hsr_register_frame_out() compares new sequence_nr vs the old one
recorded in hsr_node::seq_out and if the new sequence_nr is higher then
it will be written to hsr_node::seq_out as the new value.
This operation isn't locked so it is possible that two frames with the
same sequence number arrive (via the two slave devices) and are fed to
hsr_register_frame_out() at the same time. Both will pass the check and
update the sequence counter later to the same value. As a result the
content of the same packet is fed into the stack twice.
This was noticed by running ping and observing DUP being reported from
time to time.
Instead of using the hsr_priv::seqnr_lock for the whole receive path (as
it is for sending in the master node) add an additional lock that is only
used for sequence number checks and updates.
Add a per-node lock that is used during sequence number reads and
updates.
Fixes: f421436a591d3 ("net/hsr: Add support for the High-availability Seamless Redundancy protocol (HSRv0)") Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Sending frames via the hsr (master) device requires a sequence number
which is tracked in hsr_priv::sequence_nr and protected by
hsr_priv::seqnr_lock. Each time a new frame is sent, it will obtain a
new id and then send it via the slave devices.
Each time a packet is sent (via hsr_forward_do()) the sequence number is
checked via hsr_register_frame_out() to ensure that a frame is not
handled twice. This make sense for the receiving side to ensure that the
frame is not injected into the stack twice after it has been received
from both slave ports.
There is no locking to cover the sending path which means the following
scenario is possible:
Both skbs (or their struct hsr_frame_info) received an unique id.
However since skb2 was sent before skb1, the higher sequence number was
recorded in hsr_register_frame_out() and the late arriving skb1 was
dropped and never sent.
This scenario has been observed in a three node HSR setup, with node1 +
node2 having ping and iperf running in parallel. From time to time ping
reported a missing packet. Based on tracing that missing ping packet did
not leave the system.
It might be possible (didn't check) to drop the sequence number check on
the sending side. But if the higher sequence number leaves on wire
before the lower does and the destination receives them in that order
and it will drop the packet with the lower sequence number and never
inject into the stack.
Therefore it seems the only way is to lock the whole path from obtaining
the sequence number and sending via dev_queue_xmit() and assuming the
packets leave on wire in the same order (and don't get reordered by the
NIC).
Cover the whole path for the master interface from obtaining the ID
until after it has been forwarded via hsr_forward_skb() to ensure the
skbs are sent to the NIC in the order of the assigned sequence numbers.
Fixes: f421436a591d3 ("net/hsr: Add support for the High-availability Seamless Redundancy protocol (HSRv0)") Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The hsr device is a software device. Its
net_device_ops::ndo_start_xmit() routine will process the packet and
then pass the resulting skb to dev_queue_xmit().
During processing, hsr acquires a lock with spin_lock_bh()
(hsr_add_node()) which needs to be promoted to the _irq() suffix in
order to avoid a potential deadlock.
Then there are the warnings in dev_queue_xmit() (due to
local_bh_disable() with disabled interrupts) left.
Instead trying to address those (there is qdisc and…) for netpoll sake,
just disable netpoll on hsr.
Disable netpoll on hsr and replace the _irqsave() locking with _bh().
Fixes: f421436a591d3 ("net/hsr: Add support for the High-availability Seamless Redundancy protocol (HSRv0)") Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Due to the hashed-MAC optimisation one problem become visible:
hsr_handle_sup_frame() walks over the list of available nodes and merges
two node entries into one if based on the information in the supervision
both MAC addresses belong to one node. The list-walk happens on a RCU
protected list and delete operation happens under a lock.
If the supervision arrives on both slave interfaces at the same time
then this delete operation can occur simultaneously on two CPUs. The
result is the first-CPU deletes the from the list and the second CPUs
BUGs while attempting to dereference a poisoned list-entry. This happens
more likely with the optimisation because a new node for the mac_B entry
is created once a packet has been received and removed (merged) once the
supervision frame has been received.
Avoid removing/ cleaning up a hsr_node twice by adding a `removed' field
which is set to true after the removal and checked before the removal.
Fixes: f266a683a4804 ("net/hsr: Better frame dispatch") Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
hsr_forward_skb() a skb and keeps information in an on-stack
hsr_frame_info. hsr_get_node() assigns hsr_frame_info::node_src which is
from a RCU list. This pointer is used later in hsr_forward_do().
I don't see a reason why this pointer can't vanish midway since there is
no guarantee that hsr_forward_skb() is invoked from an RCU read section.
Use rcu_read_lock() to protect hsr_frame_info::node_src from its
assignment until it is no longer used.
Fixes: f266a683a4804 ("net/hsr: Better frame dispatch") Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Currently div2 value is applied to the wrong bits. This is caused by a
bug in the code where the shift is done only for lpl, for anything
else the mask is not shifted to the correct bits.
Fix this by correctly shift if lpl is not supported.
