powerpc/85xx: Fix oops when CONFIG_FSL_PMC=n

[ Upstream commit 3dc709e518b47386e6af937eaec37bb36539edfd ]

When CONFIG_FSL_PMC is set to n, no value is assigned to cpu_up_prepare
in the mpc85xx_pm_ops structure. As a result, oops is triggered in
smp_85xx_start_cpu().

  smp: Bringing up secondary CPUs ...
  kernel tried to execute user page (0) - exploit attempt? (uid: 0)
  BUG: Unable to handle kernel instruction fetch (NULL pointer?)
  Faulting instruction address: 0x00000000
  Oops: Kernel access of bad area, sig: 11 [#1]
  ...
  NIP [00000000] 0x0
  LR [c0021d2c] smp_85xx_kick_cpu+0xe8/0x568
  Call Trace:
  [c1051da8] [c0021cb8] smp_85xx_kick_cpu+0x74/0x568 (unreliable)
  [c1051de8] [c0011460] __cpu_up+0xc0/0x228
  [c1051e18] [c0031bbc] bringup_cpu+0x30/0x224
  [c1051e48] [c0031f3c] cpu_up.constprop.0+0x180/0x33c
  [c1051e88] [c00322e8] bringup_nonboot_cpus+0x88/0xc8
  [c1051eb8] [c07e67bc] smp_init+0x30/0x78
  [c1051ed8] [c07d9e28] kernel_init_freeable+0x118/0x2a8
  [c1051f18] [c00032d8] kernel_init+0x14/0x124
  [c1051f38] [c0010278] ret_from_kernel_thread+0x14/0x1c

Fixes: c45361abb918 ("powerpc/85xx: fix timebase sync issue when CONFIG_HOTPLUG_CPU=n")
Reported-by: Martin Kennedy <hurricos@gmail.com>
Signed-off-by: Xiaoming Ni <nixiaoming@huawei.com>
Tested-by: Martin Kennedy <hurricos@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20211126041153.16926-1-nixiaoming@huawei.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

bpf, selftests: Fix racing issue in btf_skc_cls_ingress test

[ Upstream commit c2fcbf81c332b42382a0c439bfe2414a241e4f5b ]

The libbpf CI reported occasional failure in btf_skc_cls_ingress:

  test_syncookie:FAIL:Unexpected syncookie states gen_cookie:80326634 recv_cookie:0
  bpf prog error at line 97

"error at line 97" means the bpf prog cannot find the listening socket
when the final ack is received.  It then skipped processing
the syncookie in the final ack which then led to "recv_cookie:0".

The problem is the userspace program did not do accept() and went
ahead to close(listen_fd) before the kernel (and the bpf prog) had
a chance to process the final ack.

The fix is to add accept() call so that the userspace will wait for
the kernel to finish processing the final ack first before close()-ing
everything.

Fixes: 9a856cae2217 ("bpf: selftest: Add test_btf_skc_cls_ingress")
Reported-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20211216191630.466151-1-kafai@fb.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

bpf: Fix extable fixup offset.

[ Upstream commit 433956e91200734d09958673a56df02d00a917c2 ]

The prog - start_of_ldx is the offset before the faulting ldx to the location
after it, so this will be used to adjust pt_regs->ip for jumping over it and
continuing, and with old temp it would have been fixed up to the wrong offset,
causing crash.

Fixes: 4c5de127598e ("bpf: Emit explicit NULL pointer checks for PROBE_LDX instructions.")
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Reviewed-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: kexec: Fix missing error code 'ret' warning in load_other_segments()

[ Upstream commit 9c5d89bc10551f1aecd768b00fca3339a7b8c8ee ]

Since commit ac10be5cdbfa ("arm64: Use common
of_kexec_alloc_and_setup_fdt()"), smatch reports the following warning:

  arch/arm64/kernel/machine_kexec_file.c:152 load_other_segments()
  warn: missing error code 'ret'

Return code is not set to an error code in load_other_segments() when
of_kexec_alloc_and_setup_fdt() call returns a NULL dtb. This results
in status success (return code set to 0) being returned from
load_other_segments().

Set return code to -EINVAL if of_kexec_alloc_and_setup_fdt() returns
NULL dtb.

Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com>
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Fixes: ac10be5cdbfa ("arm64: Use common of_kexec_alloc_and_setup_fdt()")
Link: https://lore.kernel.org/r/20211210010121.101823-1-nramas@linux.microsoft.com
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

afs: Fix mmap

[ Upstream commit 1744a22ae948799da7927b53ec97ccc877ff9d61 ]

Fix afs_add_open_map() to check that the vnode isn't already on the list
when it adds it.  It's possible that afs_drop_open_mmap() decremented
the cb_nr_mmap counter, but hadn't yet got into the locked section to
remove it.

Also vnode->cb_mmap_link should be initialised, so fix that too.

Fixes: 6e0e99d58a65 ("afs: Fix mmap coherency vs 3rd-party changes")
Reported-by: kafs-testing+fedora34_64checkkafs-build-300@auristor.com
Suggested-by: Marc Dionne <marc.dionne@auristor.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-by: kafs-testing+fedora34_64checkkafs-build-300@auristor.com
cc: linux-afs@lists.infradead.org
Link: https://lore.kernel.org/r/686465.1639435380@warthog.procyon.org.uk/
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

sit: do not call ipip6_dev_free() from sit_init_net()

[ Upstream commit e28587cc491ef0f3c51258fdc87fbc386b1d4c59 ]

ipip6_dev_free is sit dev->priv_destructor, already called
by register_netdevice() if something goes wrong.

Alternative would be to make ipip6_dev_free() robust against
multiple invocations, but other drivers do not implement this
strategy.

syzbot reported:

dst_release underflow
WARNING: CPU: 0 PID: 5059 at net/core/dst.c:173 dst_release+0xd8/0xe0 net/core/dst.c:173
Modules linked in:
CPU: 1 PID: 5059 Comm: syz-executor.4 Not tainted 5.16.0-rc5-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:dst_release+0xd8/0xe0 net/core/dst.c:173
Code: 4c 89 f2 89 d9 31 c0 5b 41 5e 5d e9 da d5 44 f9 e8 1d 90 5f f9 c6 05 87 48 c6 05 01 48 c7 c7 80 44 99 8b 31 c0 e8 e8 67 29 f9 <0f> 0b eb 85 0f 1f 40 00 53 48 89 fb e8 f7 8f 5f f9 48 83 c3 a8 48
RSP: 0018:ffffc9000aa5faa0 EFLAGS: 00010246
RAX: d6894a925dd15a00 RBX: 00000000ffffffff RCX: 0000000000040000
RDX: ffffc90005e19000 RSI: 000000000003ffff RDI: 0000000000040000
RBP: 0000000000000000 R08: ffffffff816a1f42 R09: ffffed1017344f2c
R10: ffffed1017344f2c R11: 0000000000000000 R12: 0000607f462b1358
R13: 1ffffffff1bfd305 R14: ffffe8ffffcb1358 R15: dffffc0000000000
FS:  00007f66c71a2700(0000) GS:ffff8880b9a00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f88aaed5058 CR3: 0000000023e0f000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <TASK>
 dst_cache_destroy+0x107/0x1e0 net/core/dst_cache.c:160
 ipip6_dev_free net/ipv6/sit.c:1414 [inline]
 sit_init_net+0x229/0x550 net/ipv6/sit.c:1936
 ops_init+0x313/0x430 net/core/net_namespace.c:140
 setup_net+0x35b/0x9d0 net/core/net_namespace.c:326
 copy_net_ns+0x359/0x5c0 net/core/net_namespace.c:470
 create_new_namespaces+0x4ce/0xa00 kernel/nsproxy.c:110
 unshare_nsproxy_namespaces+0x11e/0x180 kernel/nsproxy.c:226
 ksys_unshare+0x57d/0xb50 kernel/fork.c:3075
 __do_sys_unshare kernel/fork.c:3146 [inline]
 __se_sys_unshare kernel/fork.c:3144 [inline]
 __x64_sys_unshare+0x34/0x40 kernel/fork.c:3144
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x44/0xd0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x7f66c882ce99
Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 bc ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f66c71a2168 EFLAGS: 00000246 ORIG_RAX: 0000000000000110
RAX: ffffffffffffffda RBX: 00007f66c893ff60 RCX: 00007f66c882ce99
RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000048040200
RBP: 00007f66c8886ff1 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007fff6634832f R14: 00007f66c71a2300 R15: 0000000000022000
 </TASK>

Fixes: cf124db566e6 ("net: Fix inconsistent teardown and release of private netdev state.")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Link: https://lore.kernel.org/r/20211216111741.1387540-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: systemport: Add global locking for descriptor lifecycle

[ Upstream commit 8b8e6e782456f1ce02a7ae914bbd5b1053f0b034 ]

The descriptor list is a shared resource across all of the transmit queues, and
the locking mechanism used today only protects concurrency across a given
transmit queue between the transmit and reclaiming. This creates an opportunity
for the SYSTEMPORT hardware to work on corrupted descriptors if we have
multiple producers at once which is the case when using multiple transmit
queues.

This was particularly noticeable when using multiple flows/transmit queues and
it showed up in interesting ways in that UDP packets would get a correct UDP
header checksum being calculated over an incorrect packet length. Similarly TCP
packets would get an equally correct checksum computed by the hardware over an
incorrect packet length.

The SYSTEMPORT hardware maintains an internal descriptor list that it re-arranges
when the driver produces a new descriptor anytime it writes to the
WRITE_PORT_{HI,LO} registers, there is however some delay in the hardware to
re-organize its descriptors and it is possible that concurrent TX queues
eventually break this internal allocation scheme to the point where the
length/status part of the descriptor gets used for an incorrect data buffer.

The fix is to impose a global serialization for all TX queues in the short
section where we are writing to the WRITE_PORT_{HI,LO} registers which solves
the corruption even with multiple concurrent TX queues being used.

Fixes: 80105befdb4b ("net: systemport: add Broadcom SYSTEMPORT Ethernet MAC driver")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20211215202450.4086240-1-f.fainelli@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net/smc: Prevent smc_release() from long blocking

[ Upstream commit 5c15b3123f65f8fbb1b445d9a7e8812e0e435df2 ]

In nginx/wrk benchmark, there's a hung problem with high probability
on case likes that: (client will last several minutes to exit)

server: smc_run nginx

client: smc_run wrk -c 10000 -t 1 http://server

Client hangs with the following backtrace:

0 [ffffa7ce8Of3bbf8] __schedule at ffffffff9f9eOd5f
1 [ffffa7ce8Of3bc88] schedule at ffffffff9f9eløe6
2 [ffffa7ce8Of3bcaO] schedule_timeout at ffffffff9f9e3f3c
3 [ffffa7ce8Of3bd2O] wait_for_common at ffffffff9f9el9de
4 [ffffa7ce8Of3bd8O] __flush_work at ffffffff9fOfeOl3
5 [ffffa7ce8øf3bdfO] smc_release at ffffffffcO697d24 [smc]
6 [ffffa7ce8Of3be2O] __sock_release at ffffffff9f8O2e2d
7 [ffffa7ce8Of3be4ø] sock_close at ffffffff9f8ø2ebl
8 [ffffa7ce8øf3be48] __fput at ffffffff9f334f93
9 [ffffa7ce8Of3be78] task_work_run at ffffffff9flOlff5
10 [ffffa7ce8Of3beaO] do_exit at ffffffff9fOe5Ol2
11 [ffffa7ce8Of3bflO] do_group_exit at ffffffff9fOe592a
12 [ffffa7ce8Of3bf38] __x64_sys_exit_group at ffffffff9fOe5994
13 [ffffa7ce8Of3bf4O] do_syscall_64 at ffffffff9f9d4373
14 [ffffa7ce8Of3bfsO] entry_SYSCALL_64_after_hwframe at ffffffff9fa0007c

This issue dues to flush_work(), which is used to wait for
smc_connect_work() to finish in smc_release(). Once lots of
smc_connect_work() was pending or all executing work dangling,
smc_release() has to block until one worker comes to free, which
is equivalent to wait another smc_connnect_work() to finish.

In order to fix this, There are two changes:

1. For those idle smc_connect_work(), cancel it from the workqueue; for
   executing smc_connect_work(), waiting for it to finish. For that
   purpose, replace flush_work() with cancel_work_sync().

2. Since smc_connect() hold a reference for passive closing, if
   smc_connect_work() has been cancelled, release the reference.

Fixes: 24ac3a08e658 ("net/smc: rebuild nonblocking connect")
Reported-by: Tony Lu <tonylu@linux.alibaba.com>
Tested-by: Dust Li <dust.li@linux.alibaba.com>
Reviewed-by: Dust Li <dust.li@linux.alibaba.com>
Reviewed-by: Tony Lu <tonylu@linux.alibaba.com>
Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
Acked-by: Karsten Graul <kgraul@linux.ibm.com>
Link: https://lore.kernel.org/r/1639571361-101128-1-git-send-email-alibuda@linux.alibaba.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: Fix double 0x prefix print in SKB dump

[ Upstream commit 8a03ef676ade55182f9b05115763aeda6dc08159 ]

When printing netdev features %pNF already takes care of the 0x prefix,
remove the explicit one.

Fixes: 6413139dfc64 ("skbuff: increase verbosity when dumping skb data")
Signed-off-by: Gal Pressman <gal@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

dsa: mv88e6xxx: fix debug print for SPEED_UNFORCED

[ Upstream commit e08cdf63049b711099efff0811273449083bb958 ]

Debug print uses invalid check to detect if speed is unforced:
(speed != SPEED_UNFORCED) should be used instead of (!speed).

Found by Linux Verification Center (linuxtesting.org) with SVACE.

Signed-off-by: Andrey Eremeev <Axtone4all@yandex.ru>
Fixes: 96a2b40c7bd3 ("net: dsa: mv88e6xxx: add port's MAC speed setter")
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

sfc_ef100: potential dereference of null pointer

[ Upstream commit 407ecd1bd726f240123f704620d46e285ff30dd9 ]

The return value of kmalloc() needs to be checked.
To avoid use in efx_nic_update_stats() in case of the failure of alloc.

Fixes: b593b6f1b492 ("sfc_ef100: statistics gathering")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: stmmac: dwmac-rk: fix oob read in rk_gmac_setup

[ Upstream commit 0546b224cc7717cc8a2db076b0bb069a9c430794 ]

KASAN reports an out-of-bounds read in rk_gmac_setup on the line:

while (ops->regs[i]) {

This happens for most platforms since the regs flexible array member is
empty, so the memory after the ops structure is being read here.  It
seems that mostly this happens to contain zero anyway, so we get lucky
and everything still works.

