| Age | Commit message (Collapse) | Author |
|---|
|
For vhost-scsi with 3 vqs or more and a workload that tries to use
them in parallel like:
fio --filename=/dev/sdb --direct=1 --rw=randrw --bs=4k \
--ioengine=libaio --iodepth=128 --numjobs=3
the single vhost worker thread will become a bottlneck and we are stuck
at around 500K IOPs no matter how many jobs, virtqueues, and CPUs are
used.
To better utilize virtqueues and available CPUs, this patch allows
userspace to create workers and bind them to vqs. You can have N workers
per dev and also share N workers with M vqs on that dev.
This patch adds the interface related code and the next patch will hook
vhost-scsi into it. The patches do not try to hook net and vsock into
the interface because:
1. multiple workers don't seem to help vsock. The problem is that with
only 2 virtqueues we never fully use the existing worker when doing
bidirectional tests. This seems to match vhost-scsi where we don't see
the worker as a bottleneck until 3 virtqueues are used.
2. net already has a way to use multiple workers.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-16-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
The next patch allows userspace to create multiple workers per device,
so this patch replaces the vhost_worker pointer with an xarray so we
can store mupltiple workers and look them up.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-15-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
The next patches add new vhost worker ioctls which will need to get a
vhost_virtqueue from a userspace struct which specifies the vq's index.
This moves the vhost_vring_ioctl code to do this to a helper so it can
be shared.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-14-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
vhost_work_queue is no longer used. Each driver is using the poll or vq
based queueing, so remove vhost_work_queue.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-13-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
With one worker we will always send the scsi cmd responses then send the
TMF rsp, because LIO will always complete the scsi cmds first then call
into us to send the TMF response.
With multiple workers, the IO vq workers could be running while the
TMF/ctl vq worker is running so this has us do a flush before completing
the TMF to make sure cmds are completed when it's work is later queued
and run.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-12-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
Convert from vhost_work_queue to vhost_vq_work_queue so we can
remove vhost_work_queue.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-11-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
This patch separates the scsi cmd completion code paths so we can complete
cmds based on their vq instead of having all cmds complete on the same
worker/CPU. This will be useful with the next patches that allow us to
create mulitple worker threads and bind them to different vqs, and we can
have completions running on different threads/CPUs.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20230626232307.97930-10-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
Convert from vhost_work_queue to vhost_vq_work_queue, so we can drop
vhost_work_queue.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-9-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
This has the drivers pass in their poll to vq mapping and then converts
the core poll code to use the vq based helpers. In the next patches we
will allow vqs to be handled by different workers, so to allow drivers
to execute operations like queue, stop, flush, etc on specific polls/vqs
we need to know the mappings.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-8-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
This patch has the core work flush function take a worker. When we
support multiple workers we can then flush each worker during device
removal, stoppage, etc. It also adds a helper to flush specific
virtqueues, so vhost-scsi can flush IO vqs from it's ctl vq.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-7-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
This patch has the core work queueing function take a worker for when we
support multiple workers. It also adds a helper that takes a vq during
queueing so modules can control which vq/worker to queue work on.
This temp leaves vhost_work_queue. It will be removed when the drivers
are converted in the next patches.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-6-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
In the next patches each vq might have different workers so one could
have work but others do not. For net, we only want to check specific vqs,
so this adds a helper to check if a vq has work pending and converts
vhost-net to use it.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20230626232307.97930-5-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
This patchset allows userspace to map vqs to different workers. This
patch adds a worker pointer to the vq so in later patches in this set
we can queue/flush specific vqs and their workers.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-4-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
This patchset allows us to allocate multiple workers, so this has us
move from the vhost_worker that's embedded in the vhost_dev to
dynamically allocating it.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-3-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
vsock can start queueing work after VHOST_VSOCK_SET_GUEST_CID, so
after we have called vhost_worker_create it can be calling
vhost_work_queue and trying to access the vhost worker/task. If
vhost_dev_alloc_iovecs fails, then vhost_worker_free could free
the worker/task from under vsock.
This moves vhost_worker_create to the end of vhost_dev_set_owner
where we know we can no longer fail in that path. If it fails
after the VHOST_SET_OWNER and userspace closes the device, then
the normal vsock release handling will do the right thing.