Fixes: 4d7dc77babfe ("clk: qcom: Add support for Krait clocks") Signed-off-by: Christian Marangi <ansuelsmth@gmail.com> Signed-off-by: Bjorn Andersson <andersson@kernel.org> Link: https://lore.kernel.org/r/20221108215625.30186-1-ansuelsmth@gmail.com Signed-off-by: Sasha Levin <sashal@kernel.org>
The sc7180 lpass clock controller's pm_runtime usage wasn't broken
quite as spectacularly as the sc7280's pm_runtime usage, but it was
still broken. Putting some printouts in at boot showed me this (with
serial console enabled, which makes the prints slow and thus changes
timing):
[ 3.109951] DOUG: my_pm_clk_resume, usage=1
[ 3.114767] DOUG: my_pm_clk_resume, usage=1
[ 3.664443] DOUG: my_pm_clk_suspend, usage=0
[ 3.897566] DOUG: my_pm_clk_suspend, usage=0
[ 3.910137] DOUG: my_pm_clk_resume, usage=1
[ 3.923217] DOUG: my_pm_clk_resume, usage=0
[ 4.440116] DOUG: my_pm_clk_suspend, usage=-1
[ 4.444982] DOUG: my_pm_clk_suspend, usage=0
[ 14.170501] DOUG: my_pm_clk_resume, usage=1
[ 14.176245] DOUG: my_pm_clk_resume, usage=0
It seems to be doing roughly the right sequence of calls, but just
like with sc7280 this is more by luck than anything. Having a usage of
-1 is just not OK.
Let's fix this like we did with sc7280.
Signed-off-by: Douglas Anderson <dianders@chromium.org> Fixes: ce8c195e652f ("clk: qcom: lpasscc: Introduce pm autosuspend for SC7180") Reviewed-by: Stephen Boyd <swboyd@chromium.org> Signed-off-by: Bjorn Andersson <andersson@kernel.org> Link: https://lore.kernel.org/r/20221104064055.2.I49b25b9bda9430fc7ea21e5a708ca5a0aced2798@changeid Signed-off-by: Sasha Levin <sashal@kernel.org>
Make sure the nss is valid for nss_delta array. Return zero
if the index is invalid.
Coverity message:
Event overrun-call: Overrunning callee's array of size 4 by passing
argument "n_chains" (which evaluates to 15) in call to
"mt76_tx_power_nss_delta".
int delta = mt76_tx_power_nss_delta(n_chains);
Fixes: 07cda406308b ("mt76: fix rounding issues on converting per-chain and combined txpower") Signed-off-by: Deren Wu <deren.wu@mediatek.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Signed-off-by: Sasha Levin <sashal@kernel.org>
The radar detctor is used for dfs channel. So it will start after switching
to dfs channel and will stop after leaving. The TX will be blocked if radar
detctor isn't stopped in non-dfs channel.
This patch resets the dfs state to indicate the radar detector needs to be
stopped.
Signed-off-by: Xing Song <xing.song@mediatek.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Signed-off-by: Sasha Levin <sashal@kernel.org>
In regulatory_init_db(), when it's going to return a error, reg_pdev
should be unregistered. When load_builtin_regdb_keys() fails it doesn't
do it and makes cfg80211 can't be reload with report:
Commit f3186dd87669 ("spi: Optionally use GPIO descriptors for CS GPIOs")
has changed the user-space interface so that bogus SPI_CS_HIGH started
to appear in the mask returned by SPI_IOC_RD_MODE even for active-low CS
pins. Commit 138c9c32f090
("spi: spidev: Fix CS polarity if GPIO descriptors are used") fixed only
SPI_IOC_WR_MODE part of the problem. Let's fix SPI_IOC_RD_MODE
symmetrically.
Fixes: f3186dd87669 ("spi: Optionally use GPIO descriptors for CS GPIOs") Signed-off-by: Alexander Sverdlin <alexander.sverdlin@siemens.com> Link: https://lore.kernel.org/r/20221130162927.539512-1-alexander.sverdlin@siemens.com Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The "ignore_updelay" variable needs to be initialized to false.
Fixes: f8a65ab2f3ff ("bonding: fix link recovery in mode 2 when updelay is nonzero") Signed-off-by: Dan Carpenter <error27@gmail.com> Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com> Acked-by: Jay Vosburgh <jay.vosburgh@canonical.com> Link: https://lore.kernel.org/r/Y4SWJlh3ohJ6EPTL@kili Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
When redirecting, we use sk_msg_to_ingress() to get the BPF_F_INGRESS
flag from the msg->flags. If apply_bytes is used and it is larger than
the current data being processed, sk_psock_msg_verdict() will not be
called when sendmsg() is called again. At this time, the msg->flags is 0,
and we lost the BPF_F_INGRESS flag.
So we need to save the BPF_F_INGRESS flag in sk_psock and use it when
redirection.
Fixes: 8934ce2fd081 ("bpf: sockmap redirect ingress support") Signed-off-by: Pengcheng Yang <yangpc@wangsu.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Jakub Sitnicki <jakub@cloudflare.com> Link: https://lore.kernel.org/bpf/1669718441-2654-3-git-send-email-yangpc@wangsu.com Signed-off-by: Sasha Levin <sashal@kernel.org>
In tcp_bpf_send_verdict() redirection, the eval variable is assigned to
__SK_REDIRECT after the apply_bytes data is sent, if msg has more_data,
sock_put() will be called multiple times.
We should reset the eval variable to __SK_NONE every time more_data
starts.
This causes:
IPv4: Attempt to release TCP socket in state 1 00000000b4c925d7
------------[ cut here ]------------
refcount_t: addition on 0; use-after-free.
