fixes for 5.11-rc7, including fixes from bpf and mac80211
trees.
Current release - regressions:
- ip_tunnel: fix mtu calculation
- mlx5: fix function calculation for page trees
Previous releases - regressions:
- vsock: fix the race conditions in multi-transport support
- neighbour: prevent a dead entry from updating gc_list
- dsa: mv88e6xxx: override existent unicast portvec in port_fdb_add
Previous releases - always broken:
- bpf, cgroup: two copy_{from,to}_user() warn_on_once splats for BPF
cgroup getsockopt infra when user space is trying to race against
optlen, from Loris Reiff.
- mac80211: fix station rate table updates on assoc
- r8169: work around RTL8125 UDP HW bug
- igc: report speed and duplex as unknown when device is runtime
suspended
- rxrpc: fix deadlock around release of dst cached on udp tunnel"
* tag 'net-5.11-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (36 commits)
net: hsr: align sup_multicast_addr in struct hsr_priv to u16 boundary
net: ipa: fix two format specifier errors
net: ipa: use the right accessor in ipa_endpoint_status_skip()
net: ipa: be explicit about endianness
net: ipa: add a missing __iomem attribute
net: ipa: pass correct dma_handle to dma_free_coherent()
r8169: fix WoL on shutdown if CONFIG_DEBUG_SHIRQ is set
net/rds: restrict iovecs length for RDS_CMSG_RDMA_ARGS
net: mvpp2: TCAM entry enable should be written after SRAM data
net: lapb: Copy the skb before sending a packet
net/mlx5e: Release skb in case of failure in tc update skb
net/mlx5e: Update max_opened_tc also when channels are closed
net/mlx5: Fix leak upon failure of rule creation
net/mlx5: Fix function calculation for page trees
docs: networking: swap words in icmp_errors_use_inbound_ifaddr doc
udp: ipv4: manipulate network header of NATed UDP GRO fraglist
net: ip_tunnel: fix mtu calculation
vsock: fix the race conditions in multi-transport support
net: sched: replaced invalid qdisc tree flush helper in qdisc_replace
ibmvnic: device remove has higher precedence over reset
...
Andreas Oetken [Tue, 2 Feb 2021 09:03:04 +0000 (10:03 +0100)]
net: hsr: align sup_multicast_addr in struct hsr_priv to u16 boundary
sup_multicast_addr is passed to ether_addr_equal for address comparison
which casts the address inputs to u16 leading to an unaligned access.
Aligning the sup_multicast_addr to u16 boundary fixes the issue.
Jakub Kicinski [Tue, 2 Feb 2021 16:51:25 +0000 (08:51 -0800)]
Merge tag 'mlx5-fixes-2021-02-01' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux
Saeed Mahameed says:
====================
mlx5 fixes 2021-02-01
Please note the first patch in this series
("Fix function calculation for page trees") is fixing a regression
due to previous fix in net which you didn't include in your previous
rc pr. So I hope this series will make it into your next rc pr,
so mlx5 won't be broken in the next rc.
* tag 'mlx5-fixes-2021-02-01' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux:
net/mlx5e: Release skb in case of failure in tc update skb
net/mlx5e: Update max_opened_tc also when channels are closed
net/mlx5: Fix leak upon failure of rule creation
net/mlx5: Fix function calculation for page trees
====================
Jakub Kicinski [Tue, 2 Feb 2021 16:48:17 +0000 (08:48 -0800)]
Merge branch 'net-ipa-a-few-bug-fixes'
Alex Elder says:
====================
net: ipa: a few bug fixes
This series fixes four minor bugs. The first two are things that
sparse points out. All four are very simple and each patch should
explain itself pretty well.
====================
Alex Elder [Mon, 1 Feb 2021 23:26:08 +0000 (17:26 -0600)]
net: ipa: use the right accessor in ipa_endpoint_status_skip()
When extracting the destination endpoint ID from the status in
ipa_endpoint_status_skip(), u32_get_bits() is used. This happens to
work, but it's wrong: the structure field is only 8 bits wide
instead of 32.
Fix this by using u8_get_bits() to get the destination endpoint ID.