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Message-Id: <20230626232307.97930-2-michael.christie@oracle.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
For virtio-net we were getting CPU stall warnings, and fixed it by
calling the scheduler: see f8bb51043945 ("virtio_net: suppress cpu stall
when free_unused_bufs").
This driver is similar so theoretically the same logic applies.
Signed-off-by: Xianting Tian <xianting.tian@linux.alibaba.com>
Message-Id: <20230609131817.712867-4-xianting.tian@linux.alibaba.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
For virtio-net we were getting CPU stall warnings, and fixed it by
calling the scheduler: see f8bb51043945 ("virtio_net: suppress cpu stall
when free_unused_bufs").
This driver is similar so theoretically the same logic applies.
Signed-off-by: Xianting Tian <xianting.tian@linux.alibaba.com>
Message-Id: <20230609131817.712867-3-xianting.tian@linux.alibaba.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
For virtio-net we were getting CPU stall warnings, and fixed it by
calling the scheduler: see f8bb51043945 ("virtio_net: suppress cpu stall
when free_unused_bufs").
This driver is similar so theoretically the same logic applies.
Signed-off-by: Xianting Tian <xianting.tian@linux.alibaba.com>
Message-Id: <20230609131817.712867-2-xianting.tian@linux.alibaba.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
This commit implements a better layout of the
live migration bar, therefore the accessors for virtqueue
state have been refactored.
This commit also add a comment to the probing-ids list,
indicating this driver drives F2000X-PL virtio-net
Signed-off-by: Zhu Lingshan <lingshan.zhu@intel.com>
Message-Id: <20230612151420.1019504-4-lingshan.zhu@intel.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
Rather than a hardcode, this commit detects
and reports the max value of allowed size
of the virtqueues
Signed-off-by: Zhu Lingshan <lingshan.zhu@intel.com>
Message-Id: <20230612151420.1019504-3-lingshan.zhu@intel.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
This commit dynamically allocates the data
stores for the virtqueues based on
virtio_pci_common_cfg.num_queues.
Signed-off-by: Zhu Lingshan <lingshan.zhu@intel.com>
Message-Id: <20230612151420.1019504-2-lingshan.zhu@intel.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
While initially I thought that we couldn't move all new mount api
handling into params.{c,h} it turns out it is possible. So this just
moves a good chunk of code out of super.c and into params.{c,h}.
Signed-off-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
|
|
kdb_send_sig() is defined in the signal code and called from kdb,
but the declaration is part of the kdb internal code.
Move the declaration to the shared header to avoid the warning:
kernel/signal.c:4789:6: error: no previous prototype for 'kdb_send_sig' [-Werror=missing-prototypes]
Reported-by: Arnd Bergmann <arnd@arndb.de>
Closes: https://lore.kernel.org/lkml/20230517125423.930967-1-arnd@kernel.org/
Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org>
Link: https://lore.kernel.org/r/20230630201206.2396930-1-daniel.thompson@linaro.org
|
|
Switch to the quicinc.com id.
Fixes: bd1af6b5fffd ("Documentation: ABI: sysfs-class-net-qmi: document pass-through file")
Signed-off-by: Subash Abhinov Kasiviswanathan <quic_subashab@quicinc.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
request sockets are lockless, __tcp_oow_rate_limited() could be called
on the same object from different cpus. This is harmless.
Add READ_ONCE()/WRITE_ONCE() annotations to avoid a KCSAN report.
Fixes: 4ce7e93cb3fe ("tcp: rate limit ACK sent by SYN_RECV request sockets")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Jason A. Donenfeld says:
====================
wireguard fixes for 6.4.2/6.5-rc1
Sorry to send these patches during the merge window, but they're net
fixes, not netdev enhancements, and while I'd ordinarily wait anyway,
I just got a first bug report for one of these fixes, which I originally
had thought was mostly unlikely. So please apply the following three
patches to net:
1) Make proper use of nr_cpu_ids with cpumask_next(), rather than
awkwardly using modulo, to handle dynamic CPU topology changes.
Linus noticed this a while ago and pointed it out, and today a user
actually got hit by it.
2) Respect persistent keepalive and other staged packets when setting
the private key after the interface is already up.