To avoid adding redundant data to nearly all the ops structures, add a
new flag to indicate whether the regs field is valid and avoid this loop
when it is not.

Fixes: 3bb3d6b1c195 ("net: stmmac: Add RK3566/RK3568 SoC support")
Signed-off-by: John Keeping <john@metanate.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net/packet: rx_owner_map depends on pg_vec

[ Upstream commit ec6af094ea28f0f2dda1a6a33b14cd57e36a9755 ]

Packet sockets may switch ring versions. Avoid misinterpreting state
between versions, whose fields share a union. rx_owner_map is only
allocated with a packet ring (pg_vec) and both are swapped together.
If pg_vec is NULL, meaning no packet ring was allocated, then neither
was rx_owner_map. And the field may be old state from a tpacket_v3.

Fixes: 61fad6816fc1 ("net/packet: tpacket_rcv: avoid a producer race condition")
Reported-by: Syzbot <syzbot+1ac0994a0a0c55151121@syzkaller.appspotmail.com>
Signed-off-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20211215143937.106178-1-willemdebruijn.kernel@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

netdevsim: Zero-initialize memory for new map's value in function nsim_bpf_map_alloc

[ Upstream commit 481221775d53d6215a6e5e9ce1cce6d2b4ab9a46 ]

Zero-initialize memory for new map's value in function nsim_bpf_map_alloc
since it may cause a potential kernel information leak issue, as follows:
1. nsim_bpf_map_alloc calls nsim_map_alloc_elem to allocate elements for
a new map.
2. nsim_map_alloc_elem uses kmalloc to allocate map's value, but doesn't
zero it.
3. A user application can use IOCTL BPF_MAP_LOOKUP_ELEM to get specific
element's information in the map.
4. The kernel function map_lookup_elem will call bpf_map_copy_value to get
the information allocated at step-2, then use copy_to_user to copy to the
user buffer.
This can only leak information for an array map.

Fixes: 395cacb5f1a0 ("netdevsim: bpf: support fake map offload")
Suggested-by: Jakub Kicinski <kuba@kernel.org>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Haimin Zhang <tcs.kernel@gmail.com>
Link: https://lore.kernel.org/r/20211215111530.72103-1-tcs.kernel@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ixgbe: set X550 MDIO speed before talking to PHY

[ Upstream commit bf0a375055bd1afbbf02a0ef45f7655da7b71317 ]

The MDIO bus speed must be initialized before talking to the PHY the first
time in order to avoid talking to it using a speed that the PHY doesn't
support.

This fixes HW initialization error -17 (IXGBE_ERR_PHY_ADDR_INVALID) on
Denverton CPUs (a.k.a. the Atom C3000 family) on ports with a 10Gb network
plugged in. On those devices, HLREG0[MDCSPD] resets to 1, which combined
with the 10Gb network results in a 24MHz MDIO speed, which is apparently
too fast for the connected PHY. PHY register reads over MDIO bus return
garbage, leading to initialization failure.

Reproduced with Linux kernel 4.19 and 5.15-rc7. Can be reproduced using
the following setup:

* Use an Atom C3000 family system with at least one X552 LAN on the SoC
* Disable PXE or other BIOS network initialization if possible
  (the interface must not be initialized before Linux boots)
* Connect a live 10Gb Ethernet cable to an X550 port
* Power cycle (not reset, doesn't always work) the system and boot Linux
* Observe: ixgbe interfaces w/ 10GbE cables plugged in fail with error -17

Fixes: e84db7272798 ("ixgbe: Introduce function to control MDIO speed")
Signed-off-by: Cyril Novikov <cnovikov@lynx.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ixgbe: Document how to enable NBASE-T support

[ Upstream commit 271225fd57c2f1e0b3f8826df51be6c634affefe ]

Commit a296d665eae1 ("ixgbe: Add ethtool support to enable 2.5 and 5.0
Gbps support") introduced suppression of the advertisement of NBASE-T
speeds by default, according to Todd Fujinaka to accommodate customers
with network switches which could not cope with advertised NBASE-T
speeds, as posted in the E1000-devel mailing list:

https://sourceforge.net/p/e1000/mailman/message/37106269/

However, the suppression was not documented at all, nor was how to
enable NBASE-T support.

Properly document the NBASE-T suppression and how to enable NBASE-T
support.

Fixes: a296d665eae1 ("ixgbe: Add ethtool support to enable 2.5 and 5.0 Gbps support")
Reported-by: Robert Schlabbach <robert_s@gmx.net>
Signed-off-by: Robert Schlabbach <robert_s@gmx.net>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

igc: Fix typo in i225 LTR functions

[ Upstream commit 0182d1f3fa640888a2ed7e3f6df2fdb10adee7c8 ]

The LTR maximum value was incorrectly written using the scale from
the LTR minimum value. This would cause incorrect values to be sent,
in cases where the initial calculation lead to different min/max scales.

Fixes: 707abf069548 ("igc: Add initial LTR support")
Suggested-by: Dima Ruinskiy <dima.ruinskiy@intel.com>
Signed-off-by: Sasha Neftin <sasha.neftin@intel.com>
Tested-by: Nechama Kraus <nechamax.kraus@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

igbvf: fix double free in `igbvf_probe`

[ Upstream commit b6d335a60dc624c0d279333b22c737faa765b028 ]

In `igbvf_probe`, if register_netdev() fails, the program will go to
label err_hw_init, and then to label err_ioremap. In free_netdev() which
is just below label err_ioremap, there is `list_for_each_entry_safe` and
`netif_napi_del` which aims to delete all entries in `dev->napi_list`.
The program has added an entry `adapter->rx_ring->napi` which is added by
`netif_napi_add` in igbvf_alloc_queues(). However, adapter->rx_ring has
been freed below label err_hw_init. So this a UAF.

In terms of how to patch the problem, we can refer to igbvf_remove() and
delete the entry before `adapter->rx_ring`.

The KASAN logs are as follows:

[   35.126075] BUG: KASAN: use-after-free in free_netdev+0x1fd/0x450
[   35.127170] Read of size 8 at addr ffff88810126d990 by task modprobe/366
[   35.128360]
[   35.128643] CPU: 1 PID: 366 Comm: modprobe Not tainted 5.15.0-rc2+ #14
[   35.129789] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.12.0-59-gc9ba5276e321-prebuilt.qemu.org 04/01/2014
[   35.131749] Call Trace:
[   35.132199]  dump_stack_lvl+0x59/0x7b
[   35.132865]  print_address_description+0x7c/0x3b0
[   35.133707]  ? free_netdev+0x1fd/0x450
[   35.134378]  __kasan_report+0x160/0x1c0
[   35.135063]  ? free_netdev+0x1fd/0x450
[   35.135738]  kasan_report+0x4b/0x70
[   35.136367]  free_netdev+0x1fd/0x450
[   35.137006]  igbvf_probe+0x121d/0x1a10 [igbvf]
[   35.137808]  ? igbvf_vlan_rx_add_vid+0x100/0x100 [igbvf]
[   35.138751]  local_pci_probe+0x13c/0x1f0
[   35.139461]  pci_device_probe+0x37e/0x6c0
[   35.165526]
[   35.165806] Allocated by task 366:
[   35.166414]  ____kasan_kmalloc+0xc4/0xf0
[   35.167117]  foo_kmem_cache_alloc_trace+0x3c/0x50 [igbvf]
[   35.168078]  igbvf_probe+0x9c5/0x1a10 [igbvf]
[   35.168866]  local_pci_probe+0x13c/0x1f0
[   35.169565]  pci_device_probe+0x37e/0x6c0
[   35.179713]
[   35.179993] Freed by task 366:
[   35.180539]  kasan_set_track+0x4c/0x80
[   35.181211]  kasan_set_free_info+0x1f/0x40
[   35.181942]  ____kasan_slab_free+0x103/0x140
[   35.182703]  kfree+0xe3/0x250
[   35.183239]  igbvf_probe+0x1173/0x1a10 [igbvf]
[   35.184040]  local_pci_probe+0x13c/0x1f0

Fixes: d4e0fe01a38a0 (igbvf: add new driver to support 82576 virtual functions)
Reported-by: Zheyu Ma <zheyuma97@gmail.com>
Signed-off-by: Letu Ren <fantasquex@gmail.com>
Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

igb: Fix removal of unicast MAC filters of VFs

[ Upstream commit 584af82154f56e6b2740160fcc84a2966d969e15 ]

Move checking condition of VF MAC filter before clearing
or adding MAC filter to VF to prevent potential blackout caused
by removal of necessary and working VF's MAC filter.

Fixes: 1b8b062a99dc ("igb: add VF trust infrastructure")
Signed-off-by: Karen Sornek <karen.sornek@intel.com>
Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

soc/tegra: fuse: Fix bitwise vs. logical OR warning

[ Upstream commit a7083763619f7485ccdade160deb81737cf2732f ]

A new warning in clang points out two instances where boolean
expressions are being used with a bitwise OR instead of logical OR:

drivers/soc/tegra/fuse/speedo-tegra20.c:72:9: warning: use of bitwise '|' with boolean operands [-Wbitwise-instead-of-logical]
                reg = tegra_fuse_read_spare(i) |
                      ^~~~~~~~~~~~~~~~~~~~~~~~~~
                                               ||
drivers/soc/tegra/fuse/speedo-tegra20.c:72:9: note: cast one or both operands to int to silence this warning
drivers/soc/tegra/fuse/speedo-tegra20.c:87:9: warning: use of bitwise '|' with boolean operands [-Wbitwise-instead-of-logical]
                reg = tegra_fuse_read_spare(i) |
                      ^~~~~~~~~~~~~~~~~~~~~~~~~~
                                               ||
drivers/soc/tegra/fuse/speedo-tegra20.c:87:9: note: cast one or both operands to int to silence this warning
2 warnings generated.

The motivation for the warning is that logical operations short circuit
while bitwise operations do not.

In this instance, tegra_fuse_read_spare() is not semantically returning
a boolean, it is returning a bit value. Use u32 for its return type so
that it can be used with either bitwise or boolean operators without any
warnings.

Fixes: 25cd5a391478 ("ARM: tegra: Add speedo-based process identification")
Link: https://github.com/ClangBuiltLinux/linux/issues/1488
Suggested-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

mptcp: fix deadlock in __mptcp_push_pending()

[ Upstream commit 3d79e3756ca90f7a6087b77b62c1d9c0801e0820 ]

__mptcp_push_pending() may call mptcp_flush_join_list() with subflow
socket lock held. If such call hits mptcp_sockopt_sync_all() then
subsequently __mptcp_sockopt_sync() could try to lock the subflow
socket for itself, causing a deadlock.

sysrq: Show Blocked State
task:ss-server       state:D stack:    0 pid:  938 ppid:     1 flags:0x00000000
Call Trace:
 <TASK>
 __schedule+0x2d6/0x10c0
 ? __mod_memcg_state+0x4d/0x70
 ? csum_partial+0xd/0x20
 ? _raw_spin_lock_irqsave+0x26/0x50
 schedule+0x4e/0xc0
 __lock_sock+0x69/0x90
 ? do_wait_intr_irq+0xa0/0xa0
 __lock_sock_fast+0x35/0x50
 mptcp_sockopt_sync_all+0x38/0xc0
 __mptcp_push_pending+0x105/0x200
 mptcp_sendmsg+0x466/0x490
 sock_sendmsg+0x57/0x60
 __sys_sendto+0xf0/0x160
 ? do_wait_intr_irq+0xa0/0xa0
 ? fpregs_restore_userregs+0x12/0xd0
 __x64_sys_sendto+0x20/0x30
 do_syscall_64+0x38/0x90
 entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x7f9ba546c2d0
RSP: 002b:00007ffdc3b762d8 EFLAGS: 00000246 ORIG_RAX: 000000000000002c
RAX: ffffffffffffffda RBX: 00007f9ba56c8060 RCX: 00007f9ba546c2d0
RDX: 000000000000077a RSI: 0000000000e5e180 RDI: 0000000000000234
RBP: 0000000000cc57f0 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 00007f9ba56c8060
R13: 0000000000b6ba60 R14: 0000000000cc7840 R15: 41d8685b1d7901b8
 </TASK>

Fix the issue by using __mptcp_flush_join_list() instead of plain
mptcp_flush_join_list() inside __mptcp_push_pending(), as suggested by
Florian. The sockopt sync will be deferred to the workqueue.

Fixes: 1b3e7ede1365 ("mptcp: setsockopt: handle SO_KEEPALIVE and SO_PRIORITY")
Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/244
Suggested-by: Florian Westphal <fw@strlen.de>
Reviewed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Maxim Galaganov <max@internet.ru>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

mptcp: clear 'kern' flag from fallback sockets

[ Upstream commit d6692b3b97bdc165d150f4c1505751a323a80717 ]

The mptcp ULP extension relies on sk->sk_sock_kern being set correctly:
It prevents setsockopt(fd, IPPROTO_TCP, TCP_ULP, "mptcp", 6); from
working for plain tcp sockets (any userspace-exposed socket).

But in case of fallback, accept() can return a plain tcp sk.
In such case, sk is still tagged as 'kernel' and setsockopt will work.

This will crash the kernel, The subflow extension has a NULL ctx->conn
mptcp socket:

BUG: KASAN: null-ptr-deref in subflow_data_ready+0x181/0x2b0
Call Trace:
 tcp_data_ready+0xf8/0x370
 [..]

Fixes: cf7da0d66cc1 ("mptcp: Create SUBFLOW socket for incoming connections")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

mptcp: remove tcp ulp setsockopt support

[ Upstream commit 404cd9a22150f24acf23a8df2ad0c094ba379f57 ]

TCP_ULP setsockopt cannot be used for mptcp because its already
used internally to plumb subflow (tcp) sockets to the mptcp layer.

syzbot managed to trigger a crash for mptcp connections that are
in fallback mode:

KASAN: null-ptr-deref in range [0x0000000000000020-0x0000000000000027]
CPU: 1 PID: 1083 Comm: syz-executor.3 Not tainted 5.16.0-rc2-syzkaller #0
RIP: 0010:tls_build_proto net/tls/tls_main.c:776 [inline]
[..]
 __tcp_set_ulp net/ipv4/tcp_ulp.c:139 [inline]
 tcp_set_ulp+0x428/0x4c0 net/ipv4/tcp_ulp.c:160
 do_tcp_setsockopt+0x455/0x37c0 net/ipv4/tcp.c:3391
 mptcp_setsockopt+0x1b47/0x2400 net/mptcp/sockopt.c:638

Remove support for TCP_ULP setsockopt.

