Merge tag 'loongarch-fixes-6.1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson

Pull LoongArch fixes from Huacai Chen:
 "Fix two build warnings, a copy_thread() bug, two page table
  manipulation bugs, and some trivial cleanups"

* tag 'loongarch-fixes-6.1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson:
  docs/zh_CN/LoongArch: Fix wrong description of FPRs Note
  LoongArch: Fix unsigned comparison with less than zero
  LoongArch: Set _PAGE_DIRTY only if _PAGE_MODIFIED is set in {pmd,pte}_mkwrite()
  LoongArch: Set _PAGE_DIRTY only if _PAGE_WRITE is set in {pmd,pte}_mkdirty()
  LoongArch: Clear FPU/SIMD thread info flags for kernel thread
  LoongArch: SMP: Change prefix from loongson3 to loongson
  LoongArch: Combine acpi_boot_table_init() and acpi_boot_init()
  LoongArch: Makefile: Use "grep -E" instead of "egrep"

Merge tag 'ext4_for_linus_stable2' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4

Pull ext4 fixes from Ted Ts'o:
 "Fix a regression in the lazytime code that was introduced in v6.1-rc1,
  and a use-after-free that can be triggered by a maliciously corrupted
  file system"

* tag 'ext4_for_linus_stable2' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  fs: do not update freeing inode i_io_list
  ext4: fix use-after-free in ext4_ext_shift_extents

wifi: wilc1000: validate number of channels

There is no validation of 'e->no_of_channels' which can trigger an
out-of-bounds write in the following 'memset' call. Validate that the
number of channels does not extends beyond the size of the channel list
element.

Signed-off-by: Phil Turnbull <[email protected]>
Tested-by: Ajay Kathat <[email protected]>
Acked-by: Ajay Kathat <[email protected]>
Signed-off-by: Kalle Valo <[email protected]>
Link: https://lore.kernel.org/r/[email protected]

wifi: wilc1000: validate length of IEEE80211_P2P_ATTR_CHANNEL_LIST attribute

Validate that the IEEE80211_P2P_ATTR_CHANNEL_LIST attribute contains
enough space for a 'struct wilc_attr_oper_ch'. If the attribute is too
small then it can trigger an out-of-bounds write later in the function.

'struct wilc_attr_oper_ch' is variable sized so also check 'attr_len'
does not extend beyond the end of 'buf'.

Signed-off-by: Phil Turnbull <[email protected]>
Tested-by: Ajay Kathat <[email protected]>
Acked-by: Ajay Kathat <[email protected]>
Signed-off-by: Kalle Valo <[email protected]>
Link: https://lore.kernel.org/r/[email protected]

wifi: wilc1000: validate length of IEEE80211_P2P_ATTR_OPER_CHANNEL attribute

Validate that the IEEE80211_P2P_ATTR_OPER_CHANNEL attribute contains
enough space for a 'struct struct wilc_attr_oper_ch'. If the attribute is
too small then it triggers an out-of-bounds write later in the function.

Signed-off-by: Phil Turnbull <[email protected]>
Tested-by: Ajay Kathat <[email protected]>
Acked-by: Ajay Kathat <[email protected]>
Signed-off-by: Kalle Valo <[email protected]>
Link: https://lore.kernel.org/r/[email protected]

wifi: wilc1000: validate pairwise and authentication suite offsets

There is no validation of 'offset' which can trigger an out-of-bounds
read when extracting RSN capabilities.

Signed-off-by: Phil Turnbull <[email protected]>
Tested-by: Ajay Kathat <[email protected]>
Acked-by: Ajay Kathat <[email protected]>
Signed-off-by: Kalle Valo <[email protected]>
Link: https://lore.kernel.org/r/[email protected]

can: mcba_usb: Fix termination command argument

Microchip USB Analyzer can activate the internal termination resistors
by setting the "termination" option ON, or OFF to to deactivate them.
As I've observed, both with my oscilloscope and captured USB packets
below, you must send "0" to turn it ON, and "1" to turn it OFF.

From the schematics in the user's guide, I can confirm that you must
drive the CAN_RES signal LOW "0" to activate the resistors.

Reverse the argument value of usb_msg.termination to fix this.

These are the two commands sequence, ON then OFF.