Signed-off-by: Alex Elder <elder@linaro.org> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 1 Feb 2021 23:26:07 +0000 (17:26 -0600)]
net: ipa: be explicit about endianness
Sparse warns that the assignment of the metadata mask for a QMAP
endpoint in ipa_endpoint_init_hdr_metadata_mask() is a bad
assignment. We know we want the mask value to be big endian, even
though the value we write is in host byte order. Use a __force
tag to indicate we really mean it.
Signed-off-by: Alex Elder <elder@linaro.org> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Heiner Kallweit [Mon, 1 Feb 2021 20:50:56 +0000 (21:50 +0100)]
r8169: fix WoL on shutdown if CONFIG_DEBUG_SHIRQ is set
So far phy_disconnect() is called before free_irq(). If CONFIG_DEBUG_SHIRQ
is set and interrupt is shared, then free_irq() creates an "artificial"
interrupt by calling the interrupt handler. The "link change" flag is set
in the interrupt status register, causing phylib to eventually call
phy_suspend(). Because the net_device is detached from the PHY already,
the PHY driver can't recognize that WoL is configured and powers down the
PHY.
net/rds: restrict iovecs length for RDS_CMSG_RDMA_ARGS
syzbot found WARNING in rds_rdma_extra_size [1] when RDS_CMSG_RDMA_ARGS
control message is passed with user-controlled
0x40001 bytes of args->nr_local, causing order >= MAX_ORDER condition.
The exact value 0x40001 can be checked with UIO_MAXIOV which is 0x400.
So for kcalloc() 0x400 iovecs with sizeof(struct rds_iovec) = 0x10
is the closest limit, with 0x10 leftover.
Same condition is currently done in rds_cmsg_rdma_args().
Xie He [Mon, 1 Feb 2021 05:57:06 +0000 (21:57 -0800)]
net: lapb: Copy the skb before sending a packet
When sending a packet, we will prepend it with an LAPB header.
This modifies the shared parts of a cloned skb, so we should copy the
skb rather than just clone it, before we prepend the header.
In "Documentation/networking/driver.rst" (the 2nd point), it states
that drivers shouldn't modify the shared parts of a cloned skb when
transmitting.
The "dev_queue_xmit_nit" function in "net/core/dev.c", which is called
when an skb is being sent, clones the skb and sents the clone to
AF_PACKET sockets. Because the LAPB drivers first remove a 1-byte
pseudo-header before handing over the skb to us, if we don't copy the
skb before prepending the LAPB header, the first byte of the packets
received on AF_PACKET sockets can be corrupted.
Jakub Kicinski [Tue, 2 Feb 2021 16:37:00 +0000 (08:37 -0800)]
Merge tag 'mac80211-for-net-2021-02-02' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211
Johannes Berg says:
====================
Two fixes:
- station rate tables were not updated correctly
after association, leading to bad configuration
- rtl8723bs (staging) was initializing data incorrectly
after the previous fix and needed to move the init
later
* tag 'mac80211-for-net-2021-02-02' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211:
staging: rtl8723bs: Move wiphy setup to after reading the regulatory settings from the chip
mac80211: fix station rate table updates on assoc
====================
net/mlx5e: Update max_opened_tc also when channels are closed
max_opened_tc is used for stats, so that potentially non-zero stats
won't disappear when num_tc decreases. However, mlx5e_setup_tc_mqprio
fails to update it in the flow where channels are closed.
This commit fixes it. The new value of priv->channels.params.num_tc is
always checked on exit. In case of errors it will just be the old value,
and in case of success it will be the updated value.
Daniel Jurgens [Mon, 1 Feb 2021 16:11:10 +0000 (18:11 +0200)]
net/mlx5: Fix function calculation for page trees
The function calculation always results in a value of 0. This works
generally, but when the release all pages feature is enabled it will
result in crashes.
Fixes: 9bb0189d6ae5 ("net/mlx5: Maintain separate page trees for ECPF and PF functions") Signed-off-by: Daniel Jurgens <danielj@nvidia.com> Reported-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Jakub Kicinski [Tue, 2 Feb 2021 04:23:44 +0000 (20:23 -0800)]
Merge branch '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2021-02-01
This series contains updates to igc and i40e drivers.