Fixes: d9e4c1291810 ("mptcp: only admit explicitly supported sockopt")
Reported-by: syzbot+1fd9b69cde42967d1add@syzkaller.appspotmail.com
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amd/pm: fix a potential gpu_metrics_table memory leak

[ Upstream commit aa464957f7e660abd554f2546a588f6533720e21 ]

Memory is allocated for gpu_metrics_table in renoir_init_smc_tables(),
but not freed in int smu_v12_0_fini_smc_tables(). Free it!

Fixes: 95868b85764a ("drm/amd/powerplay: add Renoir support for gpu metrics export")
Signed-off-by: Lang Yu <lang.yu@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/amd/display: Set exit_optimized_pwr_state for DCN31

[ Upstream commit 7e4d2f30df3fb48f75ce9e96867d42bdddab83ac ]

[Why]
SMU now respects the PHY refclk disable request from driver.

This causes a hang during hotplug when PHY refclk was disabled
because it's not being re-enabled and the transmitter control
starts on dc_link_detect.

[How]
We normally would re-enable the clk with exit_optimized_pwr_state
but this is only set on DCN21 and DCN301. Set it for dcn31 as well.

This fixes DMCUB timeouts in the PHY.

Fixes: 64b1d0e8d500 ("drm/amd/display: Add DCN3.1 HWSEQ")
Reviewed-by: Eric Yang <Eric.Yang2@amd.com>
Acked-by: Pavle Kotarac <Pavle.Kotarac@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ice: Don't put stale timestamps in the skb

[ Upstream commit 37e738b6fdb14529534dca441e0222313688fde3 ]

The driver has to check if it does not accidentally put the timestamp in
the SKB before previous timestamp gets overwritten.
Timestamp values in the PHY are read only and do not get cleared except
at hardware reset or when a new timestamp value is captured.
The cached_tstamp field is used to detect the case where a new timestamp
has not yet been captured, ensuring that we avoid sending stale
timestamp data to the stack.

Fixes: ea9b847cda64 ("ice: enable transmit timestamps for E810 devices")
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Tested-by: Gurucharan G <gurucharanx.g@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ice: Use div64_u64 instead of div_u64 in adjfine

[ Upstream commit 0013881c1145d36bf26165bb70fdd7560a5507a3 ]

Change the division in ice_ptp_adjfine from div_u64 to div64_u64.
div_u64 is used when the divisor is 32 bit but in this case incval is
64 bit and it caused incorrect calculations and incval adjustments.

Fixes: 06c16d89d2cb ("ice: register 1588 PTP clock device object for E810 devices")
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Tested-by: Gurucharan G <gurucharanx.g@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

rds: memory leak in __rds_conn_create()

[ Upstream commit 5f9562ebe710c307adc5f666bf1a2162ee7977c0 ]

__rds_conn_create() did not release conn->c_path when loop_trans != 0 and
trans->t_prefer_loopback != 0 and is_outgoing == 0.

Fixes: aced3ce57cd3 ("RDS tcp loopback connection can hang")
Signed-off-by: Hangyu Hua <hbh25y@gmail.com>
Reviewed-by: Sharath Srinivasan <sharath.srinivasan@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

flow_offload: return EOPNOTSUPP for the unsupported mpls action type

[ Upstream commit 166b6a46b78bf8b9559a6620c3032f9fe492e082 ]

We need to return EOPNOTSUPP for the unsupported mpls action type when
setup the flow action.

In the original implement, we will return 0 for the unsupported mpls
action type, actually we do not setup it and the following actions
to the flow action entry.

Fixes: 9838b20a7fb2 ("net: sched: take rtnl lock in tc_setup_flow_action()")
Signed-off-by: Baowen Zheng <baowen.zheng@corigine.com>
Signed-off-by: Simon Horman <simon.horman@corigine.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: stmmac: fix tc flower deletion for VLAN priority Rx steering

[ Upstream commit aeb7c75cb77478fdbf821628e9c95c4baa9adc63 ]

To replicate the issue:-

1) Add 1 flower filter for VLAN Priority based frame steering:-
$ IFDEVNAME=eth0
$ tc qdisc add dev $IFDEVNAME ingress
$ tc qdisc add dev $IFDEVNAME root mqprio num_tc 8 \
   map 0 1 2 3 4 5 6 7 0 0 0 0 0 0 0 0 \
   queues 1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7 hw 0
$ tc filter add dev $IFDEVNAME parent ffff: protocol 802.1Q \
   flower vlan_prio 0 hw_tc 0

2) Get the 'pref' id
$ tc filter show dev $IFDEVNAME ingress

3) Delete a specific tc flower record (say pref 49151)
$ tc filter del dev $IFDEVNAME parent ffff: pref 49151

From dmesg, we will observe kernel NULL pointer ooops

[  197.170464] BUG: kernel NULL pointer dereference, address: 0000000000000000
[  197.171367] #PF: supervisor read access in kernel mode
[  197.171367] #PF: error_code(0x0000) - not-present page
[  197.171367] PGD 0 P4D 0
[  197.171367] Oops: 0000 [#1] PREEMPT SMP NOPTI

<snip>

[  197.171367] RIP: 0010:tc_setup_cls+0x20b/0x4a0 [stmmac]

<snip>

[  197.171367] Call Trace:
[  197.171367]  <TASK>
[  197.171367]  ? __stmmac_disable_all_queues+0xa8/0xe0 [stmmac]
[  197.171367]  stmmac_setup_tc_block_cb+0x70/0x110 [stmmac]
[  197.171367]  tc_setup_cb_destroy+0xb3/0x180
[  197.171367]  fl_hw_destroy_filter+0x94/0xc0 [cls_flower]

The above issue is due to previous incorrect implementation of
tc_del_vlan_flow(), shown below, that uses flow_cls_offload_flow_rule()
to get struct flow_rule *rule which is no longer valid for tc filter
delete operation.

  struct flow_rule *rule = flow_cls_offload_flow_rule(cls);
  struct flow_dissector *dissector = rule->match.dissector;

So, to ensure tc_del_vlan_flow() deletes the right VLAN cls record for
earlier configured RX queue (configured by hw_tc) in tc_add_vlan_flow(),
this patch introduces stmmac_rfs_entry as driver-side flow_cls_offload
record for 'RX frame steering' tc flower, currently used for VLAN
priority. The implementation has taken consideration for future extension
to include other type RX frame steering such as EtherType based.

v2:
 - Clean up overly extensive backtrace and rewrite git message to better
   explain the kernel NULL pointer issue.

Fixes: 0e039f5cf86c ("net: stmmac: add RX frame steering based on VLAN priority in tc flower")
Tested-by: Kurt Kanzenbach <kurt@linutronix.de>
Signed-off-by: Ong Boon Leong <boon.leong.ong@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

mac80211: fix lookup when adding AddBA extension element

[ Upstream commit 511ab0c1dfb260a6b17b8771109e8d63474473a7 ]

We should be doing the HE capabilities lookup based on the full
interface type so if P2P doesn't have HE but client has it doesn't
get confused. Fix that.

Fixes: 2ab45876756f ("mac80211: add support for the ADDBA extension element")
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Link: https://lore.kernel.org/r/iwlwifi.20211129152938.010fc1d61137.If3a468145f29d670cb00a693bed559d8290ba693@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

cfg80211: Acquire wiphy mutex on regulatory work

[ Upstream commit e08ebd6d7b90ae81f21425ca39136f5b2272580f ]

The function cfg80211_reg_can_beacon_relax() expects wiphy
mutex to be held when it is being called. However, when
reg_leave_invalid_chans() is called the mutex is not held.
Fix it by acquiring the lock before calling the function.

Fixes: a05829a7222e ("cfg80211: avoid holding the RTNL when calling the driver")
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Link: https://lore.kernel.org/r/iwlwifi.20211202152831.527686cda037.I40ad9372a47cbad53b4aae7b5a6ccc0dc3fddf8b@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

mac80211: agg-tx: don't schedule_and_wake_txq() under sta->lock

[ Upstream commit 06c41bda0ea14aa7fba932a9613c4ee239682cf0 ]

When we call ieee80211_agg_start_txq(), that will in turn call
schedule_and_wake_txq(). Called from ieee80211_stop_tx_ba_cb()
this is done under sta->lock, which leads to certain circular
lock dependencies, as reported by Chris Murphy:
https://lore.kernel.org/r/CAJCQCtSXJ5qA4bqSPY=oLRMbv-irihVvP7A2uGutEbXQVkoNaw@mail.gmail.com

In general, ieee80211_agg_start_txq() is usually not called
with sta->lock held, only in this one place. But it's always
called with sta->ampdu_mlme.mtx held, and that's therefore
clearly sufficient.

Change ieee80211_stop_tx_ba_cb() to also call it without the
sta->lock held, by factoring it out of ieee80211_remove_tid_tx()
(which is only called in this one place).

This breaks the locking chain and makes it less likely that
we'll have similar locking chain problems in the future.

Fixes: ba8c3d6f16a1 ("mac80211: add an intermediate software queue implementation")
Reported-by: Chris Murphy <lists@colorremedies.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Link: https://lore.kernel.org/r/iwlwifi.20211202152554.f519884c8784.I555fef8e67d93fff3d9a304886c4a9f8b322e591@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/i915/display: Fix an unsigned subtraction which can never be negative.

[ Upstream commit 53b3495273282aa844c4613d19c3b30558c70c84 ]

smatch warning:
drivers/gpu/drm/i915/display/intel_dmc.c:601 parse_dmc_fw() warn:
unsigned 'fw->size - offset' is never less than zero

Firmware size is size_t and offset is u32. So the subtraction is
unsigned which can never be less than zero.

Fixes: 3d5928a168a9 ("drm/i915/xelpd: Pipe A DMC plugging")
Signed-off-by: Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211210044129.12422-1-harshit.m.mogalapalli@oracle.com
(cherry picked from commit 87bb2a410dcfb617b88e4695edf4beb6336dc314)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm/ast: potential dereference of null pointer

[ Upstream commit fea3fdf975dd9f3e5248afaab8fe023db313f005 ]

The return value of kzalloc() needs to be checked.
To avoid use of null pointer '&ast_state->base' in case of the
failure of alloc.

Fixes: f0adbc382b8b ("drm/ast: Allocate initial CRTC state of the correct size")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20211214014126.2211535-1-jiasheng@iscas.ac.cn
Signed-off-by: Sasha Levin <sashal@kernel.org>

mptcp: never allow the PM to close a listener subflow

[ Upstream commit b0cdc5dbcf2ba0d99785da5aabf1b17943805b8a ]

Currently, when deleting an endpoint the netlink PM treverses
all the local MPTCP sockets, regardless of their status.

If an MPTCP listener socket is bound to the IP matching the
delete endpoint, the listener TCP socket will be closed.
That is unexpected, the PM should only affect data subflows.

Additionally, syzbot was able to trigger a NULL ptr dereference
due to the above:

general protection fault, probably for non-canonical address 0xdffffc0000000003: 0000 [#1] PREEMPT SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000018-0x000000000000001f]
CPU: 1 PID: 6550 Comm: syz-executor122 Not tainted 5.16.0-rc4-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:__lock_acquire+0xd7d/0x54a0 kernel/locking/lockdep.c:4897
Code: 0f 0e 41 be 01 00 00 00 0f 86 c8 00 00 00 89 05 69 cc 0f 0e e9 bd 00 00 00 48 b8 00 00 00 00 00 fc ff df 48 89 da 48 c1 ea 03 <80> 3c 02 00 0f 85 f3 2f 00 00 48 81 3b 20 75 17 8f 0f 84 52 f3 ff
RSP: 0018:ffffc90001f2f818 EFLAGS: 00010016
RAX: dffffc0000000000 RBX: 0000000000000018 RCX: 0000000000000000
RDX: 0000000000000003 RSI: 0000000000000000 RDI: 0000000000000001
RBP: 0000000000000000 R08: 0000000000000001 R09: 0000000000000001
R10: 0000000000000000 R11: 000000000000000a R12: 0000000000000000
R13: ffff88801b98d700 R14: 0000000000000000 R15: 0000000000000001
FS:  00007f177cd3d700(0000) GS:ffff8880b9d00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f177cd1b268 CR3: 000000001dd55000 CR4: 0000000000350ee0
Call Trace:
 <TASK>
 lock_acquire kernel/locking/lockdep.c:5637 [inline]
 lock_acquire+0x1ab/0x510 kernel/locking/lockdep.c:5602
 __raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline]
 _raw_spin_lock_irqsave+0x39/0x50 kernel/locking/spinlock.c:162
 finish_wait+0xc0/0x270 kernel/sched/wait.c:400
 inet_csk_wait_for_connect net/ipv4/inet_connection_sock.c:464 [inline]
 inet_csk_accept+0x7de/0x9d0 net/ipv4/inet_connection_sock.c:497
 mptcp_accept+0xe5/0x500 net/mptcp/protocol.c:2865
 inet_accept+0xe4/0x7b0 net/ipv4/af_inet.c:739
 mptcp_stream_accept+0x2e7/0x10e0 net/mptcp/protocol.c:3345
 do_accept+0x382/0x510 net/socket.c:1773
 __sys_accept4_file+0x7e/0xe0 net/socket.c:1816
 __sys_accept4+0xb0/0x100 net/socket.c:1846
 __do_sys_accept net/socket.c:1864 [inline]
 __se_sys_accept net/socket.c:1861 [inline]
 __x64_sys_accept+0x71/0xb0 net/socket.c:1861
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x7f177cd8b8e9
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 b1 14 00 00 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f177cd3d308 EFLAGS: 00000246 ORIG_RAX: 000000000000002b
RAX: ffffffffffffffda RBX: 00007f177ce13408 RCX: 00007f177cd8b8e9
RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000003
RBP: 00007f177ce13400 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 00007f177ce1340c
R13: 00007f177cde1004 R14: 6d705f706374706d R15: 0000000000022000
 </TASK>

Fix the issue explicitly skipping MPTCP socket in TCP_LISTEN
status.