> No.     Time           Source                Destination           Protocol Length Info
>       1 0.000000       host                  1.3.1                 USB      46     URB_BULK out
>
> Frame 1: 46 bytes on wire (368 bits), 46 bytes captured (368 bits)
> USB URB
> Leftover Capture Data: a80000000000000000000000000000000000a8
>
> No.     Time           Source                Destination           Protocol Length Info
>       2 4.372547       host                  1.3.1                 USB      46     URB_BULK out
>
> Frame 2: 46 bytes on wire (368 bits), 46 bytes captured (368 bits)
> USB URB
> Leftover Capture Data: a80100000000000000000000000000000000a9

Signed-off-by: Yasushi SHOJI <[email protected]>
Link: https://lore.kernel.org/all/[email protected]
Signed-off-by: Marc Kleine-Budde <[email protected]>

can: m_can: Add check for devm_clk_get

Since the devm_clk_get may return error,
it should be better to add check for the cdev->hclk,
as same as cdev->cclk.

Fixes: f524f829b75a ("can: m_can: Create a m_can platform framework")
Signed-off-by: Jiasheng Jiang <[email protected]>
Link: https://lore.kernel.org/all/[email protected]
Signed-off-by: Marc Kleine-Budde <[email protected]>

can: m_can: pci: add missing m_can_class_free_dev() in probe/remove methods

In m_can_pci_remove() and error handling path of m_can_pci_probe(),
m_can_class_free_dev() should be called to free resource allocated by
m_can_class_allocate_dev(), otherwise there will be memleak.

Fixes: cab7ffc0324f ("can: m_can: add PCI glue driver for Intel Elkhart Lake")
Signed-off-by: Zhang Changzhong <[email protected]>
Reviewed-by: Jarkko Nikula <[email protected]>
Link: https://lore.kernel.org/all/[email protected]
Signed-off-by: Marc Kleine-Budde <[email protected]>

can: etas_es58x: es58x_init_netdev(): free netdev when register_candev()

In case of register_candev() fails, clear
es58x_dev->netdev[channel_idx] and add free_candev(). Otherwise
es58x_free_netdevs() will unregister the netdev that has never been
registered.

Fixes: 8537257874e9 ("can: etas_es58x: add core support for ETAS ES58X CAN USB interfaces")
Signed-off-by: Zhang Changzhong <[email protected]>
Acked-by: Arunachalam Santhanam <[email protected]>
Acked-by: Vincent Mailhol <[email protected]>
Link: https://lore.kernel.org/all/[email protected]
Signed-off-by: Marc Kleine-Budde <[email protected]>

can: cc770: cc770_isa_probe(): add missing free_cc770dev()

Add the missing free_cc770dev() before return from cc770_isa_probe()
in the register_cc770dev() error handling case.

In addition, remove blanks before goto labels.

Fixes: 7e02e5433e00 ("can: cc770: legacy CC770 ISA bus driver")
Signed-off-by: Zhang Changzhong <[email protected]>
Link: https://lore.kernel.org/all/[email protected]
Signed-off-by: Marc Kleine-Budde <[email protected]>

can: sja1000_isa: sja1000_isa_probe(): add missing free_sja1000dev()

Add the missing free_sja1000dev() before return from
sja1000_isa_probe() in the register_sja1000dev() error handling case.

In addition, remove blanks before goto labels.

Fixes: 2a6ba39ad6a2 ("can: sja1000: legacy SJA1000 ISA bus driver")
Signed-off-by: Zhang Changzhong <[email protected]>
Link: https://lore.kernel.org/all/[email protected]
Signed-off-by: Marc Kleine-Budde <[email protected]>

can: sja1000: fix size of OCR_MODE_MASK define

bitfield mode in ocr register has only 2 bits not 3, so correct
the OCR_MODE_MASK define.

Signed-off-by: Heiko Schocher <[email protected]>
Link: https://lore.kernel.org/all/[email protected]
Signed-off-by: Marc Kleine-Budde <[email protected]>

can: can327: can327_feed_frame_to_netdev(): fix potential skb leak when netdev is down

In can327_feed_frame_to_netdev(), it did not free the skb when netdev
is down, and all callers of can327_feed_frame_to_netdev() did not free
allocated skb too. That would trigger skb leak.

Fix it by adding kfree_skb() in can327_feed_frame_to_netdev() when netdev
is down. Not tested, just compiled.

Fixes: 43da2f07622f ("can: can327: CAN/ldisc driver for ELM327 based OBD-II adapters")
Signed-off-by: Ziyang Xuan <[email protected]>
Link: https://lore.kernel.org/all/[email protected]
Reviewed-by: Max Staudt <[email protected]>
Cc: [email protected]
Signed-off-by: Marc Kleine-Budde <[email protected]>

Merge tag 'v6.2-rockchip-dts32-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip into arm/fixes

Disabling of the unreliable arm-global-timer on earliest
Rockchip SoCs, due to its frequency being bound to the
changing cpu clock.