Kai-Heng Feng fixes igc to report unknown speed and duplex during suspend
as an attempted read will cause errors.
Kevin Lo sets the default value to -IGC_ERR_NVM instead of success for
writing shadow RAM as this could miss a timeout. Also propagates the return
value for Flow Control configuration to properly pass on errors for igc.
Aleksandr reverts commit 346f948a5385 ("i40e: don't report link up for a VF
who hasn't enabled queues") as this can cause link flapping.
* '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
i40e: Revert "i40e: don't report link up for a VF who hasn't enabled queues"
igc: check return value of ret_val in igc_config_fc_after_link_up
igc: set the default return value to -IGC_ERR_NVM in igc_write_nvm_srwr
igc: Report speed and duplex as unknown when device is runtime suspended
====================
rmb() can be removed since:
1. pending_scrq() has dma_rmb() at the function end;
2. dma_rmb(), though weaker, is enough here.
Signed-off-by: Lijun Pan <ljp@linux.ibm.com> Acked-by: Dwip Banerjee <dnbanerg@us.ibm.com> Acked-by: Thomas Falcon <tlfalcon@linux.ibm.com> Reviewed-by: Brian King <brking@linux.vnet.ibm.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Lijun Pan [Sat, 30 Jan 2021 01:19:04 +0000 (19:19 -0600)]
ibmvnic: rework to ensure SCRQ entry reads are properly ordered
Move the dma_rmb() between pending_scrq() and ibmvnic_next_scrq()
into the end of pending_scrq() to save the duplicated code since
this dma_rmb will be used 3 times.
Signed-off-by: Lijun Pan <ljp@linux.ibm.com> Acked-by: Thomas Falcon <tlfalcon@linux.ibm.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Dongseok Yi [Fri, 29 Jan 2021 23:13:27 +0000 (08:13 +0900)]
udp: ipv4: manipulate network header of NATed UDP GRO fraglist
UDP/IP header of UDP GROed frag_skbs are not updated even after NAT
forwarding. Only the header of head_skb from ip_finish_output_gso ->
skb_gso_segment is updated but following frag_skbs are not updated.
A call path skb_mac_gso_segment -> inet_gso_segment ->
udp4_ufo_fragment -> __udp_gso_segment -> __udp_gso_segment_list
does not try to update UDP/IP header of the segment list but copy
only the MAC header.
Update port, addr and check of each skb of the segment list in
__udp_gso_segment_list. It covers both SNAT and DNAT.
Vadim Fedorenko [Fri, 29 Jan 2021 22:27:47 +0000 (01:27 +0300)]
net: ip_tunnel: fix mtu calculation
dev->hard_header_len for tunnel interface is set only when header_ops
are set too and already contains full overhead of any tunnel encapsulation.
That's why there is not need to use this overhead twice in mtu calc.
Fixes: c5ed160e7b57 ("ip_gre: set dev->hard_header_len and dev->needed_headroom properly") Reported-by: Slava Bacherikov <mail@slava.cc> Signed-off-by: Vadim Fedorenko <vfedorenko@novek.ru> Link: https://lore.kernel.org/r/1611959267-20536-1-git-send-email-vfedorenko@novek.ru Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alexander Popov [Mon, 1 Feb 2021 08:47:19 +0000 (11:47 +0300)]
vsock: fix the race conditions in multi-transport support
There are multiple similar bugs implicitly introduced by the
commit e29ab509ea9a91e6 ("vsock: add multi-transports support") and
commit 6373aba309c5f8ff ("vsock: prevent transport modules unloading").
The bug pattern:
[1] vsock_sock.transport pointer is copied to a local variable,
[2] lock_sock() is called,
[3] the local variable is used.
VSOCK multi-transport support introduced the race condition:
vsock_sock.transport value may change between [1] and [2].
Let's copy vsock_sock.transport pointer to local variables after
the lock_sock() call.
Jakub Kicinski [Tue, 2 Feb 2021 02:50:12 +0000 (18:50 -0800)]
Merge tag 'mlx5-dr-2021-01-29' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux
Saeed Mahameed says:
====================
mlx5-dr-2021-01-29
Add support for Connect-X6DX Software steering
This series adds SW Steering support for Connect-X6DX.