3) Use timer_delete_sync() instead of del_timer_sync(), per the
documentation.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The documentation says that del_timer_sync is obsolete, and code should
use the equivalent timer_delete_sync instead, so switch to it.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Packets bound for peers can queue up prior to the device private key
being set. For example, if persistent keepalive is set, a packet is
queued up to be sent as soon as the device comes up. However, if the
private key hasn't been set yet, the handshake message never sends, and
no timer is armed to retry, since that would be pointless.
But, if a user later sets a private key, the expectation is that those
queued packets, such as a persistent keepalive, are actually sent. So
adjust the configuration logic to account for this edge case, and add a
test case to make sure this works.
Maxim noticed this with a wg-quick(8) config to the tune of:
[Interface]
PostUp = wg set %i private-key somefile
[Peer]
PublicKey = ...
Endpoint = ...
PersistentKeepalive = 25
Here, the private key gets set after the device comes up using a PostUp
script, triggering the bug.
Fixes: e7096c131e51 ("net: WireGuard secure network tunnel")
Cc: stable@vger.kernel.org
Reported-by: Maxim Cournoyer <maxim.cournoyer@gmail.com>
Tested-by: Maxim Cournoyer <maxim.cournoyer@gmail.com>
Link: https://lore.kernel.org/wireguard/87fs7xtqrv.fsf@gmail.com/
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Using `% nr_cpumask_bits` is slow and complicated, and not totally
robust toward dynamic changes to CPU topologies. Rather than storing the
next CPU in the round-robin, just store the last one, and also return
that value. This simplifies the loop drastically into a much more common
pattern.
Fixes: e7096c131e51 ("net: WireGuard secure network tunnel")
Cc: stable@vger.kernel.org
Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Tested-by: Manuel Leiner <manuel.leiner@gmx.de>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Each sample script sources functions.sh before parameters.sh
which makes $APPEND undefined when trapping EXIT no matter in
append mode or not. Due to this when sample scripts finished
they always do "pgctrl reset" which resets pktgen config.
So move trap to each script after sourcing parameters.sh
and trap EXIT explicitly.
Signed-off-by: J.J. Martzki <mars14850@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When running Kselftests with the current selftests/net/config
the following problem can be seen with the net:xfrm_policy.sh
selftest:
# selftests: net: xfrm_policy.sh
[ 41.076721] IPv6: ADDRCONF(NETDEV_CHANGE): veth0: link becomes ready
[ 41.094787] IPv6: ADDRCONF(NETDEV_CHANGE): veth0: link becomes ready
[ 41.107635] IPv6: ADDRCONF(NETDEV_CHANGE): veth0: link becomes ready
# modprobe: FATAL: Module ip_tables not found in directory /lib/modules/6.1.36
# iptables v1.8.7 (legacy): can't initialize iptables table `filter': Table does not exist (do you need to insmod?)
# Perhaps iptables or your kernel needs to be upgraded.
# modprobe: FATAL: Module ip_tables not found in directory /lib/modules/6.1.36
# iptables v1.8.7 (legacy): can't initialize iptables table `filter': Table does not exist (do you need to insmod?)
# Perhaps iptables or your kernel needs to be upgraded.
# SKIP: Could not insert iptables rule
ok 1 selftests: net: xfrm_policy.sh # SKIP
This is because IPsec "policy" match support is not available
to the kernel.
This patch adds CONFIG_NETFILTER_XT_MATCH_POLICY as a module
to the selftests/net/config file, so that `make
kselftest-merge` can take this into consideration.