WARNING: CPU: 5 PID: 4482 at lib/refcount.c:25 refcount_warn_saturate+0x7d/0x110
Modules linked in:
CPU: 5 PID: 4482 Comm: sockhash_bypass Kdump: loaded Not tainted 6.0.0 #1
Hardware name: Red Hat KVM, BIOS 1.11.0-2.el7 04/01/2014
Call Trace:
<TASK>
__tcp_transmit_skb+0xa1b/0xb90
? __alloc_skb+0x8c/0x1a0
? __kmalloc_node_track_caller+0x184/0x320
tcp_write_xmit+0x22a/0x1110
__tcp_push_pending_frames+0x32/0xf0
do_tcp_sendpages+0x62d/0x640
tcp_bpf_push+0xae/0x2c0
tcp_bpf_sendmsg_redir+0x260/0x410
? preempt_count_add+0x70/0xa0
tcp_bpf_send_verdict+0x386/0x4b0
tcp_bpf_sendmsg+0x21b/0x3b0
sock_sendmsg+0x58/0x70
__sys_sendto+0xfa/0x170
? xfd_validate_state+0x1d/0x80
? switch_fpu_return+0x59/0xe0
__x64_sys_sendto+0x24/0x30
do_syscall_64+0x37/0x90
entry_SYSCALL_64_after_hwframe+0x63/0xcd
Fixes: cd9733f5d75c ("tcp_bpf: Fix one concurrency problem in the tcp_bpf_send_verdict function") Signed-off-by: Pengcheng Yang <yangpc@wangsu.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Jakub Sitnicki <jakub@cloudflare.com> Link: https://lore.kernel.org/bpf/1669718441-2654-2-git-send-email-yangpc@wangsu.com Signed-off-by: Sasha Levin <sashal@kernel.org>
icmp conntrack will set icmp redirects as RELATED, but icmpv6 will not
do this.
For icmpv6, only icmp errors (code <= 128) are examined for RELATED state.
ICMPV6 Redirects are part of neighbour discovery mechanism, those are
handled by marking a selected subset (e.g. neighbour solicitations) as
UNTRACKED, but not REDIRECT -- they will thus be flagged as INVALID.
Add minimal support for REDIRECTs. No parsing of neighbour options is
added for simplicity, so this will only check that we have the embeeded
original header (ND_OPT_REDIRECT_HDR), and then attempt to do a flow
lookup for this tuple.
Also extend the existing test case to cover redirects.
Fixes: 9fb9cbb1082d ("[NETFILTER]: Add nf_conntrack subsystem.") Reported-by: Eric Garver <eric@garver.life> Link: https://github.com/firewalld/firewalld/issues/1046 Signed-off-by: Florian Westphal <fw@strlen.de> Acked-by: Eric Garver <eric@garver.life> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by going to
err_pm instead of err_clk.
Fixes:f086ba9d5389c ("ASoC: pcm512x: Support mastering BCLK/LRCLK using the PLL")
As comment of pci_get_class() says, it returns a pci_device with its
refcount increased and decreased the refcount for the input parameter
@from if it is not NULL.
If we break the loop in amdgpu_atrm_get_bios() with 'pdev' not NULL, we
need to call pci_dev_put() to decrease the refcount. Add the missing
pci_dev_put() to avoid refcount leak.
Fixes: d38ceaf99ed0 ("drm/amdgpu: add core driver (v4)") Signed-off-by: Xiongfeng Wang <wangxiongfeng2@huawei.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
As comment of pci_get_class() says, it returns a pci_device with its
refcount increased and decreased the refcount for the input parameter
@from if it is not NULL.
If we break the loop in radeon_atrm_get_bios() with 'pdev' not NULL, we
need to call pci_dev_put() to decrease the refcount. Add the missing
pci_dev_put() to avoid refcount leak.
Fixes: d8ade3526b2a ("drm/radeon: handle non-VGA class pci devices with ATRM") Fixes: c61e2775873f ("drm/radeon: split ATRM support out from the ATPX handler (v3)") Signed-off-by: Xiongfeng Wang <wangxiongfeng2@huawei.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Return true early if ASIC is in BACO state already, no need
to talk to SMU. It can fix the issue that driver was not
calling BACO exit at all in runtime pm resume, and a timing
issue leading to a PCI AER error happened eventually.
When registering the mt8173-afe-pcm driver, we are also adding two
components: one is for the PCM DAIs and one is for the HDMI DAIs, but
when debugfs is enabled, we're getting the following issue:
[ 17.279176] debugfs: Directory '11220000.audio-controller' with parent 'mtk-rt5650' already present!
[ 17.288345] debugfs: Directory '11220000.audio-controller' with parent 'mtk-rt5650' already present!
To overcome to that without any potentially big rewrite of this driver,
similarly to what was done in mt8195-afe-pcm, add a debugfs_prefix to
the components before actually adding them.
We see kernel crashes and lockups and KASAN errors related to ax210
firmware crashes. One of the KASAN dumps pointed at the tx path,
and it appears there is indeed a way to double-free an skb.
If iwl_mvm_tx_skb_sta returns non-zero, then the 'skb' sent into the
method will be freed. But, in case where we build TSO skb buffer,
the skb may also be freed in error case. So, return 0 in that particular
error case and do cleanup manually.
BUG: KASAN: use-after-free in __list_del_entry_valid+0x12/0x90
iwlwifi 0000:06:00.0: 0x00000000 | tsf hi
Read of size 8 at addr ffff88813cfa4ba0 by task btserver/9650
The buggy address belongs to the object at ffff88813cfa4b40
which belongs to the cache skbuff_fclone_cache of size 472
The buggy address is located 96 bytes inside of
472-byte region [ffff88813cfa4b40, ffff88813cfa4d18)
Memory state around the buggy address: ffff88813cfa4a80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc ffff88813cfa4b00: fc fc fc fc fc fc fc fc fa fb fb fb fb fb fb fb
>ffff88813cfa4b80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
^ ffff88813cfa4c00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff88813cfa4c80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
==================================================================
If we're asked to recover open state while a delegation return is
outstanding, then the state manager thread cannot use a cached open, so
if the server returns a delegation, we can end up deadlocked behind the
pending delegreturn.