Since the STE and actions formats are different on this new HW,
we implemented the HW specific STEv1 layer on the infrastructure
implemented in previous mlx5 DR patchset to support all the
functionalities as previous devices.
Most of the code in this series very is low level HW specific, we
implement the function pointers for the generic SW steering layer.
* tag 'mlx5-dr-2021-01-29' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux:
net/mlx5: DR, Allow SW steering for sw_owner_v2 devices
net/mlx5: DR, Copy all 64B whenever replacing STE in the head of miss-list
net/mlx5: DR, Use HW specific logic API when writing STE
net/mlx5: DR, Use the right size when writing partial STE into HW
net/mlx5: DR, Add STEv1 modify header logic
net/mlx5: DR, Add STEv1 action apply logic
net/mlx5: DR, Add STEv1 setters and getters
net/mlx5: DR, Allow native protocol support for HW STEv1
net/mlx5: DR, Add HW STEv1 match logic
net/mlx5: DR, Add match STEv1 structs to ifc
net/mlx5: DR, Fix potential shift wrapping of 32-bit value
====================
net: sched: replaced invalid qdisc tree flush helper in qdisc_replace
Commit a9ee80847d0d ("net: sched: introduce and use qdisc tree flush/purge helpers")
introduced qdisc tree flush/purge helpers, but erroneously used flush helper
instead of purge helper in qdisc_replace function.
This issue was found in our CI, that tests various qdisc setups by configuring
qdisc and sending data through it. Call of invalid helper sporadically leads
to corruption of vt_tree/cf_tree of hfsc_class that causes kernel oops:
Lijun Pan [Fri, 29 Jan 2021 04:34:01 +0000 (22:34 -0600)]
ibmvnic: device remove has higher precedence over reset
Returning -EBUSY in ibmvnic_remove() does not actually hold the
removal procedure since driver core doesn't care for the return
value (see __device_release_driver() in drivers/base/dd.c
calling dev->bus->remove()) though vio_bus_remove
(in arch/powerpc/platforms/pseries/vio.c) records the
return value and passes it on. [1]
During the device removal precedure, checking for resetting
bit is dropped so that we can continue executing all the
cleanup calls in the rest of the remove function. Otherwise,
it can cause latent memory leaks and kernel crashes.
[1] https://lore.kernel.org/linuxppc-dev/20210117101242.dpwayq6wdgfdzirl@pengutronix.de/T/#m48f5befd96bc9842ece2a3ad14f4c27747206a53 Reported-by: Uwe Kleine-Kรถnig <u.kleine-koenig@pengutronix.de> Fixes: 76c528960bab ("ibmvnic: Do not process device remove during device reset") Signed-off-by: Lijun Pan <ljp@linux.ibm.com> Link: https://lore.kernel.org/r/20210129043402.95744-1-ljp@linux.ibm.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Florian, Andrew and Vladimir suggested at some point to use devlink for
reporting tables, features and debugging counters instead of using debugfs and
printk.
So, start by reporting the VLAN and FDB table sizes.
====================
DENG Qingfang [Sat, 30 Jan 2021 13:43:34 +0000 (21:43 +0800)]
net: dsa: mv88e6xxx: override existent unicast portvec in port_fdb_add
Having multiple destination ports for a unicast address does not make
sense.
Make port_db_load_purge override existent unicast portvec instead of
adding a new port bit.
Fixes: a943b2a0b2ac ("net: dsa: mv88e6xxx: handle multiple ports in ATU") Signed-off-by: DENG Qingfang <dqfext@gmail.com> Reviewed-by: Vladimir Oltean <olteanv@gmail.com> Link: https://lore.kernel.org/r/20210130134334.10243-1-dqfext@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
VF queues were not brought up when PF was brought up after being
downed if the VF driver disabled VFs queues during PF down.
This could happen in some older or external VF driver implementations.
The problem was that PF driver used vf->queues_enabled as a condition
to decide what link-state it would send out which caused the issue.
Remove the check for vf->queues_enabled in the VF link notify.
Now VF will always be notified of the current link status.
Also remove the queues_enabled member from i40e_vf structure as it is
not used anymore. Otherwise VNF implementation was broken and caused
a link flap.