Signed-off-by: Daniel Díaz <daniel.diaz@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
After blamed commit, we must be more careful about using
skb_transport_offset(), as reminded us by syzbot:
WARNING: CPU: 0 PID: 10 at include/linux/skbuff.h:2868 skb_transport_offset include/linux/skbuff.h:2977 [inline]
WARNING: CPU: 0 PID: 10 at include/linux/skbuff.h:2868 perf_trace_net_dev_start_xmit+0x89a/0xce0 include/trace/events/net.h:14
Modules linked in:
CPU: 0 PID: 10 Comm: kworker/u4:1 Not tainted 6.1.30-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 05/27/2023
Workqueue: bat_events batadv_iv_send_outstanding_bat_ogm_packet
RIP: 0010:skb_transport_header include/linux/skbuff.h:2868 [inline]
RIP: 0010:skb_transport_offset include/linux/skbuff.h:2977 [inline]
RIP: 0010:perf_trace_net_dev_start_xmit+0x89a/0xce0 include/trace/events/net.h:14
Code: 8b 04 25 28 00 00 00 48 3b 84 24 c0 00 00 00 0f 85 4e 04 00 00 48 8d 65 d8 5b 41 5c 41 5d 41 5e 41 5f 5d c3 cc e8 56 22 01 fd <0f> 0b e9 f6 fc ff ff 89 f9 80 e1 07 80 c1 03 38 c1 0f 8c 86 f9 ff
RSP: 0018:ffffc900002bf700 EFLAGS: 00010293
RAX: ffffffff8485d8ca RBX: 000000000000ffff RCX: ffff888100914280
RDX: 0000000000000000 RSI: 000000000000ffff RDI: 000000000000ffff
RBP: ffffc900002bf818 R08: ffffffff8485d5b6 R09: fffffbfff0f8fb5e
R10: 0000000000000000 R11: dffffc0000000001 R12: 1ffff110217d8f67
R13: ffff88810bec7b3a R14: dffffc0000000000 R15: dffffc0000000000
FS: 0000000000000000(0000) GS:ffff8881f6a00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f96cf6d52f0 CR3: 000000012224c000 CR4: 0000000000350ef0
Call Trace:
<TASK>
[<ffffffff84715e35>] trace_net_dev_start_xmit include/trace/events/net.h:14 [inline]
[<ffffffff84715e35>] xmit_one net/core/dev.c:3643 [inline]
[<ffffffff84715e35>] dev_hard_start_xmit+0x705/0x980 net/core/dev.c:3660
[<ffffffff8471a232>] __dev_queue_xmit+0x16b2/0x3370 net/core/dev.c:4324
[<ffffffff85416493>] dev_queue_xmit include/linux/netdevice.h:3030 [inline]
[<ffffffff85416493>] batadv_send_skb_packet+0x3f3/0x680 net/batman-adv/send.c:108
[<ffffffff85416744>] batadv_send_broadcast_skb+0x24/0x30 net/batman-adv/send.c:127
[<ffffffff853bc52a>] batadv_iv_ogm_send_to_if net/batman-adv/bat_iv_ogm.c:393 [inline]
[<ffffffff853bc52a>] batadv_iv_ogm_emit net/batman-adv/bat_iv_ogm.c:421 [inline]
[<ffffffff853bc52a>] batadv_iv_send_outstanding_bat_ogm_packet+0x69a/0x840 net/batman-adv/bat_iv_ogm.c:1701
[<ffffffff8151023c>] process_one_work+0x8ac/0x1170 kernel/workqueue.c:2289
[<ffffffff81511938>] worker_thread+0xaa8/0x12d0 kernel/workqueue.c:2436
Fixes: 66e4c8d95008 ("net: warn if transport header was not set")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
There was a regression introduced by the blamed commit, where pinging to
a VLAN-unaware bridge would fail with the repeated message "Couldn't
decode source port" coming from the tagging protocol driver.
When receiving packets with a bridge_vid as determined by
dsa_tag_8021q_bridge_join(), dsa_8021q_rcv() will decode:
- source_port = 0 (which isn't really valid, more like "don't know")
- switch_id = 0 (which isn't really valid, more like "don't know")
- vbid = value in range 1-7
Since the blamed patch has reversed the order of the checks, we are now
going to believe that source_port != -1 and switch_id != -1, so they're
valid, but they aren't.
The minimal solution to the problem is to only populate source_port and
switch_id with what dsa_8021q_rcv() came up with, if the vbid is zero,
i.e. the source port information is trustworthy.
Fixes: c1ae02d87689 ("net: dsa: tag_sja1105: always prefer source port information from INCL_SRCPT")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
According to the synchronization rules for .ndo_get_stats() as seen in
Documentation/networking/netdevices.rst, acquiring a plain spin_lock()
should not be illegal, but the bridge driver implementation makes it so.