To avoid this problem, let's just ask the server not to give us a
delegation unless we're explicitly reclaiming one.
Fixes: be36e185bd26 ("NFSv4: nfs4_open_recover_helper() must set share access") Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Commit 5441e9dafdfc6dc40 ("[media] s5p-mfc: Core support for MFC v7")
which adds mfc v7 support for Exynos3250 and use the same compatible
string as used by Exynos5240 but both the IPs are a bit different in
terms of IP clock.
Add variant driver data based on the new compatible string
"samsung,exynos3250-mfc" for Exynos3250 SoC.
Suggested-by: Alim Akhtar <alim.akhtar@samsung.com> Fixes: 5441e9dafdfc ("[media] s5p-mfc: Core support for MFC v7") Signed-off-by: Aakarsh Jain <aakarsh.jain@samsung.com> Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
In dvb_register_frontend(), dvb_register_device() is possible to fail
but its return value is ignored.
It will cause use-after-free when module is removed, because in
dvb_unregister_frontend() it tries to unregister a not registered
device.
BUG: KASAN: use-after-free in dvb_remove_device+0x18b/0x1f0 [dvb_core]
Read of size 4 at addr ffff88800dff4824 by task rmmod/428
CPU: 3 PID: 428 Comm: rmmod
Call Trace:
<TASK>
...
dvb_remove_device+0x18b/0x1f0 [dvb_core]
dvb_unregister_frontend+0x7b/0x130 [dvb_core]
vidtv_bridge_remove+0x6e/0x160 [dvb_vidtv_bridge]
...
Fix this by catching return value of dvb_register_device().
However the fe->refcount can't be put to zero immediately, because
there are still modules calling dvb_frontend_detach() when
dvb_register_frontend() fails.
while the imx8mn_enet_phy_sels list didn't contained audio_pll1_out for
source select bits 101b.
Fixes: 96d6392b54dbb ("clk: imx: Add support for i.MX8MN clock driver") Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com> Acked-by: Marco Felsch <m.felsch@pengutronix.de> Reviewed-by: Abel Vesa <abel.vesa@linaro.org> Signed-off-by: Abel Vesa <abel.vesa@linaro.org> Link: https://lore.kernel.org/r/20221117113637.1978703-6-dario.binacchi@amarulasolutions.com Signed-off-by: Sasha Levin <sashal@kernel.org>
while the imx8mn_sai2_sels list contained clk_ext3 and clk_ext4 for
source select bits 110b and 111b.
Fixes: 96d6392b54dbb ("clk: imx: Add support for i.MX8MN clock driver") Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com> Acked-by: Marco Felsch <m.felsch@pengutronix.de> Reviewed-by: Abel Vesa <abel.vesa@linaro.org> Signed-off-by: Abel Vesa <abel.vesa@linaro.org> Link: https://lore.kernel.org/r/20221117113637.1978703-5-dario.binacchi@amarulasolutions.com Signed-off-by: Sasha Levin <sashal@kernel.org>
There is no occurrence of the hdmi oscillator in the reference manual
(document IMX8MNRM Rev 2, 07/2022). Further, if we consider the indexes
76-81 and 134 of the "Clock Root" table of chapter 5 of the RM, there is
no entry for the source select bits 101b, which is the setting referenced
by "osc_hdmi".
Fix by renaming "osc_hdmi" with "dummy", a clock which has already been
used for missing source select bits.
Tested on the BSH SystemMaster (SMM) S2 board.
Fixes: 96d6392b54dbb ("clk: imx: Add support for i.MX8MN clock driver") Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com> Acked-by: Marco Felsch <m.felsch@pengutronix.de> Signed-off-by: Abel Vesa <abel.vesa@linaro.org> Link: https://lore.kernel.org/r/20221117113637.1978703-3-dario.binacchi@amarulasolutions.com Signed-off-by: Sasha Levin <sashal@kernel.org>
The IMX8MN platform does not have any video processing unit (VPU), and
indeed in the reference manual (document IMX8MNRM Rev 2, 07/2022) there
is no occurrence of its pll. From an analysis of the code and the RM
itself, I think vpu pll is used instead of m7 alternate pll, probably
for copy and paste of code taken from modules of similar architectures.
As an example for all, if we consider the second row of the "Clock Root"
table of chapter 5 (Clocks and Power Management) of the RM:
but in the source code, the imx8mn_m7_sels clocks list contains vpu_pll
for the source select bits 011b.
So, let's rename "vpu_pll" to "m7_alt_pll" to be consistent with the RM.
The IMX8MN_VPU_* constants have not been removed to ensure backward
compatibility of the patch.
No functional changes intended.
Fixes: 96d6392b54dbb ("clk: imx: Add support for i.MX8MN clock driver") Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com> Acked-by: Marco Felsch <m.felsch@pengutronix.de> Reviewed-by: Abel Vesa <abel.vesa@linaro.org> Signed-off-by: Abel Vesa <abel.vesa@linaro.org> Link: https://lore.kernel.org/r/20221117113637.1978703-2-dario.binacchi@amarulasolutions.com Signed-off-by: Sasha Levin <sashal@kernel.org>
The function send_packet() has a race condition as follows:
func send_packet()
{
// do work
call usb_submit_urb()
mutex_unlock()
wait_for_event_interruptible() <-- lock gone
mutex_lock()
}
func vfd_write()
{
mutex_lock()
call send_packet() <- prev call is not completed
mutex_unlock()
}
When the mutex is unlocked and the function send_packet() waits for the
call to complete, vfd_write() can start another call, which leads to the
"URB submitted while active" warning in usb_submit_urb().