Reported-and-tested-by: syzbot+e4d843bb96a9431e6331@syzkaller.appspotmail.com
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Fixes: 740d798e8767 ("mptcp: remove id 0 address")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Link: https://lore.kernel.org/r/ebc7594cdd420d241fb2172ddb8542ba64717657.1639238695.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftest/net/forwarding: declare NETIFS p9 p10

[ Upstream commit 71da1aec215290e249d09c44c768df859f3a3bba ]

The recent GRE selftests defined NUM_NETIFS=10. If the users copy
forwarding.config.sample to forwarding.config directly, they will get
error "Command line is not complete" when run the GRE tests, because
create_netif_veth() failed with no interface name defined.

Fix it by extending the NETIFS with p9 and p10.

Fixes: 2800f2485417 ("selftests: forwarding: Test multipath hashing on inner IP pkts for GRE tunnel")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: dsa: mv88e6xxx: Unforce speed & duplex in mac_link_down()

[ Upstream commit 9d591fc028b6bddb38c6585874f331267cbdadae ]

Commit 64d47d50be7a ("net: dsa: mv88e6xxx: configure interface settings
in mac_config") removed forcing of speed and duplex from
mv88e6xxx_mac_config(), where the link is forced down, and left it only
in mv88e6xxx_mac_link_up(), by which time link is unforced.

It seems that (at least on 88E6190) when changing cmode to 2500base-x,
if the link is not forced down, but the speed or duplex are still
forced, the forcing of new settings for speed & duplex doesn't take in
mv88e6xxx_mac_link_up().

Fix this by unforcing speed & duplex in mv88e6xxx_mac_link_down().

Fixes: 64d47d50be7a ("net: dsa: mv88e6xxx: configure interface settings in mac_config")
Signed-off-by: Marek Behún <kabel@kernel.org>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftests/net: toeplitz: fix udp option

[ Upstream commit a8d13611b4a7b1b20d17bf2b9a89a3efcabde56c ]

Tiny fix. Option -u ("use udp") does not take an argument.

It can cause the next argument to silently be ignored.

Fixes: 5ebfb4cc3048 ("selftests/net: toeplitz test")
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net/sched: sch_ets: don't remove idle classes from the round-robin list

[ Upstream commit c062f2a0b04d86c5b8c9d973bea43493eaca3d32 ]

Shuang reported that the following script:

 1) tc qdisc add dev ddd0 handle 10: parent 1: ets bands 8 strict 4 priomap 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7
 2) mausezahn ddd0  -A 10.10.10.1 -B 10.10.10.2 -c 0 -a own -b 00:c1:a0:c1:a0:00 -t udp &
 3) tc qdisc change dev ddd0 handle 10: ets bands 4 strict 2 quanta 2500 2500 priomap 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

crashes systematically when line 2) is commented:

 list_del corruption, ffff8e028404bd30->next is LIST_POISON1 (dead000000000100)
 ------------[ cut here ]------------
 kernel BUG at lib/list_debug.c:47!
 invalid opcode: 0000 [#1] PREEMPT SMP NOPTI
 CPU: 0 PID: 954 Comm: tc Not tainted 5.16.0-rc4+ #478
 Hardware name: Red Hat KVM, BIOS 1.11.1-4.module+el8.1.0+4066+0f1aadab 04/01/2014
 RIP: 0010:__list_del_entry_valid.cold.1+0x12/0x47
 Code: fe ff 0f 0b 48 89 c1 4c 89 c6 48 c7 c7 08 42 1b 87 e8 1d c5 fe ff 0f 0b 48 89 fe 48 89 c2 48 c7 c7 98 42 1b 87 e8 09 c5 fe ff <0f> 0b 48 c7 c7 48 43 1b 87 e8 fb c4 fe ff 0f 0b 48 89 f2 48 89 fe
 RSP: 0018:ffffae46807a3888 EFLAGS: 00010246
 RAX: 000000000000004e RBX: 0000000000000007 RCX: 0000000000000202
 RDX: 0000000000000000 RSI: ffffffff871ac536 RDI: 00000000ffffffff
 RBP: ffffae46807a3a10 R08: 0000000000000000 R09: c0000000ffff7fff
 R10: 0000000000000001 R11: ffffae46807a36a8 R12: ffff8e028404b800
 R13: ffff8e028404bd30 R14: dead000000000100 R15: ffff8e02fafa2400
 FS:  00007efdc92e4480(0000) GS:ffff8e02fb600000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
 CR2: 0000000000682f48 CR3: 00000001058be000 CR4: 0000000000350ef0
 Call Trace:
  <TASK>
  ets_qdisc_change+0x58b/0xa70 [sch_ets]
  tc_modify_qdisc+0x323/0x880
  rtnetlink_rcv_msg+0x169/0x4a0
  netlink_rcv_skb+0x50/0x100
  netlink_unicast+0x1a5/0x280
  netlink_sendmsg+0x257/0x4d0
  sock_sendmsg+0x5b/0x60
  ____sys_sendmsg+0x1f2/0x260
  ___sys_sendmsg+0x7c/0xc0
  __sys_sendmsg+0x57/0xa0
  do_syscall_64+0x3a/0x80
  entry_SYSCALL_64_after_hwframe+0x44/0xae
 RIP: 0033:0x7efdc8031338
 Code: 89 02 48 c7 c0 ff ff ff ff eb b5 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 8d 05 25 43 2c 00 8b 00 85 c0 75 17 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 58 c3 0f 1f 80 00 00 00 00 41 54 41 89 d4 55
 RSP: 002b:00007ffdf1ce9828 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
 RAX: ffffffffffffffda RBX: 0000000061b37a97 RCX: 00007efdc8031338
 RDX: 0000000000000000 RSI: 00007ffdf1ce9890 RDI: 0000000000000003
 RBP: 0000000000000000 R08: 0000000000000001 R09: 000000000078a940
 R10: 000000000000000c R11: 0000000000000246 R12: 0000000000000001
 R13: 0000000000688880 R14: 0000000000000000 R15: 0000000000000000
  </TASK>
 Modules linked in: sch_ets sch_tbf dummy rfkill iTCO_wdt iTCO_vendor_support intel_rapl_msr intel_rapl_common joydev pcspkr i2c_i801 virtio_balloon i2c_smbus lpc_ich ip_tables xfs libcrc32c crct10dif_pclmul crc32_pclmul crc32c_intel serio_raw ghash_clmulni_intel ahci libahci libata virtio_blk virtio_console virtio_net net_failover failover sunrpc dm_mirror dm_region_hash dm_log dm_mod [last unloaded: sch_ets]
 ---[ end trace f35878d1912655c2 ]---
 RIP: 0010:__list_del_entry_valid.cold.1+0x12/0x47
 Code: fe ff 0f 0b 48 89 c1 4c 89 c6 48 c7 c7 08 42 1b 87 e8 1d c5 fe ff 0f 0b 48 89 fe 48 89 c2 48 c7 c7 98 42 1b 87 e8 09 c5 fe ff <0f> 0b 48 c7 c7 48 43 1b 87 e8 fb c4 fe ff 0f 0b 48 89 f2 48 89 fe
 RSP: 0018:ffffae46807a3888 EFLAGS: 00010246
 RAX: 000000000000004e RBX: 0000000000000007 RCX: 0000000000000202
 RDX: 0000000000000000 RSI: ffffffff871ac536 RDI: 00000000ffffffff
 RBP: ffffae46807a3a10 R08: 0000000000000000 R09: c0000000ffff7fff
 R10: 0000000000000001 R11: ffffae46807a36a8 R12: ffff8e028404b800
 R13: ffff8e028404bd30 R14: dead000000000100 R15: ffff8e02fafa2400
 FS:  00007efdc92e4480(0000) GS:ffff8e02fb600000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
 CR2: 0000000000682f48 CR3: 00000001058be000 CR4: 0000000000350ef0
 Kernel panic - not syncing: Fatal exception in interrupt
 Kernel Offset: 0x4e00000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff)
 ---[ end Kernel panic - not syncing: Fatal exception in interrupt ]---

we can remove 'q->classes[i].alist' only if DRR class 'i' was part of the
active list. In the ETS scheduler DRR classes belong to that list only if
the queue length is greater than zero: we need to test for non-zero value
of 'q->classes[i].qdisc->q.qlen' before removing from the list, similarly
to what has been done elsewhere in the ETS code.

Fixes: de6d25924c2a ("net/sched: sch_ets: don't peek at classes beyond 'nbands'")
Reported-by: Shuang Li <shuali@redhat.com>
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

drm: simpledrm: fix wrong unit with pixel clock

[ Upstream commit 5cf06065bd1f7b94fbb80e7eeb033899f77ab5ba ]

Pixel clock has to be set in kHz.

Signed-off-by: Alejandro Concepcion-Rodriguez <asconcepcion@acoro.eu>
Fixes: 11e8f5fd223b ("drm: Add simpledrm driver")
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/6f8554ef-1305-0dda-821c-f7d2e5644a48@acoro.eu
Signed-off-by: Sasha Levin <sashal@kernel.org>

dmaengine: st_fdma: fix MODULE_ALIAS

[ Upstream commit 822c9f2b833c53fc67e8adf6f63ecc3ea24d502c ]

modprobe can't handle spaces in aliases.

Fixes: 6b4cd727eaf1 ("dmaengine: st_fdma: Add STMicroelectronics FDMA engine driver support")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Link: https://lore.kernel.org/r/20211125154441.2626214-1-hi@alyssa.is
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

dmaengine: idxd: fix missed completion on abort path

[ Upstream commit 8affd8a4b5ce356c8900cfb037674f3a4a11fbdb ]

Ming reported that with the abort path of the descriptor submission, there
can be a window where a completed descriptor can be missed to be completed
by the irq completion thread:

CPU A CPU B
Submit (successful)

Submit (fail)
irq_process_work_list() // empty

llist_abort_desc()
// remove all descs from pending list

irq_process_pending_llist() // empty
exit idxd_wq_thread() with no processing

Add opportunistic descriptor completion in the abort path in order to
remove the missed completion.

Fixes: 6b4b87f2c31a ("dmaengine: idxd: fix submission race window")
Reported-by: Ming Li <ming4.li@intel.com>
Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/163898288714.443911.16084982766671976640.stgit@djiang5-desk3.ch.intel.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftests: Fix IPv6 address bind tests

[ Upstream commit 28a2686c185e84b6aa6a4d9c9a972360eb7ca266 ]

IPv6 allows binding a socket to a device then binding to an address
not on the device (__inet6_bind -> ipv6_chk_addr with strict flag
not set). Update the bind tests to reflect legacy behavior.

Fixes: 34d0302ab861 ("selftests: Add ipv6 address bind tests to fcnal-test")
Reported-by: Li Zhijian <lizhijian@fujitsu.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftests: Fix raw socket bind tests with VRF

[ Upstream commit 0f108ae4452025fef529671998f6c7f1c4526790 ]

Commit referenced below added negative socket bind tests for VRF. The
socket binds should fail since the address to bind to is in a VRF yet
the socket is not bound to the VRF or a device within it. Update the
expected return code to check for 1 (bind failure) so the test passes
when the bind fails as expected. Add a 'show_hint' comment to explain
why the bind is expected to fail.

Fixes: 75b2b2b3db4c ("selftests: Add ipv4 address bind tests to fcnal-test")
Reported-by: Li Zhijian <lizhijian@fujitsu.com>
Signed-off-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftests: Add duplicate config only for MD5 VRF tests

[ Upstream commit 7e0147592b5c4f9e2eb8c54a7857a56d4863f74e ]

Commit referenced below added configuration in the default VRF that
duplicates a VRF to check MD5 passwords are properly used and fail
when expected. That config should not be added all the time as it
can cause tests to pass that should not (by matching on default VRF
setup when it should not). Move the duplicate setup to a function
that is only called for the MD5 tests and add a cleanup function
to remove it after the MD5 tests.

Fixes: 5cad8bce26e0 ("fcnal-test: Add TCP MD5 tests for VRF")
Signed-off-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: hns3: fix race condition in debugfs

[ Upstream commit 6dde452bceca3f2ed2b33bc46a16ff5682a03a2e ]

When multiple threads concurrently access the debugfs content, data
and pointer exceptions may occur. Therefore, mutex lock protection is
added for debugfs.

Fixes: 5e69ea7ee2a6 ("net: hns3: refactor the debugfs process")
Signed-off-by: Yufeng Mo <moyufeng@huawei.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: hns3: fix use-after-free bug in hclgevf_send_mbx_msg

[ Upstream commit 27cbf64a766e86f068ce6214f04c00ceb4db1af4 ]

Currently, the hns3_remove function firstly uninstall client instance,
and then uninstall acceletion engine device. The netdevice is freed in
client instance uninstall process, but acceletion engine device uninstall
process still use it to trace runtime information. This causes a use after
free problem.

So fixes it by check the instance register state to avoid use after free.

Fixes: d8355240cf8f ("net: hns3: add trace event support for PF/VF mailbox")
Signed-off-by: Jie Wang <wangjie125@huawei.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftests: icmp_redirect: pass xfail=0 to log_test()

[ Upstream commit 3748939bce3fc7a15ef07161826507fbe410bb7a ]

If any sub-test in this icmp_redirect.sh is failing but not expected
to fail. The script will complain:
    ./icmp_redirect.sh: line 72: [: 1: unary operator expected

This is because when the sub-test is not expected to fail, we won't
pass any value for the xfail local variable in log_test() and thus
it's empty. Fix this by passing 0 as the 4th variable to log_test()
for non-xfail cases.

v2: added fixes tag

Fixes: 0a36a75c6818 ("selftests: icmp_redirect: support expected failures")
Signed-off-by: Po-Hsu Lin <po-hsu.lin@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

netdevsim: don't overwrite read only ethtool parms

[ Upstream commit ee60e626d536da4c710b3634afe68fe7c6d69b59 ]

Ethtool ring feature has _max_pending attributes read-only.
Set only read-write attributes in nsim_set_ringparam.

This patch is useful, if netdevsim device is set-up using NetworkManager,
because NetworkManager sends 0 as MAX values, as it is pointless to
retrieve them in extra call, because they should be read-only. Then,
the device is left in incosistent state (value > MAX).