After running these commands, I am being faced with the following
lockdep splat:
$ ip link add link swp0 name macsec0 type macsec encrypt on && ip link set swp0 up
$ ip link add dev br0 type bridge vlan_filtering 1 && ip link set br0 up
$ ip link set macsec0 master br0 && ip link set macsec0 up
========================================================
WARNING: possible irq lock inversion dependency detected
6.4.0-04295-g31b577b4bd4a #603 Not tainted
--------------------------------------------------------
swapper/1/0 just changed the state of lock:
ffff6bd348724cd8 (&br->lock){+.-.}-{3:3}, at: br_forward_delay_timer_expired+0x34/0x198
but this lock took another, SOFTIRQ-unsafe lock in the past:
(&ocelot->stats_lock){+.+.}-{3:3}
and interrupts could create inverse lock ordering between them.
other info that might help us debug this:
Chain exists of:
&br->lock --> &br->hash_lock --> &ocelot->stats_lock
Possible interrupt unsafe locking scenario:
CPU0 CPU1
---- ----
lock(&ocelot->stats_lock);
local_irq_disable();
lock(&br->lock);
lock(&br->hash_lock);
<Interrupt>
lock(&br->lock);
*** DEADLOCK ***
(details about the 3 locks skipped)
swp0 is instantiated by drivers/net/dsa/ocelot/felix.c, and this
only matters to the extent that its .ndo_get_stats64() method calls
spin_lock(&ocelot->stats_lock).
Documentation/locking/lockdep-design.rst says:
| A lock is irq-safe means it was ever used in an irq context, while a lock
| is irq-unsafe means it was ever acquired with irq enabled.
(...)
| Furthermore, the following usage based lock dependencies are not allowed
| between any two lock-classes::
|
| <hardirq-safe> -> <hardirq-unsafe>
| <softirq-safe> -> <softirq-unsafe>
Lockdep marks br->hash_lock as softirq-safe, because it is sometimes
taken in softirq context (for example br_fdb_update() which runs in
NET_RX softirq), and when it's not in softirq context it blocks softirqs
by using spin_lock_bh().
Lockdep marks ocelot->stats_lock as softirq-unsafe, because it never
blocks softirqs from running, and it is never taken from softirq
context. So it can always be interrupted by softirqs.
There is a call path through which a function that holds br->hash_lock:
fdb_add_hw_addr() will call a function that acquires ocelot->stats_lock:
ocelot_port_get_stats64(). This can be seen below:
ocelot_port_get_stats64+0x3c/0x1e0
felix_get_stats64+0x20/0x38
dsa_slave_get_stats64+0x3c/0x60
dev_get_stats+0x74/0x2c8
rtnl_fill_stats+0x4c/0x150
rtnl_fill_ifinfo+0x5cc/0x7b8
rtmsg_ifinfo_build_skb+0xe4/0x150
rtmsg_ifinfo+0x5c/0xb0
__dev_notify_flags+0x58/0x200
__dev_set_promiscuity+0xa0/0x1f8
dev_set_promiscuity+0x30/0x70
macsec_dev_change_rx_flags+0x68/0x88
__dev_set_promiscuity+0x1a8/0x1f8
__dev_set_rx_mode+0x74/0xa8
dev_uc_add+0x74/0xa0
fdb_add_hw_addr+0x68/0xd8
fdb_add_local+0xc4/0x110
br_fdb_add_local+0x54/0x88
br_add_if+0x338/0x4a0
br_add_slave+0x20/0x38
do_setlink+0x3a4/0xcb8
rtnl_newlink+0x758/0x9d0
rtnetlink_rcv_msg+0x2f0/0x550
netlink_rcv_skb+0x128/0x148
rtnetlink_rcv+0x24/0x38
the plain English explanation for it is:
The macsec0 bridge port is created without p->flags & BR_PROMISC,
because it is what br_manage_promisc() decides for a VLAN filtering
bridge with a single auto port.
As part of the br_add_if() procedure, br_fdb_add_local() is called for
the MAC address of the device, and this results in a call to
dev_uc_add() for macsec0 while the softirq-safe br->hash_lock is taken.
Because macsec0 does not have IFF_UNICAST_FLT, dev_uc_add() ends up
calling __dev_set_promiscuity() for macsec0, which is propagated by its
implementation, macsec_dev_change_rx_flags(), to the lower device: swp0.
This triggers the call path:
dev_set_promiscuity(swp0)
-> rtmsg_ifinfo()
-> dev_get_stats()
-> ocelot_port_get_stats64()
with a calling context that lockdep doesn't like (br->hash_lock held).