Fix this by removing the mutex_unlock() call in send_packet() and using
mutex_lock_interruptible().
In vimc_init(), when platform_driver_register(&vimc_pdrv) fails,
platform_driver_unregister(&vimc_pdrv) is wrongly called rather than
platform_device_unregister(&vimc_pdev), which causes kernel warning:
A submitter workqueue is dynamically allocated by init_submitter()
called by drbd_create_device(), we should destroy it when this
device is not needed or destroyed.
Fixes: 113fef9e20e0 ("drbd: prepare to queue write requests on a submit worker") Signed-off-by: Wang ShaoBo <bobo.shaobowang@huawei.com> Link: https://lore.kernel.org/r/20221124015817.2729789-3-bobo.shaobowang@huawei.com Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Sasha Levin <sashal@kernel.org>
This revert c2258ffc56f2 ("drbd: poison free'd device, resource and
connection structs"), add memset is odd here for debugging, there are
some methods to accurately show what happened, such as kdump.
Before this change when a bond in mode 2 lost link, all of its slaves
lost link, the bonding device would never recover even after the
expiration of updelay. This change removes the updelay when the bond
currently has no usable links. Conforming to bonding.txt section 13.1
paragraph 4.
Fixes: 41f891004063 ("bonding: ignore updelay param when there is no active slave") Signed-off-by: Jonathan Toppins <jtoppins@redhat.com> Acked-by: Jay Vosburgh <jay.vosburgh@canonical.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
As comment of pci_get_domain_bus_and_slot() says, it returns
a pci device with refcount increment, when finish using it,
the caller must decrement the reference count by calling
pci_dev_put().
So before returning from amdgpu_device_resume|suspend_display_audio(),
pci_dev_put() is called to avoid refcount leak.
Fixes: 3f12acc8d6d4 ("drm/amdgpu: put the audio codec into suspend state before gpu reset V3") Reviewed-by: Evan Quan <evan.quan@amd.com> Signed-off-by: Yang Yingliang <yangyingliang@huawei.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
If clk_register() fails, @pll->rate_table may have allocated memory by
kmemdup(), so it needs to be freed, otherwise will cause memory leak
issue, this patch fixes it.
Fixes: 90c590254051 ("clk: rockchip: add clock type for pll clocks and pll used on rk3066") Signed-off-by: Xiu Jianfeng <xiujianfeng@huawei.com> Link: https://lore.kernel.org/r/20221123091201.199819-1-xiujianfeng@huawei.com Signed-off-by: Heiko Stuebner <heiko@sntech.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
Use kfree_const() to free supply_name conditionally in create_regulator()
as supply_name may be allocated from kmalloc() or directly from .rodata
section.
Fixes: 87fe29b61f95 ("regulator: push allocations in create_regulator() outside of lock") Signed-off-by: Wang ShaoBo <bobo.shaobowang@huawei.com> Link: https://lore.kernel.org/r/20221123034616.3609537-1-bobo.shaobowang@huawei.com Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Shifting signed 32-bit value by 31 bits is undefined, so changing
significant bit to unsigned. The UBSAN warning calltrace like below:
UBSAN: shift-out-of-bounds in sound/core/seq/seq_clientmgr.c:509:22
left shift of 1 by 31 places cannot be represented in type 'int'
...
Call Trace:
<TASK>
dump_stack_lvl+0x8d/0xcf
ubsan_epilogue+0xa/0x44
__ubsan_handle_shift_out_of_bounds+0x1e7/0x208
snd_seq_deliver_single_event.constprop.21+0x191/0x2f0
snd_seq_deliver_event+0x1a2/0x350
snd_seq_kernel_client_dispatch+0x8b/0xb0
snd_seq_client_notify_subscription+0x72/0xa0
snd_seq_ioctl_subscribe_port+0x128/0x160
snd_seq_kernel_client_ctl+0xce/0xf0
snd_seq_oss_create_client+0x109/0x15b
alsa_seq_oss_init+0x11c/0x1aa
do_one_initcall+0x80/0x440
kernel_init_freeable+0x370/0x3c3
kernel_init+0x1b/0x190
ret_from_fork+0x1f/0x30
</TASK>
Shifting signed 32-bit value by 31 bits is undefined, so changing
significant bit to unsigned. The UBSAN warning calltrace like below:
UBSAN: shift-out-of-bounds in sound/core/pcm_native.c:2676:21
left shift of 1 by 31 places cannot be represented in type 'int'
...
Call Trace:
<TASK>
dump_stack_lvl+0x8d/0xcf
ubsan_epilogue+0xa/0x44
__ubsan_handle_shift_out_of_bounds+0x1e7/0x208
snd_pcm_open_substream+0x9f0/0xa90
snd_pcm_oss_open.part.26+0x313/0x670
snd_pcm_oss_open+0x30/0x40
soundcore_open+0x18b/0x2e0
chrdev_open+0xe2/0x270
do_dentry_open+0x2f7/0x620
path_openat+0xd66/0xe70
do_filp_open+0xe3/0x170
do_sys_openat2+0x357/0x4a0
do_sys_open+0x87/0xd0
do_syscall_64+0x34/0x80
Since for_each_available_child_of_node() will increase the refcount of
node, we need to call of_node_put() manually when breaking out of the
iteration.
This is no bugfix (so no Fixes: tag is necessary) as it is
taken care of in hid_sensor_custom_add_attributes().
The motivation for this patch is that:
hid_sensor_custom_field.attr_name and
hid_sensor_custom_field.attrs
has the size of HID_CUSTOM_TOTAL_ATTRS and used in same context.