Fixes: a7fc6db099b5 ("netdevsim: support ethtool ring and coalesce settings")
Signed-off-by: Filip Pokryvka <fpokryvk@redhat.com>
Link: https://lore.kernel.org/r/20211210175032.411872-1-fpokryvk@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

inet_diag: fix kernel-infoleak for UDP sockets

[ Upstream commit 71ddeac8cd1d217744a0e060ff520e147c9328d1 ]

KMSAN reported a kernel-infoleak [1], that can exploited
by unpriv users.

After analysis it turned out UDP was not initializing
r->idiag_expires. Other users of inet_sk_diag_fill()
might make the same mistake in the future, so fix this
in inet_sk_diag_fill().

[1]
BUG: KMSAN: kernel-infoleak in instrument_copy_to_user include/linux/instrumented.h:121 [inline]
BUG: KMSAN: kernel-infoleak in copyout lib/iov_iter.c:156 [inline]
BUG: KMSAN: kernel-infoleak in _copy_to_iter+0x69d/0x25c0 lib/iov_iter.c:670
 instrument_copy_to_user include/linux/instrumented.h:121 [inline]
 copyout lib/iov_iter.c:156 [inline]
 _copy_to_iter+0x69d/0x25c0 lib/iov_iter.c:670
 copy_to_iter include/linux/uio.h:155 [inline]
 simple_copy_to_iter+0xf3/0x140 net/core/datagram.c:519
 __skb_datagram_iter+0x2cb/0x1280 net/core/datagram.c:425
 skb_copy_datagram_iter+0xdc/0x270 net/core/datagram.c:533
 skb_copy_datagram_msg include/linux/skbuff.h:3657 [inline]
 netlink_recvmsg+0x660/0x1c60 net/netlink/af_netlink.c:1974
 sock_recvmsg_nosec net/socket.c:944 [inline]
 sock_recvmsg net/socket.c:962 [inline]
 sock_read_iter+0x5a9/0x630 net/socket.c:1035
 call_read_iter include/linux/fs.h:2156 [inline]
 new_sync_read fs/read_write.c:400 [inline]
 vfs_read+0x1631/0x1980 fs/read_write.c:481
 ksys_read+0x28c/0x520 fs/read_write.c:619
 __do_sys_read fs/read_write.c:629 [inline]
 __se_sys_read fs/read_write.c:627 [inline]
 __x64_sys_read+0xdb/0x120 fs/read_write.c:627
 do_syscall_x64 arch/x86/entry/common.c:51 [inline]
 do_syscall_64+0x54/0xd0 arch/x86/entry/common.c:82
 entry_SYSCALL_64_after_hwframe+0x44/0xae

Uninit was created at:
 slab_post_alloc_hook mm/slab.h:524 [inline]
 slab_alloc_node mm/slub.c:3251 [inline]
 __kmalloc_node_track_caller+0xe0c/0x1510 mm/slub.c:4974
 kmalloc_reserve net/core/skbuff.c:354 [inline]
 __alloc_skb+0x545/0xf90 net/core/skbuff.c:426
 alloc_skb include/linux/skbuff.h:1126 [inline]
 netlink_dump+0x3d5/0x16a0 net/netlink/af_netlink.c:2245
 __netlink_dump_start+0xd1c/0xee0 net/netlink/af_netlink.c:2370
 netlink_dump_start include/linux/netlink.h:254 [inline]
 inet_diag_handler_cmd+0x2e7/0x400 net/ipv4/inet_diag.c:1343
 sock_diag_rcv_msg+0x24a/0x620
 netlink_rcv_skb+0x447/0x800 net/netlink/af_netlink.c:2491
 sock_diag_rcv+0x63/0x80 net/core/sock_diag.c:276
 netlink_unicast_kernel net/netlink/af_netlink.c:1319 [inline]
 netlink_unicast+0x1095/0x1360 net/netlink/af_netlink.c:1345
 netlink_sendmsg+0x16f3/0x1870 net/netlink/af_netlink.c:1916
 sock_sendmsg_nosec net/socket.c:704 [inline]
 sock_sendmsg net/socket.c:724 [inline]
 sock_write_iter+0x594/0x690 net/socket.c:1057
 do_iter_readv_writev+0xa7f/0xc70
 do_iter_write+0x52c/0x1500 fs/read_write.c:851
 vfs_writev fs/read_write.c:924 [inline]
 do_writev+0x63f/0xe30 fs/read_write.c:967
 __do_sys_writev fs/read_write.c:1040 [inline]
 __se_sys_writev fs/read_write.c:1037 [inline]
 __x64_sys_writev+0xe5/0x120 fs/read_write.c:1037
 do_syscall_x64 arch/x86/entry/common.c:51 [inline]
 do_syscall_64+0x54/0xd0 arch/x86/entry/common.c:82
 entry_SYSCALL_64_after_hwframe+0x44/0xae

Bytes 68-71 of 312 are uninitialized
Memory access of size 312 starts at ffff88812ab54000
Data copied to user address 0000000020001440

CPU: 1 PID: 6365 Comm: syz-executor801 Not tainted 5.16.0-rc3-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011

Fixes: 3c4d05c80567 ("inet_diag: Introduce the inet socket dumping routine")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Link: https://lore.kernel.org/r/20211209185058.53917-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

sch_cake: do not call cake_destroy() from cake_init()

[ Upstream commit ab443c53916730862cec202078d36fd4008bea79 ]

qdiscs are not supposed to call their own destroy() method
from init(), because core stack already does that.

syzbot was able to trigger use after free:

DEBUG_LOCKS_WARN_ON(lock->magic != lock)
WARNING: CPU: 0 PID: 21902 at kernel/locking/mutex.c:586 __mutex_lock_common kernel/locking/mutex.c:586 [inline]
WARNING: CPU: 0 PID: 21902 at kernel/locking/mutex.c:586 __mutex_lock+0x9ec/0x12f0 kernel/locking/mutex.c:740
Modules linked in:
CPU: 0 PID: 21902 Comm: syz-executor189 Not tainted 5.16.0-rc4-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:__mutex_lock_common kernel/locking/mutex.c:586 [inline]
RIP: 0010:__mutex_lock+0x9ec/0x12f0 kernel/locking/mutex.c:740
Code: 08 84 d2 0f 85 19 08 00 00 8b 05 97 38 4b 04 85 c0 0f 85 27 f7 ff ff 48 c7 c6 20 00 ac 89 48 c7 c7 a0 fe ab 89 e8 bf 76 ba ff <0f> 0b e9 0d f7 ff ff 48 8b 44 24 40 48 8d b8 c8 08 00 00 48 89 f8
RSP: 0018:ffffc9000627f290 EFLAGS: 00010282
RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
RDX: ffff88802315d700 RSI: ffffffff815f1db8 RDI: fffff52000c4fe44
RBP: ffff88818f28e000 R08: 0000000000000000 R09: 0000000000000000
R10: ffffffff815ebb5e R11: 0000000000000000 R12: 0000000000000000
R13: dffffc0000000000 R14: ffffc9000627f458 R15: 0000000093c30000
FS:  0000555556abc400(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007fda689c3303 CR3: 000000001cfbb000 CR4: 0000000000350ef0
Call Trace:
 <TASK>
 tcf_chain0_head_change_cb_del+0x2e/0x3d0 net/sched/cls_api.c:810
 tcf_block_put_ext net/sched/cls_api.c:1381 [inline]
 tcf_block_put_ext net/sched/cls_api.c:1376 [inline]
 tcf_block_put+0xbc/0x130 net/sched/cls_api.c:1394
 cake_destroy+0x3f/0x80 net/sched/sch_cake.c:2695
 qdisc_create.constprop.0+0x9da/0x10f0 net/sched/sch_api.c:1293
 tc_modify_qdisc+0x4c5/0x1980 net/sched/sch_api.c:1660
 rtnetlink_rcv_msg+0x413/0xb80 net/core/rtnetlink.c:5571
 netlink_rcv_skb+0x153/0x420 net/netlink/af_netlink.c:2496
 netlink_unicast_kernel net/netlink/af_netlink.c:1319 [inline]
 netlink_unicast+0x533/0x7d0 net/netlink/af_netlink.c:1345
 netlink_sendmsg+0x904/0xdf0 net/netlink/af_netlink.c:1921
 sock_sendmsg_nosec net/socket.c:704 [inline]
 sock_sendmsg+0xcf/0x120 net/socket.c:724
 ____sys_sendmsg+0x6e8/0x810 net/socket.c:2409
 ___sys_sendmsg+0xf3/0x170 net/socket.c:2463
 __sys_sendmsg+0xe5/0x1b0 net/socket.c:2492
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x7f1bb06badb9
Code: Unable to access opcode bytes at RIP 0x7f1bb06bad8f.
RSP: 002b:00007fff3012a658 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
RAX: ffffffffffffffda RBX: 0000000000000003 RCX: 00007f1bb06badb9
RDX: 0000000000000000 RSI: 00000000200007c0 RDI: 0000000000000003
RBP: 0000000000000000 R08: 0000000000000003 R09: 0000000000000003
R10: 0000000000000003 R11: 0000000000000246 R12: 00007fff3012a688
R13: 00007fff3012a6a0 R14: 00007fff3012a6e0 R15: 00000000000013c2
 </TASK>

Fixes: 046f6fd5daef ("sched: Add Common Applications Kept Enhanced (cake) qdisc")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Acked-by: Toke Høiland-Jørgensen <toke@toke.dk>
Link: https://lore.kernel.org/r/20211210142046.698336-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

s390/kexec_file: fix error handling when applying relocations

[ Upstream commit 41967a37b8eedfee15b81406a9f3015be90d3980 ]

arch_kexec_apply_relocations_add currently ignores all errors returned
by arch_kexec_do_relocs. This means that every unknown relocation is
silently skipped causing unpredictable behavior while the relocated code
runs. Fix this by checking for errors and fail kexec_file_load if an
unknown relocation type is encountered.

The problem was found after gcc changed its behavior and used
R_390_PLT32DBL relocations for brasl instruction and relied on ld to
resolve the relocations in the final link in case direct calls are
possible. As the purgatory code is only linked partially (option -r)
ld didn't resolve the relocations leaving them for arch_kexec_do_relocs.
But arch_kexec_do_relocs doesn't know how to handle R_390_PLT32DBL
relocations so they were silently skipped. This ultimately caused an
endless loop in the purgatory as the brasl instructions kept branching
to itself.

Fixes: 71406883fd35 ("s390/kexec_file: Add kexec_file_load system call")
Reported-by: Tao Liu <ltao@redhat.com>
Signed-off-by: Philipp Rudo <prudo@redhat.com>
Link: https://lore.kernel.org/r/20211208130741.5821-3-prudo@redhat.com
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftests: net: Correct ping6 expected rc from 2 to 1

[ Upstream commit 92816e2629808726af015c7f5b14adc8e4f8b147 ]

./fcnal-test.sh -v -t ipv6_ping
TEST: ping out, VRF bind - ns-B IPv6 LLA                                      [FAIL]
TEST: ping out, VRF bind - multicast IP                                       [FAIL]

ping6 is failing as it should.
COMMAND: ip netns exec ns-A /bin/ping6 -c1 -w1 fe80::7c4c:bcff:fe66:a63a%red
strace of ping6 shows it is failing with '1',
so change the expected rc from 2 to 1.

Fixes: c0644e71df33 ("selftests: Add ipv6 ping tests to fcnal-test")
Reported-by: kernel test robot <lkp@intel.com>
Suggested-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Jie2x Zhou <jie2x.zhou@intel.com>
Link: https://lore.kernel.org/r/20211209020230.37270-1-jie2x.zhou@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

Revert "drm/fb-helper: improve DRM fbdev emulation device names"

[ Upstream commit 842470c4e211f284a224842849b1fa81b130c154 ]

This reverts commit b3484d2b03e4c940a9598aa841a52d69729c582a.

That change attempted to improve the DRM drivers fbdev emulation device
names to avoid having confusing names like "simpledrmdrmfb" in /proc/fb.

But unfortunately, there are user-space programs such as pm-utils that
match against the fbdev names and so broke after the mentioned commit.

Since the names in /proc/fb are used by tools that consider it an uAPI,
let's restore the old names even when this lead to silly names like the
one mentioned above.

Fixes: b3484d2b03e4 ("drm/fb-helper: improve DRM fbdev emulation device names")
Reported-by: Johannes Stezenbach <js@sig21.net>
Signed-off-by: Javier Martinez Canillas <javierm@redhat.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211020165740.3011927-1-javierm@redhat.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

vdpa: Consider device id larger than 31

[ Upstream commit bb47620be322c5e9e372536cb6b54e17b3a00258 ]

virtio device id value can be more than 31. Hence, use BIT_ULL in
assignment.

Fixes: 33b347503f01 ("vdpa: Define vdpa mgmt device, ops and a netlink interface")
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Parav Pandit <parav@nvidia.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Link: https://lore.kernel.org/r/20211130042949.88958-1-parav@nvidia.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

virtio/vsock: fix the transport to work with VMADDR_CID_ANY

[ Upstream commit 1db8f5fc2e5c66a5c51e1f6488e0ba7d45c29ae4 ]

The VMADDR_CID_ANY flag used by a socket means that the socket isn't bound
to any specific CID. For example, a host vsock server may want to be bound
with VMADDR_CID_ANY, so that a guest vsock client can connect to the host
server with CID=VMADDR_CID_HOST (i.e. 2), and meanwhile, a host vsock
client can connect to the same local server with CID=VMADDR_CID_LOCAL
(i.e. 1).

The current implementation sets the destination socket's svm_cid to a
fixed CID value after the first client's connection, which isn't an
expected operation. For example, if the guest client first connects to the
host server, the server's svm_cid gets set to VMADDR_CID_HOST, then other
host clients won't be able to connect to the server anymore.

Reproduce steps:
1. Run the host server:
   socat VSOCK-LISTEN:1234,fork -
2. Run a guest client to connect to the host server:
   socat - VSOCK-CONNECT:2:1234
3. Run a host client to connect to the host server:
   socat - VSOCK-CONNECT:1:1234

Without this patch, step 3. above fails to connect, and socat complains
"socat[1720] E connect(5, AF=40 cid:1 port:1234, 16): Connection
reset by peer".
With this patch, the above works well.