Normally we don't see this, because even though many drivers that can be
bridge ports don't support IFF_UNICAST_FLT, we need a driver that
(a) doesn't support IFF_UNICAST_FLT, *and*
(b) it forwards the IFF_PROMISC flag to another driver, and
(c) *that* driver implements ndo_get_stats64() using a softirq-unsafe
spinlock.
Condition (b) is necessary because the first __dev_set_rx_mode() calls
__dev_set_promiscuity() with "bool notify=false", and thus, the
rtmsg_ifinfo() code path won't be entered.
The same criteria also hold true for DSA switches which don't report
IFF_UNICAST_FLT. When the DSA master uses a spin_lock() in its
ndo_get_stats64() method, the same lockdep splat can be seen.
I think the deadlock possibility is real, even though I didn't reproduce
it, and I'm thinking of the following situation to support that claim:
fdb_add_hw_addr() runs on a CPU A, in a context with softirqs locally
disabled and br->hash_lock held, and may end up attempting to acquire
ocelot->stats_lock.
In parallel, ocelot->stats_lock is currently held by a thread B (say,
ocelot_check_stats_work()), which is interrupted while holding it by a
softirq which attempts to lock br->hash_lock.
Thread B cannot make progress because br->hash_lock is held by A. Whereas
thread A cannot make progress because ocelot->stats_lock is held by B.
When taking the issue at face value, the bridge can avoid that problem
by simply making the ports promiscuous from a code path with a saner
calling context (br->hash_lock not held). A bridge port without
IFF_UNICAST_FLT is going to become promiscuous as soon as we call
dev_uc_add() on it (which we do unconditionally), so why not be
preemptive and make it promiscuous right from the beginning, so as to
not be taken by surprise.
With this, we've broken the links between code that holds br->hash_lock
or br->lock and code that calls into the ndo_change_rx_flags() or
ndo_get_stats64() ops of the bridge port.
Fixes: 2796d0c648c9 ("bridge: Automatically manage port promiscuous mode.")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Pull iomap updates from Darrick Wong:
- Fix a type signature mismatch
- Drop Christoph as maintainer
* tag 'iomap-6.5-merge-1' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
iomap: drop me [hch] from MAINTAINERS for iomap
fs: iomap: Change the type of blocksize from 'int' to 'unsigned int' in iomap_file_buffered_write_punch_delalloc
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull vfs fix from Christian Brauner:
"A fix for the backing file work from this cycle.
When init_file() failed it would call file_free_rcu() on the file
allocated by the caller of init_file(). It naively assumed that the
correct cleanup operation would be called depending on whether it is a
regular file or a backing file. However, that presupposes that the
FMODE_BACKING flag would already be set which it won't be as that is
done in the caller of init_file().
Fix that bug by moving the cleanup of the allocated file into the
caller where it belongs in the first place. There's no good reason for
init_file() to consume resources it didn't allocate. This is a
mainline only fix and was reported by syzbot. The fix was validated by
syzbot against the provided reproducer"
* tag 'v6.5/vfs.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
fs: move cleanup from init_file() into its callers
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux
Pull i2c updates from Wolfram Sang:
- I2C has now a co-maintainer taking care of the host drivers. Welcome
Andi Shyti and have fun!
- platform remove callback converted to return void in drivers
- simplify drivers by using devm_clk_get_enabled()
- introduce i2c_get_match_data() to avoid more boilerplate code
(especially since the core stopped delivering an i2c_device_id)
- and the usual bunch of driver updates
* tag 'i2c-for-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux: (38 commits)
i2c: uniphier: Use devm_clk_get_enabled()
i2c: uniphier-f: Use devm_clk_get_enabled()
i2c: owl: Use devm_clk_get_enabled()
i2c: lpc2k: Use devm_clk_get_enabled()
i2c: hix5hd2: Use devm_clk_get_enabled()
i2c: sun6i-p2wi: Use devm_clk_get_enabled()
i2c: pasemi-platform: Use devm_clk_get_enabled()
i2c: mt7621: Use devm_clk_get_enabled()
i2c: xiic: Use devm_clk_get_enabled()
i2c: davinci: Use platform table macro over module_alias
i2c: ocores: use devm_ managed clks
i2c: nomadik: Use dev_err_probe() whenever possible
i2c: nomadik: Use devm_clk_get_enabled()
i2c: nomadik: Remove unnecessary goto label
usb: typec: ucsi: Mark dGPUs as DEVICE scope
i2c: wmt: Use devm_platform_get_and_ioremap_resource()
i2c: versatile: Use devm_platform_get_and_ioremap_resource()
i2c: hix5hd2: Add I2C_M_STOP flag support for i2c-hix5hd2 driver.
i2c: mpc: Use of_property_read_reg() to parse "reg"
i2c: imx-lpi2c: Don't open-code DIV_ROUND_UP
...