We compare against HID_CUSTOM_TOTAL_ATTRS when
looping through hid_custom_attrs.
We will silent the smatch error:
hid_sensor_custom_add_attributes() error: buffer overflow
'hid_custom_attrs' 8 <= 10
Signed-off-by: Marcus Folkesson <marcus.folkesson@gmail.com> Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz> Signed-off-by: Sasha Levin <sashal@kernel.org>
The number of bytes used by spi_nor_spimem_check_readop() may be
incorrect for the dummy cycles. Since nor->read_dummy is not initialized
before spi_nor_spimem_adjust_hwcaps().
We use both mode and wait state clock cycles instead of nor->read_dummy.
dma_alloc_coherent does not return a physical address, but a DMA address,
which might be remapped or have an offset. Passing the DMA address to
vm_iomap_memory is thus broken.
Use the proper dma_mmap_coherent helper instead, and stop passing
__GFP_COMP to dma_alloc_coherent, as the memory management inside the
DMA allocator is hidden from the callers and does not require it.
With this the gfp_t argument to __videobuf_dc_alloc can be removed and
hard coded to GFP_KERNEL.
A problem about modprobe s5p_fimc failed is triggered with the
following log given:
[ 272.075275] Error: Driver 'exynos4-fimc' is already registered, aborting...
modprobe: ERROR: could not insert 's5p_fimc': Device or resource busy
The reason is that fimc_md_init() returns platform_driver_register()
directly without checking its return value, if platform_driver_register()
failed, it returns without unregister fimc_driver, resulting the
s5p_fimc can never be installed later.
A simple call graph is shown as below:
fimc_md_init()
fimc_register_driver() # register fimc_driver
platform_driver_register()
platform_driver_register()
driver_register()
bus_add_driver()
dev = kzalloc(...) # OOM happened
# return without unregister fimc_driver
Fix by unregister fimc_driver when platform_driver_register() returns
error.
Fixes: d3953223b090 ("[media] s5p-fimc: Add the media device driver") Signed-off-by: Yuan Can <yuancan@huawei.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Sasha Levin <sashal@kernel.org>
If device_register() returns error in solo_sysfs_init(), the
name allocated by dev_set_name() need be freed. As comment of
device_register() says, it should use put_device() to give up
the reference in the error path. So fix this by calling
put_device(), then the name can be freed in kobject_cleanup().
Fixes: dcae5dacbce5 ("[media] solo6x10: sync to latest code from Bluecherry's git repo") Signed-off-by: Yang Yingliang <yangyingliang@huawei.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Sasha Levin <sashal@kernel.org>
It is because the error handling in vidtv_bridge_dvb_init() is wrong.
First, vidtv_bridge_dmx(dev)_init() will clean themselves when fail, but
goto fail_dmx(_dev): calls release functions again, which causes
use-after-free.
Also, in fail_fe, fail_tuner_probe and fail_demod_probe, j = i will cause
out-of-bound when i finished its loop (i == NUM_FE). And the loop
releasing is wrong, although now NUM_FE is 1 so it won't cause problem.
Fix this by correctly releasing everything.
Fixes: f90cf6079bf6 ("media: vidtv: add a bridge driver") Signed-off-by: Chen Zhongjin <chenzhongjin@huawei.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Sasha Levin <sashal@kernel.org>
As can be seen in elants_i2c_power_off(), we want the reset GPIO
asserted when power is off. The reset GPIO is active low so we need
the reset line logic low when power is off to avoid leakage.
We have a problem, though, at probe time. At probe time we haven't
powered the regulators on yet but we have:
While that _looks_ right, it turns out that it's not. The
GPIOD_OUT_LOW doesn't mean to init the GPIO to low. It means init the
GPIO to "not asserted". Since this is an active low GPIO that inits it
to be high.
Let's fix this to properly init the GPIO. Now after both probe and
power off the state of the GPIO is consistent (it's "asserted" or
level low).
Once we fix this, we can see that at power on time we no longer to
assert the reset GPIO as the first thing. The reset GPIO is _always_
asserted before powering on. Let's fix powering on to account for
this.
If we get an error (other than -ENOENT) we need to propagate that up the
stack. Otherwise if the nvmem driver hasn't probed yet, we'll end up
end up claiming that we support all the OPPs which is not likely to be
true (and on some generations impossible to be true, ie. if there are
conflicting OPPs).
Fixes: fe7952c629da ("drm/msm: Add speed-bin support to a618 gpu") Signed-off-by: Rob Clark <robdclark@chromium.org> Reviewed-by: Douglas Anderson <dianders@chromium.org> Reviewed-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/511690/ Link: https://lore.kernel.org/r/20221115154637.1613968-1-robdclark@gmail.com Signed-off-by: Sasha Levin <sashal@kernel.org>
This driver is attempting to register to support two different buses.
if either of these is successful then ath10k_pci_init() should return 0
so that hardware attached to the successful bus can be probed and
supported. only if both of these are unsuccessful should ath10k_pci_init()
return an errno.
Fixes: 0b523ced9a3c ("ath10k: add basic skeleton to support ahb") Signed-off-by: Xiu Jianfeng <xiujianfeng@huawei.com> Reviewed-by: Jeff Johnson <quic_jjohnson@quicinc.com> Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com> Link: https://lore.kernel.org/r/20221110061926.18163-1-xiujianfeng@huawei.com Signed-off-by: Sasha Levin <sashal@kernel.org>
Zero out the pointer to ->slave_dir so that the holder code doesn't
incorrectly treat the object as alive when add_disk failed or after
del_gendisk was called.