The original commit was a workaround to avoid breaking existing VFs though
it's really a fault of the VF code not the PF. The commit should be safe to
revert as all of the VFs we know of have been fixed. Also, since we now
know there is a related bug in the workaround, removing it is preferred.
Fixes: 346f948a5385 ("i40e: don't report link up for a VF who hasn't enabled") Signed-off-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com> Signed-off-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com> Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Linus Torvalds [Mon, 1 Feb 2021 19:15:57 +0000 (11:15 -0800)]
Merge tag 'media/v5.11-3' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media
Pull media fixes from Mauro Carvalho Chehab:
"The rockship rkisp1 driver will be promoted from staging in 5.11.
While not too late, do a few uAPI changes which are needed to better
support its functionalities"
* tag 'media/v5.11-3' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
media: rockchip: rkisp1: extend uapi array sizes
media: rockchip: rkisp1: carry ip version information
media: rockchip: rkisp1: reduce number of histogram grid elements in uapi
media: rkisp1: stats: mask the hist_bins values
media: rkisp1: stats: remove a wrong cast to u8
media: rkisp1: uapi: change hist_bins array type from __u16 to __u32
Hans de Goede [Mon, 1 Feb 2021 15:29:56 +0000 (16:29 +0100)]
staging: rtl8723bs: Move wiphy setup to after reading the regulatory settings from the chip
Commit 73c6fe37ed09 ("staging: rtl8723bs: fix wireless regulatory API
misuse") moved the wiphy_apply_custom_regulatory() call to earlier in the
driver's init-sequence, so that it gets called before wiphy_register().
But at this point in time the eFuses which code the regulatory-settings
for the chip have not been read by the driver yet, causing
_rtw_reg_apply_flags() to set the IEEE80211_CHAN_DISABLED flag on *all*
channels.
On the device where I initially tested the fix, a Jumper EZpad 7 tablet,
this does not cause any problems because shortly after init the
rtw_reg_notifier() gets called fixing things up. I guess this happens
into response to receiving a (broadcast) packet with regulatory info
from the access-point ?
But on another device with a RTL8723BS wifi chip, an Acer Switch 10E
(SW3-016), the rtw_reg_notifier() never gets called. I assume that some
fuse has been set on this device to ignore regulatory info received from
access-points.
This means that on the Acer the driver is stuck in a state with all
channels disabled, leading to non working Wifi.
We cannot move the wiphy_apply_custom_regulatory() call back, because
that call must be made before the wiphy_register() call.
Instead move the entire rtw_wdev_alloc() call to after the Efuses have
been read, fixing all channels being disabled in the initial channel-map.
Fixes: 73c6fe37ed09 ("staging: rtl8723bs: fix wireless regulatory API misuse") Cc: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Link: https://lore.kernel.org/r/20210201152956.370186-2-hdegoede@redhat.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Kevin Lo [Thu, 7 Jan 2021 06:10:38 +0000 (14:10 +0800)]
igc: check return value of ret_val in igc_config_fc_after_link_up
Check return value from ret_val to make error check actually work.
Fixes: 3a56c17c0297 ("igc: Add setup link functionality") Signed-off-by: Kevin Lo <kevlo@kevlo.org> Acked-by: Sasha Neftin <sasha.neftin@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Kevin Lo [Sun, 20 Dec 2020 14:18:19 +0000 (22:18 +0800)]
igc: set the default return value to -IGC_ERR_NVM in igc_write_nvm_srwr
This patch sets the default return value to -IGC_ERR_NVM in
igc_write_nvm_srwr. Without this change it wouldn't lead to a shadow RAM
write EEWR timeout.
Fixes: d6c58def7f21 ("igc: Add NVM support") Signed-off-by: Kevin Lo <kevlo@kevlo.org> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Kai-Heng Feng [Wed, 2 Dec 2020 07:50:17 +0000 (15:50 +0800)]
igc: Report speed and duplex as unknown when device is runtime suspended
Similar to commit 778be3147ac3 ("igb: Report speed and duplex as unknown
when device is runtime suspended"), if we try to read speed and duplex
sysfs while the device is runtime suspended, igc will complain and
stops working:
The more generic approach will be wrap get_link_ksettings() with begin()
and complete() callbacks, and calls runtime resume and runtime suspend
routine respectively. However, igc is like igb, runtime resume routine
uses rtnl_lock() which upper ethtool layer also uses.