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux
Pull parisc updates from Helge Deller:
- Add missing cacheflush() syscall
- Fix STI console on 64-bit-only machines
- Move kernel debug options to Kconfig.debug
- Lots of warning fixes in arch/parisc/ and drivers/parisc/ when
compiled with W=1
- Enable some more graphics drivers in refreshed defconfigs
* tag 'parisc-for-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux: (29 commits)
parisc: Refresh defconfigs
parisc: irq: Add irq-related function declarations
parisc: Move init function declarations into header file
parisc: dino: Make dino_init() returning void
parisc: lba_pci: Mark two variables __maybe_unused
parisc: unaligned: Include header file to avoid missing prototype warnings
parisc: signal: Mark do_notify_resume() and sys_rt_sigreturn() asmlinkage
parisc: unwind: Mark start and stop variables __maybe_unused
parisc: init: Drop unused variable end_paddr
parisc: traps: Mark functions static
parisc: processor: Fix kdoc for init_cpu_profiler()
parisc: sys_parisc: parisc_personality() is called from asm code
parisc: ccio-dma: Fix kdoc and compiler warnings
parisc: pdc_stable: Fix kdoc and compiler warnings
parisc: pci-dma: Make pcxl_alloc_range() static
parisc: Mark image_size __maybe_unused in perf_write()
parisc: module: Mark symindex __maybe_unused
parisc: pdc_chassis: Fix kdoc warnings
parisc: firmware: Fix kdoc warnings
parisc: drivers: Fix kdoc warnings
...
|
|
The value of "end" should be "start + length - 1".
Signed-off-by: Shiyang Ruan <ruansy.fnst@fujitsu.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
|
|
Use struct initializers to ensure that the xfs_btree_irecs passed into
the query_range function are completely initialized. No functional
changes, just closing some sloppy hygiene.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Improve the validation of the fsmap offset fields in the query keys and
move the validation to the top of the function now that we have pushed
the low key adjustment code downwards.
Also fix some indenting issues that aren't worth a separate patch.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
The external log device fsmap backend doesn't have an rmapbt to query,
so it's wasteful to spend time initializing the rmap_irec objects.
Worse yet, the log could (someday) be longer than 2^32 fsblocks, so
using the rmap irec structure will result in integer overflows.
Fix this mess by computing the start address that we want from keys[0]
directly, and use the daddr-based record filtering algorithm that we
also use for rtbitmap queries.
Fixes: e89c041338ed ("xfs: implement the GETFSMAP ioctl")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
The rtbitmap fsmap backend doesn't query the rmapbt, so it's wasteful to
spend time initializing the rmap_irec objects. Worse yet, the logic to
query the rtbitmap is spread across three separate functions, which is
unnecessarily difficult to follow.
Compute the start rtextent that we want from keys[0] directly and
combine the functions to avoid passing parameters around everywhere, and
consolidate all the logic into a single function. At one point many
years ago I intended to use __xfs_getfsmap_rtdev as the launching point
for realtime rmapbt queries, but this hasn't been the case for a long
time.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
The realtime section ends at the last rt extent. If the user configures
the rt geometry with an extent size that is not an integer factor of the
number of rt blocks, it's possible for there to be rt blocks past the
end of the last rt extent. These tail blocks cannot ever be allocated
and will cause corruption reports if the last extent coincides with the
end of an rt bitmap block, so do not report consider them for the
GETFSMAP output.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
It's not correct to use the rmap irec structure to hold query key
information to query the rtbitmap because the realtime volume can be
longer than 2^32 fsblocks in length. Because the rt volume doesn't have
allocation groups, introduce a daddr-based record filtering algorithm
and compute the rtextent values using 64-bit variables. The same
problem exists in the external log device fsmap implementation, so use
the same solution to fix it too.
After this patch, all the code that touches info->low and info->high
under xfs_getfsmap_logdev and __xfs_getfsmap_rtdev are unnecessary.