The input parameter @fields is type of struct ima_template_field ***, so
when allocates array memory for @fields, the size of element should be
sizeof(**field) instead of sizeof(*field).
Actually the original code would not cause any runtime error, but it's
better to make it logically right.
In the PP_OD_EDIT_VDDC_CURVE case the "input_index" variable is capped at
2 but not checked for negative values so it results in an out of bounds
read. This value comes from the user via sysfs.
Fixes: d5bf26539494 ("drm/amd/powerplay: added vega20 overdrive support V3") Signed-off-by: Dan Carpenter <error27@gmail.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
In of_get_regulator(), the node is returned from of_parse_phandle()
with refcount incremented, after using it, of_node_put() need be called.
Fixes: 69511a452e6d ("regulator: map consumer regulator based on device tree") Signed-off-by: Yang Yingliang <yangyingliang@huawei.com> Link: https://lore.kernel.org/r/20221115091508.900752-1-yangyingliang@huawei.com Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
GCC 11.3.0 fails to compile btf_dump.c due to the following error,
which seems to originate in btf_dump_struct_data where the returned
value would be uninitialized if btf_vlen returns zero.
btf_dump.c: In function ‘btf_dump_dump_type_data’:
btf_dump.c:2363:12: error: ‘err’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
2363 | if (err < 0)
| ^
Fixes: 920d16af9b42 ("libbpf: BTF dumper support for typed data") Signed-off-by: David Michael <fedora.dm0@gmail.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Stanislav Fomichev <sdf@google.com> Acked-by: Alan Maguire <alan.maguire@oracle.com> Link: https://lore.kernel.org/bpf/87zgcu60hq.fsf@gmail.com Signed-off-by: Sasha Levin <sashal@kernel.org>
When the radeon driver reads the bios information from ACPI
table in radeon_acpi_vfct_bios(), it misses to call acpi_put_table()
to release the ACPI memory after the init, so add acpi_put_table()
properly to fix the memory leak.
This fixes crashes in bfq_add_bfqq_busy due to waker_bfqq being NULL,
but woken_list_node still being hashed. This would happen when
bfq_init_rq() expects a brand new allocated queue to be returned from
bfq_get_bfqq_handle_split() and unconditionally updates waker_bfqq
without resetting woken_list_node. Since we can always return oom_bfqq
when attempting to allocate, we cannot assume waker_bfqq starts as NULL.
Avoid setting woken_bfqq for oom_bfqq entirely, as it's not useful.
Provide a CONFIG_PROC_FS=n fallback for proc_create_net_single_write().
Also provide a fallback for proc_create_net_data_write().
Fixes: 564def71765c ("proc: Add a way to make network proc files writable") Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: David Howells <dhowells@redhat.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org
cc: netdev@vger.kernel.org Signed-off-by: Sasha Levin <sashal@kernel.org>
It is required to return the received buffers, if streaming can not be
started. For instance media_pipeline_start() may fail with EPIPE, if
a link validation between entities is not passed, and in such a case
a user gets a kernel warning:
When using wpa_supplicant v2.10, this driver is no longer able to
associate with any AP and fails in the EAPOL 4-way handshake while
sending the 2/4 message to the AP. The problem is not present in
wpa_supplicant v2.9 or older. The problem stems from HostAP commit 144314eaa ("wpa_supplicant: Send EAPOL frames over nl80211 where available")
which changes the way EAPOL frames are sent, from them being send
at L2 frames to them being sent via nl80211 control port.
An EAPOL frame sent as L2 frame is passed to the WiFi driver with
skb->protocol ETH_P_PAE, while EAPOL frame sent via nl80211 control
port has skb->protocol set to ETH_P_802_3 . The later happens in
ieee80211_tx_control_port(), where the EAPOL frame is encapsulated
into 802.3 frame.
The rsi_91x driver handles ETH_P_PAE EAPOL frames as high-priority
frames and sends them via highest-priority transmit queue, while
the ETH_P_802_3 frames are sent as regular frames. The EAPOL 4-way
handshake frames must be sent as highest-priority, otherwise the
4-way handshake times out.
Therefore, to fix this problem, inspect the skb control flags and
if flag IEEE80211_TX_CTRL_PORT_CTRL_PROTO is set, assume this is
an EAPOL frame and transmit the frame via high-priority queue just
like other ETH_P_PAE frames.
Fixes: 0eb42586cf87 ("rsi: data packet descriptor enhancements") Signed-off-by: Marek Vasut <marex@denx.de> Signed-off-by: Kalle Valo <kvalo@kernel.org> Link: https://lore.kernel.org/r/20221104163339.227432-1-marex@denx.de Signed-off-by: Sasha Levin <sashal@kernel.org>
USB controllers on SM8250 doesn't work after coming back from suspend.
This can be fixed by keeping the USB GDSCs in retention mode so that
hardware can keep them ON and put into rentention mode once the parent
domain goes to a low power state.
When equivalent completed state is found and it has additional precision
restrictions, BPF verifier propagates precision to
currently-being-verified state chain (i.e., including parent states) so
that if some of the states in the chain are not yet completed, necessary
precision restrictions are enforced.
Unfortunately, right now this happens only for the last frame (deepest
active subprogram's frame), not all the frames. This can lead to
incorrect matching of states due to missing precision marker. Currently
this doesn't seem possible as BPF verifier forces everything to precise
when validated BPF program has any subprograms. But with the next patch
lifting this restriction, this becomes problematic.
In fact, without this fix, we'll start getting failure in one of the
existing test_verifier test cases:
Every 8 bytes of the stack is tracked by a bpf_stack_state.