So to prevent a deadlock on rtnl, take a different approach, use
pm_runtime_suspended() to avoid reading register while device is runtime
suspended.
Felix Fietkau [Mon, 1 Feb 2021 08:33:24 +0000 (09:33 +0100)]
mac80211: fix station rate table updates on assoc
If the driver uses .sta_add, station entries are only uploaded after the sta
is in assoc state. Fix early station rate table updates by deferring them
until the sta has been uploaded.
Cc: stable@vger.kernel.org Signed-off-by: Felix Fietkau <nbd@nbd.name> Link: https://lore.kernel.org/r/20210201083324.3134-1-nbd@nbd.name
[use rcu_access_pointer() instead since we won't dereference here] Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Linus Torvalds [Sun, 31 Jan 2021 19:48:12 +0000 (11:48 -0800)]
Merge tag 'x86_entry_for_v5.11_rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fix from Borislav Petkov:
"A single fix for objtool to generate proper unwind info for newer
toolchains which do not generate section symbols anymore. And a
cleanup ontop.
This was originally going to go during the next merge window but
people can already trigger a build error with binutils-2.36 which
doesn't emit section symbols - something which objtool relies on - so
let's expedite it"
* tag 'x86_entry_for_v5.11_rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/entry: Remove put_ret_addr_in_rdi THUNK macro argument
x86/entry: Emit a symbol for register restoring thunk
Linus Torvalds [Sun, 31 Jan 2021 19:40:57 +0000 (11:40 -0800)]
Merge tag 'timers-urgent-2021-01-31' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer fix from Thomas Gleixner:
"A fix for handling advertised, but non-existent 146818 RTCs correctly.
With the recent UIP handling changes the time readout of non-existent
RTCs hangs forever as the read returns always 0xFF which means the UIP
bit is set.
Sanity check the RTC before registering by checking the RTC_VALID
register for correctness"
* tag 'timers-urgent-2021-01-31' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
rtc: mc146818: Detect and handle broken RTCs
Linus Torvalds [Sun, 31 Jan 2021 19:39:32 +0000 (11:39 -0800)]
Merge tag 'core-urgent-2021-01-31' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull single stepping fix from Thomas Gleixner:
"A single fix for the single step reporting regression caused by
getting the condition wrong when moving SYSCALL_EMU away from TIF
flags"
[ There's apparently another problem too, fix pending ]
* tag 'core-urgent-2021-01-31' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
entry: Unbreak single step reporting behaviour
Linus Torvalds [Sun, 31 Jan 2021 19:37:43 +0000 (11:37 -0800)]
Merge tag 'powerpc-5.11-6' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux
Pull powerpc fix from Michael Ellerman:
"One fix for a bug in our soft interrupt masking, which could lead to
interrupt replaying recursing, causing spurious interrupts.
Thanks to Nicholas Piggin"
* tag 'powerpc-5.11-6' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/64s: prevent recursive replay_soft_interrupts causing superfluous interrupt
Linus Torvalds [Sun, 31 Jan 2021 19:19:12 +0000 (11:19 -0800)]
Merge tag 'nfs-for-5.11-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs
Pull NFS client fixes from Trond Myklebust:
- SUNRPC: Handle 0 length opaque XDR object data properly
- Fix a layout segment leak in pnfs_layout_process()
- pNFS/NFSv4: Update the layout barrier when we schedule a layoutreturn
- pNFS/NFSv4: Improve rejection of out-of-order layouts
- pNFS/NFSv4: Try to return invalid layout in pnfs_layout_process()
* tag 'nfs-for-5.11-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
SUNRPC: Handle 0 length opaque XDR object data properly
SUNRPC: Move simple_get_bytes and simple_get_netobj into private header
pNFS/NFSv4: Improve rejection of out-of-order layouts
pNFS/NFSv4: Update the layout barrier when we schedule a layoutreturn
pNFS/NFSv4: Try to return invalid layout in pnfs_layout_process()
pNFS/NFSv4: Fix a layout segment leak in pnfs_layout_process()
Linus Walleij [Sat, 2 Jan 2021 23:15:10 +0000 (00:15 +0100)]
leds: rt8515: Add Richtek RT8515 LED driver
This adds a driver for the Richtek RT8515 dual channel
torch/flash white LED driver.