Fixes: c0cfa2d8a788 ("vsock: add multi-transports support")
Signed-off-by: Wei Wang <wei.w.wang@intel.com>
Link: https://lore.kernel.org/r/20211126011823.1760-1-wei.w.wang@intel.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

virtio: always enter drivers/virtio/

[ Upstream commit 27d9839f17940e8edc475df616bbd9cf7ede8d05 ]

When neither VIRTIO_PCI_LIB nor VIRTIO are enabled, but the alibaba
vdpa driver is, the kernel runs into a link error because the legacy
virtio module never gets built:

x86_64-linux-ld: drivers/vdpa/alibaba/eni_vdpa.o: in function `eni_vdpa_set_features':
eni_vdpa.c:(.text+0x23f): undefined reference to `vp_legacy_set_features'
x86_64-linux-ld: drivers/vdpa/alibaba/eni_vdpa.o: in function `eni_vdpa_set_vq_state':
eni_vdpa.c:(.text+0x2fe): undefined reference to `vp_legacy_get_queue_enable'
x86_64-linux-ld: drivers/vdpa/alibaba/eni_vdpa.o: in function `eni_vdpa_set_vq_address':
eni_vdpa.c:(.text+0x376): undefined reference to `vp_legacy_set_queue_address'
x86_64-linux-ld: drivers/vdpa/alibaba/eni_vdpa.o: in function `eni_vdpa_set_vq_ready':
eni_vdpa.c:(.text+0x3b4): undefined reference to `vp_legacy_set_queue_address'
x86_64-linux-ld: drivers/vdpa/alibaba/eni_vdpa.o: in function `eni_vdpa_free_irq':
eni_vdpa.c:(.text+0x460): undefined reference to `vp_legacy_queue_vector'
x86_64-linux-ld: eni_vdpa.c:(.text+0x4b7): undefined reference to `vp_legacy_config_vector'
x86_64-linux-ld: drivers/vdpa/alibaba/eni_vdpa.o: in function `eni_vdpa_reset':

When VIRTIO_PCI_LIB was added, it was correctly added to drivers/Makefile
as well, but for the legacy module, this is missing.  Solve this by always
entering drivers/virtio during the build and letting its Makefile take
care of the individual options, rather than having a separate line for
each sub-option.

Fixes: 64b9f64f80a6 ("vdpa: introduce virtio pci driver")
Fixes: e85087beedca ("eni_vdpa: add vDPA driver for Alibaba ENI")
Fixes: d89c8169bd70 ("virtio-pci: introduce legacy device module")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20211206085034.2836099-1-arnd@kernel.org
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

iwlwifi: mvm: don't crash on invalid rate w/o STA

[ Upstream commit d599f714b73e4177dfdfe64fce09175568288ee9 ]

If we get to the WARN_ONCE(..., "Got a HT rate (...)", ...)
here with a NULL sta, then we crash because mvmsta is bad
and we try to dereference it. Fix that by printing -1 as the
state if no station was given.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Fixes: 6761a718263a ("iwlwifi: mvm: add explicit check for non-data frames in get Tx rate")
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/iwlwifi.20211203140410.1a1541d7dcb5.I606c746e11447fe168cf046376b70b04e278c3b4@changeid
Signed-off-by: Sasha Levin <sashal@kernel.org>

soc: imx: Register SoC device only on i.MX boards

[ Upstream commit 4ebd29f91629e69da7d57390cdc953772eee03ab ]

At the moment, using the ARM32 multi_v7_defconfig always results in two
SoCs being exposed in sysfs. This is wrong, as far as I'm aware the
Qualcomm DragonBoard 410c does not actually make use of a i.MX SoC. :)

  qcom-db410c:/sys/devices/soc0$ grep . *
  family:Freescale i.MX
  machine:Qualcomm Technologies, Inc. APQ 8016 SBC
  revision:0.0
  serial_number:0000000000000000
  soc_id:Unknown

  qcom-db410c:/sys/devices/soc1$ grep . *
  family:Snapdragon
  machine:APQ8016
  ...

This happens because imx_soc_device_init() registers the soc device
unconditionally, even when running on devices that do not make use of i.MX.
Arnd already reported this more than a year ago and even suggested a fix
similar to this commit, but for some reason it was never submitted.

Fix it by checking if the "__mxc_cpu_type" variable was actually
initialized by earlier platform code. On devices without i.MX it will
simply stay 0.

Cc: Peng Fan <peng.fan@nxp.com>
Fixes: d2199b34871b ("ARM: imx: use device_initcall for imx_soc_device_init")
Reported-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/CAK8P3a0hxO1TmK6oOMQ70AHSWJnP_CAq57YMOutrxkSYNjFeuw@mail.gmail.com/
Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

clk: Don't parent clks until the parent is fully registered

[ Upstream commit 54baf56eaa40aa5cdcd02b3c20d593e4e1211220 ]

Before commit fc0c209c147f ("clk: Allow parents to be specified without
string names") child clks couldn't find their parent until the parent
clk was added to a list in __clk_core_init(). After that commit, child
clks can reference their parent clks directly via a clk_hw pointer, or
they can lookup that clk_hw pointer via DT if the parent clk is
registered with an OF clk provider.

The common clk framework treats hw->core being non-NULL as "the clk is
registered" per the logic within clk_core_fill_parent_index():

parent = entry->hw->core;
/*
 * We have a direct reference but it isn't registered yet?
 * Orphan it and let clk_reparent() update the orphan status
 * when the parent is registered.
 */
if (!parent)

Therefore we need to be extra careful to not set hw->core until the clk
is fully registered with the clk framework. Otherwise we can get into a
situation where a child finds a parent clk and we move the child clk off
the orphan list when the parent isn't actually registered, wrecking our
enable accounting and breaking critical clks.

Consider the following scenario:

  CPU0                                     CPU1
  ----                                     ----
  struct clk_hw clkBad;
  struct clk_hw clkA;

  clkA.init.parent_hws = { &clkBad };

  clk_hw_register(&clkA)                   clk_hw_register(&clkBad)
   ...                                      __clk_register()
     hw->core = core
     ...
   __clk_register()
    __clk_core_init()
     clk_prepare_lock()
     __clk_init_parent()
      clk_core_get_parent_by_index()
       clk_core_fill_parent_index()
        if (entry->hw) {
 parent = entry->hw->core;

At this point, 'parent' points to clkBad even though clkBad hasn't been
fully registered yet. Ouch! A similar problem can happen if a clk
controller registers orphan clks that are referenced in the DT node of
another clk controller.

Let's fix all this by only setting the hw->core pointer underneath the
clk prepare lock in __clk_core_init(). This way we know that
clk_core_fill_parent_index() can't see hw->core be non-NULL until the
clk is fully registered.

Fixes: fc0c209c147f ("clk: Allow parents to be specified without string names")
Signed-off-by: Mike Tipton <quic_mdtipton@quicinc.com>
Link: https://lore.kernel.org/r/20211109043438.4639-1-quic_mdtipton@quicinc.com
[sboyd@kernel.org: Reword commit text, update comment]
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: imx8mq: remove interconnect property from lcdif

[ Upstream commit e5e6268f77badf18bd6ab435364cfe21c7396c31 ]

The mxsfb driver handling imx8mq lcdif doesn't yet request the
interconnect bandwidth that's needed at runtime when the description is
present in the DT node.

So remove that description and bring it back when it's supported.

Fixes: ad1abc8a03fd ("arm64: dts: imx8mq: Add interconnect for lcdif")
Signed-off-by: Martin Kepplinger <martin.kepplinger@puri.sm>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ARM: socfpga: dts: fix qspi node compatible

[ Upstream commit cb25b11943cbcc5a34531129952870420f8be858 ]

The QSPI flash node needs to have the required "jedec,spi-nor" in the
compatible string.

Fixes: 1df99da8953 ("ARM: dts: socfpga: Enable QSPI in Arria10 devkit")
Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ceph: initialize pathlen variable in reconnect_caps_cb

[ Upstream commit ee2a095d3b24f300a5e11944d208801e928f108c ]

The smatch static checker warned about an uninitialized symbol usage in
this function, in the case where ceph_mdsc_build_path returns an error.

It turns out that that case is harmless, but it just looks sketchy.
Initialize the variable at declaration time, and remove the unneeded
setting of it later.

Fixes: a33f6432b3a6 ("ceph: encode inodes' parent/d_name in cap reconnect message")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Xiubo Li <xiubli@redhat.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ceph: fix duplicate increment of opened_inodes metric

[ Upstream commit 973e5245637accc4002843f6b888495a6a7762bc ]

opened_inodes is incremented twice when the same inode is opened twice
with O_RDONLY and O_WRONLY respectively.

To reproduce, run this python script, then check the metrics:

import os
for _ in range(10000):
    fd_r = os.open('a', os.O_RDONLY)
    fd_w = os.open('a', os.O_WRONLY)
    os.close(fd_r)
    os.close(fd_w)

Fixes: 1dd8d4708136 ("ceph: metrics for opened files, pinned caps and opened inodes")
Signed-off-by: Hu Weiwen <sehuww@mail.scut.edu.cn>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

tee: amdtee: fix an IS_ERR() vs NULL bug

[ Upstream commit 9d7482771fac8d8e38e763263f2ca0ca12dd22c6 ]

The __get_free_pages() function does not return error pointers it returns
NULL so fix this condition to avoid a NULL dereference.

Fixes: 757cc3e9ff1d ("tee: add AMD-TEE driver")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Rijo Thomas <Rijo-john.Thomas@amd.com>
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

mac80211: track only QoS data frames for admission control

[ Upstream commit d5e568c3a4ec2ddd23e7dc5ad5b0c64e4f22981a ]

For admission control, obviously all of that only works for
QoS data frames, otherwise we cannot even access the QoS
field in the header.

Syzbot reported (see below) an uninitialized value here due
to a status of a non-QoS nullfunc packet, which isn't even
long enough to contain the QoS header.

Fix this to only do anything for QoS data packets.

Reported-by: syzbot+614e82b88a1a4973e534@syzkaller.appspotmail.com
Fixes: 02219b3abca5 ("mac80211: add WMM admission control support")
Link: https://lore.kernel.org/r/20211122124737.dad29e65902a.Ieb04587afacb27c14e0de93ec1bfbefb238cc2a0@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

dmaengine: idxd: fix calling wq quiesce inside spinlock

[ Upstream commit fa51b16d05583c7aebbc06330afb50276243d198 ]

Dan reports that smatch has found idxd_wq_quiesce() is being called inside
the idxd->dev_lock. idxd_wq_quiesce() calls wait_for_completion() and
therefore it can sleep. Move the call outside of the spinlock as it does
not need device lock.

Fixes: 5b0c68c473a1 ("dmaengine: idxd: support reporting of halt interrupt")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/163716858508.1721911.15051495873516709923.stgit@djiang5-desk3.ch.intel.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

dmaengine: idxd: add halt interrupt support

[ Upstream commit 88d97ea82cbe352851a8654ee952d3a694c8c2c6 ]

Add halt interrupt support. Given that the misc interrupt handler already
check halt state, the driver just need to run the halt handling code when
receiving the halt interrupt.

Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/163114224352.846654.14334468363464318828.stgit@djiang5-desk3.ch.intel.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: rockchip: fix poweroff on helios64

[ Upstream commit aef4b9a89a376a9cabe5e744729914e7766c59bb ]

Adding the rockchip,system-power-controller property here will use the
rk808 to power off the system.

Fixes: 09e006cfb43e ("arm64: dts: rockchip: Add basic support for Kobol's Helios64")
Signed-off-by: Florian Klink <flokli@flokli.de>
Tested-by: Dennis Gilmore <dgilmore@redhat.com>
Link: https://lore.kernel.org/r/20211020095926.735938-2-flokli@flokli.de
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: rockchip: fix audio-supply for Rock Pi 4

[ Upstream commit 8240e87f16d17a9592c9d67857a3dcdbcb98f10d ]

As stated in the schematics [1] and [2] P5 the APIO5 domain is supplied
by RK808-D Buck4, which in our case vcc1v8_codec - i.e. a 1.8 V regulator.

Currently only white noise comes from the ES8316's output, which - for
whatever reason - came up only after the the correct switch from i2s0_8ch_bus
to i2s0_2ch_bus for i2s0's pinctrl was done.

Fix this by setting the correct regulator for audio-supply.

[1] https://dl.radxa.com/rockpi4/docs/hw/rockpi4/rockpi4_v13_sch_20181112.pdf
[2] https://dl.radxa.com/rockpi4/docs/hw/rockpi4/rockpi_4c_v12_sch_20200620.pdf

Fixes: 1b5715c602fd ("arm64: dts: rockchip: add ROCK Pi 4 DTS support")
Signed-off-by: Alex Bee <knaerzche@gmail.com>
Link: https://lore.kernel.org/r/20211027143726.165809-1-knaerzche@gmail.com
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: rockchip: fix rk3399-leez-p710 vcc3v3-lan supply

[ Upstream commit 2b454a90e2ccdd6e03f88f930036da4df577be76 ]

Correct a typo in the vin-supply property.  The input supply is
always-on, so this mistake doesn't affect whether the supply is actually
enabled correctly.

Fixes: fc702ed49a86 ("arm64: dts: rockchip: Add dts for Leez RK3399 P710 SBC")
Signed-off-by: John Keeping <john@metanate.com>
Link: https://lore.kernel.org/r/20211102182908.3409670-3-john@metanate.com
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: rockchip: fix rk3308-roc-cc vcc-sd supply

[ Upstream commit 772fb46109f635dd75db20c86b7eaf48efa46cef ]

Correct a typo in the vin-supply property.  The input supply is
always-on, so this mistake doesn't affect whether the supply is actually
enabled correctly.

Fixes: 4403e1237be3 ("arm64: dts: rockchip: Add devicetree for board roc-rk3308-cc")
Signed-off-by: John Keeping <john@metanate.com>
Link: https://lore.kernel.org/r/20211102182908.3409670-2-john@metanate.com
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arm64: dts: rockchip: remove mmc-hs400-enhanced-strobe from rk3399-khadas-edge

[ Upstream commit 6dd0053683804427529ef3523f7872f473440a19 ]

Remove mmc-hs400-enhanced-strobe from the rk3399-khadas-edge dts to
improve compatibility with a wider range of eMMC chips.