* tag 'v6.2-rockchip-dts32-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip:
  ARM: dts: rockchip: disable arm_global_timer on rk3066 and rk3188

net: use %pS for kfree_skb tracing event location

For the cases where 'reason' doesn't give any clue, it's still
nice to be able to track the kfree_skb caller location. %p doesn't
help much so let's use %pS which prints the symbol+offset.

Signed-off-by: Stanislav Fomichev <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>

MAINTAINERS: add S390 MM section

Alexander Gordeev and Gerald Schaefer are covering the whole s390 specific
memory management code. Reflect that by adding a new S390 MM section to
MAINTAINERS.

Also rename the S390 section to S390 ARCHITECTURE to be a bit more precise.

Acked-by: Gerald Schaefer <[email protected]>
Acked-by: Christian Borntraeger <[email protected]>
Acked-by: Alexander Gordeev <[email protected]>
Acked-by: Vasily Gorbik <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Alexander Gordeev <[email protected]>

s390/crashdump: fix TOD programmable field size

The size of the TOD programmable field was incorrectly increased from
four to eight bytes with commit 1a2c5840acf9 ("s390/dump: cleanup CPU
save area handling").
This leads to an elf notes section NT_S390_TODPREG which has a size of
eight instead of four bytes in case of kdump, however even worse is
that the contents is incorrect: it is supposed to contain only the
contents of the TOD programmable field, but in fact contains a mix of
the TOD programmable field (32 bit upper bits) and parts of the CPU
timer register (lower 32 bits).

Fix this by simply changing the size of the todpreg field within the
save area structure. This will implicitly also fix the size of the
corresponding elf notes sections.

This also gets rid of this compile time warning:

in function ‘fortify_memcpy_chk’,
    inlined from ‘save_area_add_regs’ at arch/s390/kernel/crash_dump.c:99:2:
./include/linux/fortify-string.h:413:25: error: call to ‘__read_overflow2_field’
   declared with attribute warning: detected read beyond size of field
   (2nd parameter); maybe use struct_group()? [-Werror=attribute-warning]
  413 |                         __read_overflow2_field(q_size_field, size);
      |                         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Fixes: 1a2c5840acf9 ("s390/dump: cleanup CPU save area handling")
Reviewed-by: Christian Borntraeger <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Alexander Gordeev <[email protected]>

net: thunderx: Fix the ACPI memory leak

The ACPI buffer memory (string.pointer) should be freed as the buffer is
not used after returning from bgx_acpi_match_id(), free it to prevent
memory leak.

Fixes: 46b903a01c05 ("net, thunder, bgx: Add support to get MAC address from ACPI.")
Signed-off-by: Yu Liao <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>

perf: Consider OS filter fail

Some PMUs (notably the traditional hardware kind) have boundary issues
with the OS filter. Specifically, it is possible for
perf_event_attr::exclude_kernel=1 events to trigger in-kernel due to
SKID or errata.

This can upset the sigtrap logic some and trigger the WARN.

However, if this invalid sample is the first we must not loose the
SIGTRAP, OTOH if it is the second, it must not override the
pending_addr with a (possibly) invalid one.

Fixes: ca6c21327c6a ("perf: Fix missing SIGTRAPs")
Reported-by: Pengfei Xu <[email protected]>
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Reviewed-by: Marco Elver <[email protected]>
Tested-by: Pengfei Xu <[email protected]>
Link: https://lkml.kernel.org/r/[email protected]

perf: Fixup SIGTRAP and sample_flags interaction

The perf_event_attr::sigtrap functionality relies on data->addr being
set. However commit 7b0846301531 ("perf: Use sample_flags for addr")
changed this to only initialize data->addr when not 0.

Fixes: 7b0846301531 ("perf: Use sample_flags for addr")
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Link: https://lkml.kernel.org/r/Y3426b4OimE%2FI5po%40hirez.programming.kicks-ass.net

octeontx2-af: Fix reference count issue in rvu_sdp_init()

pci_get_device() will decrease the reference count for the *from*
parameter. So we don't need to call put_device() to decrease the
reference. Let's remove the put_device() in the loop and only decrease
the reference count of the returned 'pdev' for the last loop because it
will not be passed to pci_get_device() as input parameter. We don't need
to check if 'pdev' is NULL because it is already checked inside
pci_dev_put(). Also add pci_dev_put() for the error path.