Within each bpf_stack_state, there is a 'u8 slot_type[8]' to track
the type of each byte. Verifier tests slot_type[0] == STACK_SPILL
to decide if the spilled reg state is saved. Verifier currently only
saves the reg state if the whole 8 bytes are spilled to the stack,
so checking the slot_type[7] is the same as checking slot_type[0].
The later patch will allow verifier to save the bounded scalar
reg also for <8 bytes spill. There is a llvm patch [1] to ensure
the <8 bytes spill will be 8-byte aligned, so checking
slot_type[7] instead of slot_type[0] is required.
While at it, this patch refactors the slot_type[0] == STACK_SPILL
test into a new function is_spilled_reg() and change the
slot_type[0] check to slot_type[7] check in there also.
Signed-off-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20210922004934.624194-1-kafai@fb.com
Stable-dep-of: 529409ea92d5 ("bpf: propagate precision across all frames, not just the last one") Signed-off-by: Sasha Levin <sashal@kernel.org>
When processing ALU/ALU64 operations (apart from BPF_MOV, which is
handled correctly already; and BPF_NEG and BPF_END are special and don't
have source register), if destination register is already marked
precise, this causes problem with potentially missing precision tracking
for the source register. E.g., when we have r1 >>= r5 and r1 is marked
precise, but r5 isn't, this will lead to r5 staying as imprecise. This
is due to the precision backtracking logic stopping early when it sees
r1 is already marked precise. If r1 wasn't precise, we'd keep
backtracking and would add r5 to the set of registers that need to be
marked precise. So there is a discrepancy here which can lead to invalid
and incompatible states matched due to lack of precision marking on r5.
If r1 wasn't precise, precision backtracking would correctly mark both
r1 and r5 as precise.
This is simple to fix, though. During the forward instruction simulation
pass, for arithmetic operations of `scalar <op>= scalar` form (where
<op> is ALU or ALU64 operations), if destination register is already
precise, mark source register as precise. This applies only when both
involved registers are SCALARs. `ptr += scalar` and `scalar += ptr`
cases are already handled correctly.
This does have (negative) effect on some selftest programs and few
Cilium programs. ~/baseline-tmp-results.csv are veristat results with
this patch, while ~/baseline-results.csv is without it. See post
scriptum for instructions on how to make Cilium programs testable with
veristat. Correctness has a price.
$ ./veristat -C -e file,prog,insns,states ~/baseline-results.csv ~/baseline-tmp-results.csv | grep -v '+0'
File Program Total insns (A) Total insns (B) Total insns (DIFF) Total states (A) Total states (B) Total states (DIFF)
----------------------- -------------------- --------------- --------------- ------------------ ---------------- ---------------- -------------------
bpf_cubic.bpf.linked1.o bpf_cubic_cong_avoid 997 1700 +703 (+70.51%) 62 90 +28 (+45.16%)
test_l4lb.bpf.linked1.o balancer_ingress 4559 5469 +910 (+19.96%) 118 126 +8 (+6.78%)
----------------------- -------------------- --------------- --------------- ------------------ ---------------- ---------------- -------------------
$ ./veristat -C -e file,prog,verdict,insns,states ~/baseline-results-cilium.csv ~/baseline-tmp-results-cilium.csv | grep -v '+0'
File Program Total insns (A) Total insns (B) Total insns (DIFF) Total states (A) Total states (B) Total states (DIFF)
------------- ------------------------------ --------------- --------------- ------------------ ---------------- ---------------- -------------------
bpf_host.o tail_nodeport_nat_ingress_ipv6 4448 5261 +813 (+18.28%) 234 247 +13 (+5.56%)
bpf_host.o tail_nodeport_nat_ipv6_egress 3396 3446 +50 (+1.47%) 201 203 +2 (+1.00%)
bpf_lxc.o tail_nodeport_nat_ingress_ipv6 4448 5261 +813 (+18.28%) 234 247 +13 (+5.56%)
bpf_overlay.o tail_nodeport_nat_ingress_ipv6 4448 5261 +813 (+18.28%) 234 247 +13 (+5.56%)
bpf_xdp.o tail_lb_ipv4 71736 73442 +1706 (+2.38%) 4295 4370 +75 (+1.75%)
------------- ------------------------------ --------------- --------------- ------------------ ---------------- ---------------- -------------------
P.S. To make Cilium ([0]) programs libbpf-compatible and thus
veristat-loadable, apply changes from topmost commit in [1], which does
minimal changes to Cilium source code, mostly around SEC() annotations
and BPF map definitions.
devm_pinctrl_get() may return ERR_PTR(-EPROBE_DEFER), add a minus sign
to fix it.
Fixes: 4163851f7b99 ("[media] s5p-fimc: Use pinctrl API for camera ports configuration") Signed-off-by: Yang Yingliang <yangyingliang@huawei.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Sasha Levin <sashal@kernel.org>
This is because we forget to check boundary after adjust compose->height
int V4L2_SEL_TGT_CROP case. Add v4l2_rect_map_inside() to fix this problem
for this case.
Fixes: ef834f7836ec ("[media] vivid: add the video capture and output parts") Signed-off-by: Liu Shixin <liushixin2@huawei.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Sasha Levin <sashal@kernel.org>
For the case where allow_ptr_leaks is false, code is checking whether
slot type is STACK_INVALID and STACK_SPILL and rejecting other cases.
This is a consequence of incorrectly checking for register type instead
of the slot type (NOT_INIT and SCALAR_VALUE respectively). Fix the
check.
projects / users / dwmw2 / linux.git / log