Cleaning this up will be done in subsequent patches.
Fixes: 4c934c7dd60c ("xfs: report realtime space information via the rtbitmap")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
I noticed a bug in ranged GETFSMAP queries:
# xfs_io -c 'fsmap -vvvv' /opt
EXT: DEV BLOCK-RANGE OWNER FILE-OFFSET AG AG-OFFSET TOTAL
0: 8:80 [0..7]: static fs metadata 0 (0..7) 8
<snip>
9: 8:80 [192..223]: 137 0..31 0 (192..223) 32
# xfs_io -c 'fsmap -vvvv -d 208 208' /opt
#
That's not right -- we asked what block maps block 208, and we should've
received a mapping for inode 137 offset 16. Instead, we get nothing.
The root cause of this problem is a mis-interaction between the fsmap
code and how btree ranged queries work. xfs_btree_query_range returns
any btree record that overlaps with the query interval, even if the
record starts before or ends after the interval. Similarly, GETFSMAP is
supposed to return a recordset containing all records that overlap the
range queried.
However, it's possible that the recordset is larger than the buffer that
the caller provided to convey mappings to userspace. In /that/ case,
userspace is supposed to copy the last record returned to fmh_keys[0]
and call GETFSMAP again. In this case, we do not want to return
mappings that we have already supplied to the caller. The call to
xfs_btree_query_range is the same, but now we ignore any records that
start before fmh_keys[0].
Unfortunately, we didn't implement the filtering predicate correctly.
The predicate should only be called when we're calling back for more
records. Accomplish this by setting info->low.rm_blockcount to a
nonzero value and ensuring that it is cleared as necessary. As a
result, we no longer want to adjust dkeys[0] in the main setup function
because that's confusing.
This patch doesn't touch the logdev/rtbitmap backends because they have
bigger problems that will be addressed by subsequent patches.
Found via xfs/556 with parent pointers enabled.
Fixes: e89c041338ed ("xfs: implement the GETFSMAP ioctl")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
|
|
Hariprasad Kelam says:
====================
octeontx2-af: MAC block fixes for CN10KB
This patch set contains fixes for the issues encountered in testing
CN10KB MAC block RPM_USX.
Patch1: firmware to kernel communication is not working due to wrong
interrupt configuration. CSR addresses are corrected.
Patch2: NIX to RVU PF mapping errors encountered due to wrong firmware
config. Corrects this mapping error.
Patch3: Driver is trying to access non exist cgx/lmac which is resulting
in kernel panic. Address this issue by adding proper checks.
Patch4: MAC features are not getting reset on FLR. Fix the issue by
resetting the stale config.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
AF driver configures MAC features like internal loopback and PFC upon
receiving the request from PF and its VF netdev. But these
features are not getting reset in FLR. This patch fixes the issue by
resetting the same.
Fixes: 23999b30ae67 ("octeontx2-af: Enable or disable CGX internal loopback")
Fixes: 1121f6b02e7a ("octeontx2-af: Priority flow control configuration support")
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
with the addition of new MAC blocks like CN10K RPM and CN10KB
RPM_USX, LMACs are noncontiguous and CGX blocks are also
noncontiguous. But during RVU driver initialization, the driver
is assuming they are contiguous and trying to access
cgx or lmac with their id which is resulting in kernel panic.
This patch fixes the issue by adding proper checks.
[ 23.219150] pc : cgx_lmac_read+0x38/0x70
[ 23.219154] lr : rvu_program_channels+0x3f0/0x498
[ 23.223852] sp : ffff000100d6fc80
[ 23.227158] x29: ffff000100d6fc80 x28: ffff00010009f880 x27:
000000000000005a
[ 23.234288] x26: ffff000102586768 x25: 0000000000002500 x24:
fffffffffff0f000
Fixes: 91c6945ea1f9 ("octeontx2-af: cn10k: Add RPM MAC support")
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Firmware configures NIX block mapping for all MAC blocks.
The current implementation reads the configuration and
creates the mapping between RVU PF and NIX blocks. But
this configuration is only valid for silicons that support
multiple blocks. For all other silicons, all MAC blocks
map to NIX0.
This patch corrects the mapping by adding a check for the same.
Fixes: c5a73b632b90 ("octeontx2-af: Map NIX block from CGX connection")
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