This LED driver is found in some mobile phones from
Samsung such as the GT-S7710 and GT-I8190.
A V4L interface is added.
We do not have a proper datasheet for the RT8515 but
it turns out that RT9387A has a public datasheet and
is essentially the same chip. We designed the driver
in accordance with this datasheet. The day someone
needs to drive a RT9387A this driver can probably
easily be augmented to handle that chip too.
Sakari Ailus, Pavel Machek and Andy Shevchenko helped
significantly in getting this driver right.
Cc: Sakari Ailus <sakari.ailus@iki.fi> Cc: newbytee@protonmail.com Cc: Stephan Gerhold <stephan@gerhold.net> Cc: linux-media@vger.kernel.org Cc: phone-devel@vger.kernel.org Reviewed-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Pavel Machek <pavel@ucw.cz>
Andrea Righi [Wed, 25 Nov 2020 15:18:22 +0000 (16:18 +0100)]
leds: trigger: fix potential deadlock with libata
We have the following potential deadlock condition:
========================================================
WARNING: possible irq lock inversion dependency detected
5.10.0-rc2+ #25 Not tainted
--------------------------------------------------------
swapper/3/0 just changed the state of lock: ffff8880063bd618 (&host->lock){-...}-{2:2}, at: ata_bmdma_interrupt+0x27/0x200
but this lock took another, HARDIRQ-READ-unsafe lock in the past:
(&trig->leddev_list_lock){.+.?}-{2:2}
and interrupts could create inverse lock ordering between them.
other info that might help us debug this:
Possible interrupt unsafe locking scenario:
This lockdep splat is reported after:
commit 6836e0c29878 ("locking: More accurate annotations for read_lock()")
To clarify:
- read-locks are recursive only in interrupt context (when
in_interrupt() returns true)
- after acquiring host->lock in CPU1, another cpu (i.e. CPU2) may call
write_lock(&trig->leddev_list_lock) that would be blocked by CPU0
that holds trig->leddev_list_lock in read-mode
- when CPU1 (ata_ac_complete()) tries to read-lock
trig->leddev_list_lock, it would be blocked by the write-lock waiter
on CPU2 (because we are not in interrupt context, so the read-lock is
not recursive)
- at this point if an interrupt happens on CPU0 and
ata_bmdma_interrupt() is executed it will try to acquire host->lock,
that is held by CPU1, that is currently blocked by CPU2, so:
* CPU0 blocked by CPU1
* CPU1 blocked by CPU2
* CPU2 blocked by CPU0
*** DEADLOCK ***
The deadlock scenario is better represented by the following schema
(thanks to Boqun Feng <boqun.feng@gmail.com> for the schema and the
detailed explanation of the deadlock condition):
CPU 0: CPU 1: CPU 2:
----- ----- -----
led_trigger_event():
read_lock(&trig->leddev_list_lock);
<workqueue>
ata_hsm_qc_complete():
spin_lock_irqsave(&host->lock);
write_lock(&trig->leddev_list_lock);
ata_port_freeze():
ata_do_link_abort():
ata_qc_complete():
ledtrig_disk_activity():
led_trigger_blink_oneshot():
read_lock(&trig->leddev_list_lock);
// ^ not in in_interrupt() context, so could get blocked by CPU 2
<interrupt>
ata_bmdma_interrupt():
spin_lock_irqsave(&host->lock);
Fix by using read_lock_irqsave/irqrestore() in led_trigger_event(), so
that no interrupt can happen in between, preventing the deadlock
condition.
Apply the same change to led_trigger_blink_setup() as well, since the
same deadlock scenario can also happen in power_supply_update_bat_leds()
-> led_trigger_blink() -> led_trigger_blink_setup() (workqueue context),
and potentially prevent other similar usages.