Before (BJTD4R 29.1 GiB):

[    7.001493] mmc2: CQHCI version 5.10
[    7.027971] mmc2: SDHCI controller on fe330000.mmc [fe330000.mmc] using ADMA
.......
[    7.207086] mmc2: mmc_select_hs400es failed, error -110
[    7.207129] mmc2: error -110 whilst initialising MMC card
[    7.308893] mmc2: mmc_select_hs400es failed, error -110
[    7.308921] mmc2: error -110 whilst initialising MMC card
[    7.427524] mmc2: mmc_select_hs400es failed, error -110
[    7.427546] mmc2: error -110 whilst initialising MMC card
[    7.590993] mmc2: mmc_select_hs400es failed, error -110
[    7.591012] mmc2: error -110 whilst initialising MMC card

After:

[    6.960785] mmc2: CQHCI version 5.10
[    6.984672] mmc2: SDHCI controller on fe330000.mmc [fe330000.mmc] using ADMA
[    7.175021] mmc2: Command Queue Engine enabled
[    7.175053] mmc2: new HS400 MMC card at address 0001
[    7.175808] mmcblk2: mmc2:0001 BJTD4R 29.1 GiB
[    7.176033] mmcblk2boot0: mmc2:0001 BJTD4R 4.00 MiB
[    7.176245] mmcblk2boot1: mmc2:0001 BJTD4R 4.00 MiB
[    7.176495] mmcblk2rpmb: mmc2:0001 BJTD4R 4.00 MiB, chardev (242:0)

Fixes: c2aacceedc86 ("arm64: dts: rockchip: Add support for Khadas Edge/Edge-V/Captain boards")
Signed-off-by: Artem Lapkin <art@khadas.com>
Link: https://lore.kernel.org/r/20211115083321.2627461-1-art@khadas.com
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>

pinctrl: amd: Fix wakeups when IRQ is shared with SCI

commit 2d54067fcd23aae61e23508425ae5b29e973573d upstream.

On some Lenovo AMD Gen2 platforms the IRQ for the SCI and pinctrl drivers
are shared.  Due to how the s2idle loop handling works, this case needs
an extra explicit check whether the interrupt was caused by SCI or by
the GPIO controller.

To fix this rework the existing IRQ handler function to function as a
checker and an IRQ handler depending on the calling arguments.

BugLink: https://gitlab.freedesktop.org/drm/amd/-/issues/1738
Reported-by: Joerie de Gram <j.de.gram@gmail.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Acked-by: Basavaraj Natikar <Basavaraj.Natikar@amd.com>
Link: https://lore.kernel.org/r/20211101014853.6177-2-mario.limonciello@amd.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

drm/i915/hdmi: Turn DP++ TMDS output buffers back on in encoder->shutdown()

commit cecbc0c7eba7983965cac94f88d2db00b913253b upstream.

Looks like our VBIOS/GOP generally fail to turn the DP dual mode adater
TMDS output buffers back on after a reboot. This leads to a black screen
after reboot if we turned the TMDS output buffers off prior to reboot.
And if i915 decides to do a fastboot the black screen will persist even
after i915 takes over.

Apparently this has been a problem ever since commit b2ccb822d376 ("drm/i915:
Enable/disable TMDS output buffers in DP++ adaptor as needed") if one
rebooted while the display was turned off. And things became worse with
commit fe0f1e3bfdfe ("drm/i915: Shut down displays gracefully on reboot")
since now we always turn the display off before a reboot.

This was reported on a RKL, but I confirmed the same behaviour on my
SNB as well. So looks pretty universal.

Let's fix this by explicitly turning the TMDS output buffers back on
in the encoder->shutdown() hook. Note that this gets called after irqs
have been disabled, so the i2c communication with the DP dual mode
adapter has to be performed via polling (which the gmbus code is
perfectly happy to do for us).

We also need a bit of care in handling DDI encoders which may or may
not be set up for HDMI output. Specifically ddc_pin will not be
populated for a DP only DDI encoder, in which case we don't want to
call intel_gmbus_get_adapter(). We can handle that by simply doing
the dual mode adapter type check before calling
intel_gmbus_get_adapter().

Cc: <stable@vger.kernel.org> # v5.11+
Fixes: fe0f1e3bfdfe ("drm/i915: Shut down displays gracefully on reboot")
Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/4371
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211029191802.18448-2-ville.syrjala@linux.intel.com
Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
(cherry picked from commit 49c55f7b035b87371a6d3c53d9af9f92ddc962db)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

drm/i915/hdmi: convert intel_hdmi_to_dev to intel_hdmi_to_i915

commit 7ceb751b615900086eed1d65955933923f127d99 upstream.

Prefer i915 over drm pointer.

Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210921110244.8666-1-jani.nikula@intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

scsi: ufs: core: Retry START_STOP on UNIT_ATTENTION

commit af21c3fd5b3ec540a97b367a70b26616ff7e0c55 upstream.

Commit 57d104c153d3 ("ufs: add UFS power management support") made the UFS
driver submit a REQUEST SENSE command before submitting a power management
command to a WLUN to clear the POWER ON unit attention. Instead of
submitting a REQUEST SENSE command before submitting a power management
command, retry the power management command until it succeeds.

This is the preparation to get rid of all UNIT ATTENTION code which should
be handled by users.

Link: https://lore.kernel.org/r/20211001182015.1347587-2-jaegeuk@kernel.org
Cc: Adrian Hunter <adrian.hunter@intel.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

btrfs: remove stale comment about the btrfs_show_devname

Commit cdccc03a8a369b59cff5e7ea3292511cfa551120 upstream.

There were few lockdep warnings because btrfs_show_devname() was using
device_list_mutex as recorded in the commits:

  0ccd05285e7f ("btrfs: fix a possible umount deadlock")
  779bf3fefa83 ("btrfs: fix lock dep warning, move scratch dev out of device_list_mutex and uuid_mutex")

And finally, commit 88c14590cdd6 ("btrfs: use RCU in btrfs_show_devname
for device list traversal") removed the device_list_mutex from
btrfs_show_devname for performance reasons.

This patch removes a stale comment about the function
btrfs_show_devname and device_list_mutex.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

btrfs: update latest_dev when we create a sprout device

Commit b7cb29e666fe79dda5dbe5f57fb7c92413bf161c upstream.

When we add a device to the seed filesystem (sprouting) it is a new
filesystem (and fsid) on the device added. Update the latest_dev so
that /proc/self/mounts shows the correct device.

Example:

  $ btrfstune -S1 /dev/vg/seed
  $ mount /dev/vg/seed /btrfs
  mount: /btrfs: WARNING: device write-protected, mounted read-only.

  $ cat /proc/self/mounts | grep btrfs
  /dev/mapper/vg-seed /btrfs btrfs ro,relatime,space_cache,subvolid=5,subvol=/ 0 0

  $ btrfs dev add -f /dev/vg/new /btrfs

Before:

  $ cat /proc/self/mounts | grep btrfs
  /dev/mapper/vg-seed /btrfs btrfs ro,relatime,space_cache,subvolid=5,subvol=/ 0 0

After:

  $ cat /proc/self/mounts | grep btrfs
  /dev/mapper/vg-new /btrfs btrfs ro,relatime,space_cache,subvolid=5,subvol=/ 0 0

Tested-by: Su Yue <l@damenly.su>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

btrfs: use latest_dev in btrfs_show_devname

Commit 6605fd2f394bba0a0059df2b6cfc87b0b6d393a2 upstream.

The test case btrfs/238 reports the warning below:

 WARNING: CPU: 3 PID: 481 at fs/btrfs/super.c:2509 btrfs_show_devname+0x104/0x1e8 [btrfs]
 CPU: 2 PID: 1 Comm: systemd Tainted: G        W  O 5.14.0-rc1-custom #72
 Hardware name: QEMU QEMU Virtual Machine, BIOS 0.0.0 02/06/2015
 Call trace:
   btrfs_show_devname+0x108/0x1b4 [btrfs]
   show_mountinfo+0x234/0x2c4
   m_show+0x28/0x34
   seq_read_iter+0x12c/0x3c4
   vfs_read+0x29c/0x2c8
   ksys_read+0x80/0xec
   __arm64_sys_read+0x28/0x34
   invoke_syscall+0x50/0xf8
   do_el0_svc+0x88/0x138
   el0_svc+0x2c/0x8c
   el0t_64_sync_handler+0x84/0xe4
   el0t_64_sync+0x198/0x19c

Reason:
While btrfs_prepare_sprout() moves the fs_devices::devices into
fs_devices::seed_list, the btrfs_show_devname() searches for the devices
and found none, leading to the warning as in above.

Fix:
latest_dev is updated according to the changes to the device list.
That means we could use the latest_dev->name to show the device name in
/proc/self/mounts, the pointer will be always valid as it's assigned
before the device is deleted from the list in remove or replace.
The RCU protection is sufficient as the device structure is freed after
synchronization.

Reported-by: Su Yue <l@damenly.su>
Tested-by: Su Yue <l@damenly.su>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

btrfs: convert latest_bdev type to btrfs_device and rename

Commit d24fa5c1da08026be9959baca309fa0adf8708bf upstream.

In preparation to fix a bug in btrfs_show_devname().

Convert fs_devices::latest_bdev type from struct block_device to struct
btrfs_device and, rename the member to fs_devices::latest_dev.
So that btrfs_show_devname() can use fs_devices::latest_dev::name.

Tested-by: Su Yue <l@damenly.su>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

audit: improve robustness of the audit queue handling

commit f4b3ee3c85551d2d343a3ba159304066523f730f upstream.

If the audit daemon were ever to get stuck in a stopped state the
kernel's kauditd_thread() could get blocked attempting to send audit
records to the userspace audit daemon.  With the kernel thread
blocked it is possible that the audit queue could grow unbounded as
certain audit record generating events must be exempt from the queue
limits else the system enter a deadlock state.

This patch resolves this problem by lowering the kernel thread's
socket sending timeout from MAX_SCHEDULE_TIMEOUT to HZ/10 and tweaks
the kauditd_send_queue() function to better manage the various audit
queues when connection problems occur between the kernel and the
audit daemon.  With this patch, the backlog may temporarily grow
beyond the defined limits when the audit daemon is stopped and the
system is under heavy audit pressure, but kauditd_thread() will
continue to make progress and drain the queues as it would for other
connection problems.  For example, with the audit daemon put into a
stopped state and the system configured to audit every syscall it
was still possible to shutdown the system without a kernel panic,
deadlock, etc.; granted, the system was slow to shutdown but that is
to be expected given the extreme pressure of recording every syscall.

The timeout value of HZ/10 was chosen primarily through
experimentation and this developer's "gut feeling".  There is likely
no one perfect value, but as this scenario is limited in scope (root
privileges would be needed to send SIGSTOP to the audit daemon), it
is likely not worth exposing this as a tunable at present.  This can
always be done at a later date if it proves necessary.

Cc: stable@vger.kernel.org
Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking")
Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com>
Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com>
Reviewed-by: Richard Guy Briggs <rgb@redhat.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

dm btree remove: fix use after free in rebalance_children()

commit 1b8d2789dad0005fd5e7d35dab26a8e1203fb6da upstream.

Move dm_tm_unlock() after dm_tm_dec().

Cc: stable@vger.kernel.org
Signed-off-by: Joe Thornber <ejt@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ceph: fix up non-directory creation in SGID directories

commit fd84bfdddd169c219c3a637889a8b87f70a072c2 upstream.

Ceph always inherits the SGID bit if it is set on the parent inode,
while the generic inode_init_owner does not do this in a few cases where
it can create a possible security problem (cf. [1]).

Update ceph to strip the SGID bit just as inode_init_owner would.

This bug was detected by the mapped mount testsuite in [3]. The
testsuite tests all core VFS functionality and semantics with and
without mapped mounts. That is to say it functions as a generic VFS
testsuite in addition to a mapped mount testsuite. While working on
mapped mount support for ceph, SIGD inheritance was the only failing
test for ceph after the port.

The same bug was detected by the mapped mount testsuite in XFS in
January 2021 (cf. [2]).

[1]: commit 0fa3ecd87848 ("Fix up non-directory creation in SGID directories")
[2]: commit 01ea173e103e ("xfs: fix up non-directory creation in SGID directories")
[3]: https://git.kernel.org/fs/xfs/xfstests-dev.git

Cc: stable@vger.kernel.org
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

arm64: dts: ten64: remove redundant interrupt declaration for gpio-keys

commit c88c5e461939a06ae769a01649d5c6b5a156f883 upstream.

gpio-keys already 'inherits' the interrupts from the controller
of the specified GPIO, so having another declaration is redundant.
On >=v5.15 this started causing an oops under gpio_keys_probe as
the IRQ was already claimed.

Signed-off-by: Mathew McBride <matt@traverse.com.au>
Fixes: 418962eea358 ("arm64: dts: add device tree for Traverse Ten64 (LS1088A)")
Cc: stable@vger.kernel.org
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

recordmcount.pl: look for jgnop instruction as well as bcrl on s390

commit 85bf17b28f97ca2749968d8786dc423db320d9c2 upstream.

On s390, recordmcount.pl is looking for "bcrl 0,<xxx>" instructions in
the objdump -d outpout. However since binutils 2.37, objdump -d
display "jgnop <xxx>" for the same instruction. Update the
mcount_regex so that it accepts both.

Signed-off-by: Jerome Marchand <jmarchan@redhat.com>
Reviewed-by: Miroslav Benes <mbenes@suse.cz>
Acked-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20211210093827.1623286-1-jmarchan@redhat.com
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

s390/entry: fix duplicate tracking of irq nesting level

commit c9b12b59e2ea4c3c7cedec7efb071b649652f3a9 upstream.

In the current code, when exiting from idle, rcu_irq_enter() is
called twice during irq entry:

irq_entry_enter()-> rcu_irq_enter()
irq_enter() -> rcu_irq_enter()

This may lead to wrong results from rcu_is_cpu_rrupt_from_idle()
because of a wrong dynticks nmi nesting count. Fix this by only
calling irq_enter_rcu().

Cc: <stable@vger.kernel.org> # 5.12+
Reported-by: Mark Rutland <mark.rutland@arm.com>
Fixes: 56e62a737028 ("s390: convert to generic entry")
Signed-off-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

vdpa: check that offsets are within bounds

commit 3ed21c1451a14d139e1ceb18f2fa70865ce3195a upstream.

In this function "c->off" is a u32 and "size" is a long.  On 64bit systems
if "c->off" is greater than "size" then "size - c->off" is a negative and
we always return -E2BIG.  But on 32bit systems the subtraction is type
promoted to a high positive u32 value and basically any "c->len" is
accepted.