Linus Torvalds [Sun, 31 Jan 2021 01:51:06 +0000 (17:51 -0800)]
Merge tag '5.11-rc5-smb3' of git://git.samba.org/sfrench/cifs-2.6
Pull cifs fixes from Steve French:
"Four cifs patches found in additional testing of the conversion to the
new mount API: three small option processing ones, and one fixing domain
based DFS referrals"
* tag '5.11-rc5-smb3' of git://git.samba.org/sfrench/cifs-2.6:
cifs: fix dfs domain referrals
cifs: returning mount parm processing errors correctly
cifs: fix mounts to subdirectories of target
cifs: ignore auto and noauto options if given
Linus Torvalds [Sun, 31 Jan 2021 01:42:42 +0000 (17:42 -0800)]
Merge tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Pull SCSI fixes from James Bottomley:
"Two minor fixes in drivers. Both changing strings (one in a comment,
one in a module help text) with no code impact"
* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
scsi: qla2xxx: Fix description for parameter ql2xenforce_iocb_limit
scsi: target: iscsi: Fix typo in comment
Linus Torvalds [Sat, 30 Jan 2021 19:48:57 +0000 (11:48 -0800)]
Merge tag 's390-5.11-4' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux
Pull s390 fixes from Vasily Gorbik:
- Fix max number of VCPUs reported via ultravisor information sysfs
interface.
- Fix memory leaks during vfio-ap resources clean up on KVM pointer
invalidation notification.
- Fix potential specification exception by avoiding unnecessary
interrupts disable after queue reset in vfio-ap.
* tag 's390-5.11-4' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
s390: uv: Fix sysfs max number of VCPUs reporting
s390/vfio-ap: No need to disable IRQ after queue reset
s390/vfio-ap: clean up vfio_ap resources when KVM pointer invalidated
Chinmay Agarwal [Wed, 27 Jan 2021 16:54:54 +0000 (22:24 +0530)]
neighbour: Prevent a dead entry from updating gc_list
Following race condition was detected:
<CPU A, t0> - neigh_flush_dev() is under execution and calls
neigh_mark_dead(n) marking the neighbour entry 'n' as dead.
<CPU B, t1> - Executing: __netif_receive_skb() ->
__netif_receive_skb_core() -> arp_rcv() -> arp_process().arp_process()
calls __neigh_lookup() which takes a reference on neighbour entry 'n'.
<CPU A, t2> - Moves further along neigh_flush_dev() and calls
neigh_cleanup_and_release(n), but since reference count increased in t2,
'n' couldn't be destroyed.
<CPU B, t3> - Moves further along, arp_process() and calls
neigh_update()-> __neigh_update() -> neigh_update_gc_list(), which adds
the neighbour entry back in gc_list(neigh_mark_dead(), removed it
earlier in t0 from gc_list)
This leads to 'n' still being part of gc_list, but the actual
neighbour structure has been freed.
The situation can be prevented from happening if we disallow a dead
entry to have any possibility of updating gc_list. This is what the
patch intends to achieve.
Fixes: 6547d8e2146f ("neighbor: Fix locking order for gc_list changes") Signed-off-by: Chinmay Agarwal <chinagar@codeaurora.org> Reviewed-by: Cong Wang <xiyou.wangcong@gmail.com> Reviewed-by: David Ahern <dsahern@kernel.org> Link: https://lore.kernel.org/r/20210127165453.GA20514@chinagar-linux.qualcomm.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Neal Cardwell [Fri, 29 Jan 2021 18:54:38 +0000 (13:54 -0500)]
tcp: shrink inet_connection_sock icsk_mtup enabled and probe_size
This commit shrinks inet_connection_sock by 4 bytes, by shrinking
icsk_mtup.enabled from 32 bits to 1 bit, and shrinking
icsk_mtup.probe_size from s32 to an unsuigned 31 bit field.
This is to save space to compensate for the recent introduction of a
new u32 in inet_connection_sock, icsk_probes_tstamp, in the recent bug
fix commit 8a248b911650 ("tcp: fix TCP_USER_TIMEOUT with zero window").
This should not change functionality, since icsk_mtup.enabled is only
ever set to 0 or 1, and icsk_mtup.probe_size can only be either 0
or a positive MTU value returned by tcp_mss_to_mtu()