Fixes: 4c8cf31885f6 ("vhost: introduce vDPA-based backend")
Reported-by: Xie Yongji <xieyongji@bytedance.com>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Link: https://lore.kernel.org/r/20211208103337.GA4047@kili
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

virtio_ring: Fix querying of maximum DMA mapping size for virtio device

commit 817fc978b5a29b039db0418a91072b31c9aab152 upstream.

virtio_max_dma_size() returns the maximum DMA mapping size of the virtio
device by querying dma_max_mapping_size() for the device when the DMA
API is in use for the vring. Unfortunately, the device passed is
initialised by register_virtio_device() and does not inherit the DMA
configuration from its parent, resulting in SWIOTLB errors when bouncing
is enabled and the default 256K mapping limit (IO_TLB_SEGSIZE) is not
respected:

  | virtio-pci 0000:00:01.0: swiotlb buffer is full (sz: 294912 bytes), total 1024 (slots), used 725 (slots)

Follow the pattern used elsewhere in the virtio_ring code when calling
into the DMA layer and pass the parent device to dma_max_mapping_size()
instead.

Cc: Marc Zyngier <maz@kernel.org>
Cc: Quentin Perret <qperret@google.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Jason Wang <jasowang@redhat.com>
Signed-off-by: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20211201112018.25276-1-will@kernel.org
Acked-by: Jason Wang <jasowang@redhat.com>
Tested-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Fixes: e6d6dd6c875e ("virtio: Introduce virtio_max_dma_size()")
Cc: Joerg Roedel <jroedel@suse.de>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Steven Price <steven.price@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Cc: stable@vger.kernel.org
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

vduse: check that offset is within bounds in get_config()

commit dc1db0060c02d119fd4196924eff2d1129e9a442 upstream.

This condition checks "len" but it does not check "offset" and that
could result in an out of bounds read if "offset > dev->config_size".
The problem is that since both variables are unsigned the
"dev->config_size - offset" subtraction would result in a very high
unsigned value.

I think these checks might not be necessary because "len" and "offset"
are supposed to already have been validated using the
vhost_vdpa_config_validate() function.  But I do not know the code
perfectly, and I like to be safe.

Fixes: c8a6153b6c59 ("vduse: Introduce VDUSE - vDPA Device in Userspace")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Link: https://lore.kernel.org/r/20211208150956.GA29160@kili
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

vduse: fix memory corruption in vduse_dev_ioctl()

commit ff9f9c6e74848170fcb45c8403c80d661484c8c9 upstream.

The "config.offset" comes from the user.  There needs to a check to
prevent it being out of bounds.  The "config.offset" and
"dev->config_size" variables are both type u32.  So if the offset if
out of bounds then the "dev->config_size - config.offset" subtraction
results in a very high u32 value.  The out of bounds offset can result
in memory corruption.

Fixes: c8a6153b6c59 ("vduse: Introduce VDUSE - vDPA Device in Userspace")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Link: https://lore.kernel.org/r/20211208103307.GA3778@kili
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

bpf, selftests: Update test case for atomic cmpxchg on r0 with pointer

commit e523102cb719cbad1673b6aa2a4d5c1fa6f13799 upstream.

Fix up unprivileged test case results for 'Dest pointer in r0' verifier tests
given they now need to reject R0 containing a pointer value, and add a couple
of new related ones with 32bit cmpxchg as well.

  root@foo:~/bpf/tools/testing/selftests/bpf# ./test_verifier
  #0/u invalid and of negative number OK
  #0/p invalid and of negative number OK
  [...]
  #1268/p XDP pkt read, pkt_meta' <= pkt_data, bad access 1 OK
  #1269/p XDP pkt read, pkt_meta' <= pkt_data, bad access 2 OK
  #1270/p XDP pkt read, pkt_data <= pkt_meta', good access OK
  #1271/p XDP pkt read, pkt_data <= pkt_meta', bad access 1 OK
  #1272/p XDP pkt read, pkt_data <= pkt_meta', bad access 2 OK
  Summary: 1900 PASSED, 0 SKIPPED, 0 FAILED

Acked-by: Brendan Jackman <jackmanb@google.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

bpf: Fix kernel address leakage in atomic cmpxchg's r0 aux reg

commit a82fe085f344ef20b452cd5f481010ff96b5c4cd upstream.

The implementation of BPF_CMPXCHG on a high level has the following parameters:

  .-[old-val]                                          .-[new-val]
  BPF_R0 = cmpxchg{32,64}(DST_REG + insn->off, BPF_R0, SRC_REG)
                          `-[mem-loc]          `-[old-val]

Given a BPF insn can only have two registers (dst, src), the R0 is fixed and
used as an auxilliary register for input (old value) as well as output (returning
old value from memory location). While the verifier performs a number of safety
checks, it misses to reject unprivileged programs where R0 contains a pointer as
old value.

Through brute-forcing it takes about ~16sec on my machine to leak a kernel pointer
with BPF_CMPXCHG. The PoC is basically probing for kernel addresses by storing the
guessed address into the map slot as a scalar, and using the map value pointer as
R0 while SRC_REG has a canary value to detect a matching address.

Fix it by checking R0 for pointers, and reject if that's the case for unprivileged
programs.

Fixes: 5ffa25502b5a ("bpf: Add instructions for atomic_[cmp]xchg")
Reported-by: Ryota Shiga (Flatt Security)
Acked-by: Brendan Jackman <jackmanb@google.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

bpf, selftests: Add test case trying to taint map value pointer

commit b1a7288dedc6caf9023f2676b4f5ed34cf0d4029 upstream.

Add a test case which tries to taint map value pointer arithmetic into a
unknown scalar with subsequent export through the map.

Before fix:

  # ./test_verifier 1186
  #1186/u map access: trying to leak tained dst reg FAIL
  Unexpected success to load!
  verification time 24 usec
  stack depth 8
  processed 15 insns (limit 1000000) max_states_per_insn 0 total_states 1 peak_states 1 mark_read 1
  #1186/p map access: trying to leak tained dst reg FAIL
  Unexpected success to load!
  verification time 8 usec
  stack depth 8
  processed 15 insns (limit 1000000) max_states_per_insn 0 total_states 1 peak_states 1 mark_read 1
  Summary: 0 PASSED, 0 SKIPPED, 2 FAILED

After fix:

  # ./test_verifier 1186
  #1186/u map access: trying to leak tained dst reg OK
  #1186/p map access: trying to leak tained dst reg OK
  Summary: 2 PASSED, 0 SKIPPED, 0 FAILED

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

bpf: Make 32->64 bounds propagation slightly more robust

commit e572ff80f05c33cd0cb4860f864f5c9c044280b6 upstream.

Make the bounds propagation in __reg_assign_32_into_64() slightly more
robust and readable by aligning it similarly as we did back in the
__reg_combine_64_into_32() counterpart. Meaning, only propagate or
pessimize them as a smin/smax pair.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

bpf: Fix signed bounds propagation after mov32

commit 3cf2b61eb06765e27fec6799292d9fb46d0b7e60 upstream.

For the case where both s32_{min,max}_value bounds are positive, the
__reg_assign_32_into_64() directly propagates them to their 64 bit
counterparts, otherwise it pessimises them into [0,u32_max] universe and
tries to refine them later on by learning through the tnum as per comment
in mentioned function. However, that does not always happen, for example,
in mov32 operation we call zext_32_to_64(dst_reg) which invokes the
__reg_assign_32_into_64() as is without subsequent bounds update as
elsewhere thus no refinement based on tnum takes place.

Thus, not calling into the __update_reg_bounds() / __reg_deduce_bounds() /
__reg_bound_offset() triplet as we do, for example, in case of ALU ops via
adjust_scalar_min_max_vals(), will lead to more pessimistic bounds when
dumping the full register state:

Before fix:

  0: (b4) w0 = -1
  1: R0_w=invP4294967295
     (id=0,imm=ffffffff,
      smin_value=4294967295,smax_value=4294967295,
      umin_value=4294967295,umax_value=4294967295,
      var_off=(0xffffffff; 0x0),
      s32_min_value=-1,s32_max_value=-1,
      u32_min_value=-1,u32_max_value=-1)

  1: (bc) w0 = w0
  2: R0_w=invP4294967295
     (id=0,imm=ffffffff,
      smin_value=0,smax_value=4294967295,
      umin_value=4294967295,umax_value=4294967295,
      var_off=(0xffffffff; 0x0),
      s32_min_value=-1,s32_max_value=-1,
      u32_min_value=-1,u32_max_value=-1)

Technically, the smin_value=0 and smax_value=4294967295 bounds are not
incorrect, but given the register is still a constant, they break assumptions
about const scalars that smin_value == smax_value and umin_value == umax_value.

After fix:

  0: (b4) w0 = -1
  1: R0_w=invP4294967295
     (id=0,imm=ffffffff,
      smin_value=4294967295,smax_value=4294967295,
      umin_value=4294967295,umax_value=4294967295,
      var_off=(0xffffffff; 0x0),
      s32_min_value=-1,s32_max_value=-1,
      u32_min_value=-1,u32_max_value=-1)

  1: (bc) w0 = w0
  2: R0_w=invP4294967295
     (id=0,imm=ffffffff,
      smin_value=4294967295,smax_value=4294967295,
      umin_value=4294967295,umax_value=4294967295,
      var_off=(0xffffffff; 0x0),
      s32_min_value=-1,s32_max_value=-1,
      u32_min_value=-1,u32_max_value=-1)

Without the smin_value == smax_value and umin_value == umax_value invariant
being intact for const scalars, it is possible to leak out kernel pointers
from unprivileged user space if the latter is enabled. For example, when such
registers are involved in pointer arithmtics, then adjust_ptr_min_max_vals()
will taint the destination register into an unknown scalar, and the latter
can be exported and stored e.g. into a BPF map value.

Fixes: 3f50f132d840 ("bpf: Verifier, do explicit ALU32 bounds tracking")
Reported-by: Kuee K1r0a <liulin063@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

bpf, selftests: Add test case for atomic fetch on spilled pointer

commit 180486b430f4e22cc00a478163d942804baae4b5 upstream.

Test whether unprivileged would be able to leak the spilled pointer either
by exporting the returned value from the atomic{32,64} operation or by reading
and exporting the value from the stack after the atomic operation took place.

Note that for unprivileged, the below atomic cmpxchg test case named "Dest
pointer in r0 - succeed" is failing. The reason is that in the dst memory
location (r10 -8) there is the spilled register r10:

  0: R1=ctx(id=0,off=0,imm=0) R10=fp0
  0: (bf) r0 = r10
  1: R0_w=fp0 R1=ctx(id=0,off=0,imm=0) R10=fp0
  1: (7b) *(u64 *)(r10 -8) = r0
  2: R0_w=fp0 R1=ctx(id=0,off=0,imm=0) R10=fp0 fp-8_w=fp
  2: (b7) r1 = 0
  3: R0_w=fp0 R1_w=invP0 R10=fp0 fp-8_w=fp
  3: (db) r0 = atomic64_cmpxchg((u64 *)(r10 -8), r0, r1)
  4: R0_w=fp0 R1_w=invP0 R10=fp0 fp-8_w=mmmmmmmm
  4: (79) r1 = *(u64 *)(r0 -8)
  5: R0_w=fp0 R1_w=invP(id=0) R10=fp0 fp-8_w=mmmmmmmm
  5: (b7) r0 = 0
  6: R0_w=invP0 R1_w=invP(id=0) R10=fp0 fp-8_w=mmmmmmmm
  6: (95) exit

However, allowing this case for unprivileged is a bit useless given an
update with a new pointer will fail anyway:

  0: R1=ctx(id=0,off=0,imm=0) R10=fp0
  0: (bf) r0 = r10
  1: R0_w=fp0 R1=ctx(id=0,off=0,imm=0) R10=fp0
  1: (7b) *(u64 *)(r10 -8) = r0
  2: R0_w=fp0 R1=ctx(id=0,off=0,imm=0) R10=fp0 fp-8_w=fp
  2: (db) r0 = atomic64_cmpxchg((u64 *)(r10 -8), r0, r10)
  R10 leaks addr into mem

Acked-by: Brendan Jackman <jackmanb@google.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
[only backport one test for 5.15.y - gregkh]
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

bpf: Fix kernel address leakage in atomic fetch

commit 7d3baf0afa3aa9102d6a521a8e4c41888bb79882 upstream.

The change in commit 37086bfdc737 ("bpf: Propagate stack bounds to registers
in atomics w/ BPF_FETCH") around check_mem_access() handling is buggy since
this would allow for unprivileged users to leak kernel pointers. For example,
an atomic fetch/and with -1 on a stack destination which holds a spilled
pointer will migrate the spilled register type into a scalar, which can then
be exported out of the program (since scalar != pointer) by dumping it into
a map value.

The original implementation of XADD was preventing this situation by using
a double call to check_mem_access() one with BPF_READ and a subsequent one
with BPF_WRITE, in both cases passing -1 as a placeholder value instead of
register as per XADD semantics since it didn't contain a value fetch. The
BPF_READ also included a check in check_stack_read_fixed_off() which rejects
the program if the stack slot is of __is_pointer_value() if dst_regno < 0.
The latter is to distinguish whether we're dealing with a regular stack spill/
fill or some arithmetical operation which is disallowed on non-scalars, see
also 6e7e63cbb023 ("bpf: Forbid XADD on spilled pointers for unprivileged
users") for more context on check_mem_access() and its handling of placeholder
value -1.

One minimally intrusive option to fix the leak is for the BPF_FETCH case to
initially check the BPF_READ case via check_mem_access() with -1 as register,
followed by the actual load case with non-negative load_reg to propagate
stack bounds to registers.

Fixes: 37086bfdc737 ("bpf: Propagate stack bounds to registers in atomics w/ BPF_FETCH")
Reported-by: <n4ke4mry@gmail.com>
Acked-by: Brendan Jackman <jackmanb@google.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

firmware: arm_scpi: Fix string overflow in SCPI genpd driver

commit 865ed67ab955428b9aa771d8b4f1e4fb7fd08945 upstream.

Without the bound checks for scpi_pd->name, it could result in the buffer
overflow when copying the SCPI device name from the corresponding device
tree node as the name string is set at maximum size of 30.

Let us fix it by using devm_kasprintf so that the string buffer is
allocated dynamically.

Fixes: 8bec4337ad40 ("firmware: scpi: add device power domain support using genpd")
Reported-by: Pedro Batista <pedbap.g@gmail.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Cc: stable@vger.kernel.org
Cc: Cristian Marussi <cristian.marussi@arm.com>
Link: https://lore.kernel.org/r/20211209120456.696879-1-sudeep.holla@arm.com'
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>