Fixes: fe1939bb2340 ("octeontx2-af: Add SDP interface support")
Signed-off-by: Xiongfeng Wang <[email protected]>
Reviewed-by: Saeed Mahameed <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>

net: altera_tse: release phylink resources in tse_shutdown()

Call phylink_disconnect_phy() in tse_shutdown() to release the
resources occupied by phylink_of_phy_connect() in the tse_open().

Fixes: fef2998203e1 ("net: altera: tse: convert to phylink")
Signed-off-by: Liu Jian <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>

virtio_net: Fix probe failed when modprobe virtio_net

When doing the following test steps, an error was found:
  step 1: modprobe virtio_net succeeded
    # modprobe virtio_net        <-- OK

  step 2: fault injection in register_netdevice()
    # modprobe -r virtio_net     <-- OK
    # ...
      FAULT_INJECTION: forcing a failure.
      name failslab, interval 1, probability 0, space 0, times 0
      CPU: 0 PID: 3521 Comm: modprobe
      Hardware name: QEMU Standard PC (i440FX + PIIX, 1996),
      Call Trace:
       <TASK>
       ...
       should_failslab+0xa/0x20
       ...
       dev_set_name+0xc0/0x100
       netdev_register_kobject+0xc2/0x340
       register_netdevice+0xbb9/0x1320
       virtnet_probe+0x1d72/0x2658 [virtio_net]
       ...
       </TASK>
      virtio_net: probe of virtio0 failed with error -22

  step 3: modprobe virtio_net failed
    # modprobe virtio_net        <-- failed
      virtio_net: probe of virtio0 failed with error -2

The root cause of the problem is that the queues are not
disable on the error handling path when register_netdevice()
fails in virtnet_probe(), resulting in an error "-ENOENT"
returned in the next modprobe call in setup_vq().

virtio_pci_modern_device uses virtqueues to send or
receive message, and "queue_enable" records whether the
queues are available. In vp_modern_find_vqs(), all queues
will be selected and activated, but once queues are enabled
there is no way to go back except reset.

Fix it by reset virtio device on error handling path. This
makes error handling follow the same order as normal device
cleanup in virtnet_remove() which does: unregister, destroy
failover, then reset. And that flow is better tested than
error handling so we can be reasonably sure it works well.

Fixes: 024655555021 ("virtio_net: fix use after free on allocation failure")
Signed-off-by: Li Zetao <[email protected]>
Acked-by: Michael S. Tsirkin <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>

net/mlx5e: MACsec, block offload requests with encrypt off

Currently offloading MACsec with authentication only (encrypt
property set to off) is not supported, block such requests
when adding/updating a macsec device.

Fixes: 8ff0ac5be144 ("net/mlx5: Add MACsec offload Tx command support")
Signed-off-by: Emeel Hakim <[email protected]>
Reviewed-by: Raed Salem <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5e: MACsec, fix Tx SA active field update

Currently during update Tx security association (SA) flow, the Tx SA
active state is updated only if the Tx SA in question is the same SA
that the MACsec interface is using for Tx,in consequence when the
MACsec interface chose to work with this Tx SA later, where this SA
for example should have been updated to active state and it was not,
the relevant Tx SA HW context won't be installed, hence the MACSec
flow won't be offloaded.

Fix by update Tx SA active state as part of update flow regardless
whether the SA in question is the same Tx SA used by the MACsec
interface.

Fixes: 8ff0ac5be144 ("net/mlx5: Add MACsec offload Tx command support")
Signed-off-by: Raed Salem <[email protected]>
Reviewed-by: Emeel Hakim <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5e: MACsec, remove replay window size limitation in offload path

Currently offload path limits replay window size to 32/64/128/256 bits,
such a limitation should not exist since software allows it.
Remove such limitation.

Fixes: eb43846b43c3 ("net/mlx5e: Support MACsec offload replay window")
Signed-off-by: Emeel Hakim <[email protected]>
Reviewed-by: Raed Salem <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5e: MACsec, fix add Rx security association (SA) rule memory leak

Currently MACsec's add Rx SA flow steering (fs) rule routine
uses a spec object which is dynamically allocated and do
not free it upon leaving. The above led to a memory leak.

Fix by freeing dynamically allocated objects.

Fixes: 3b20949cb21b ("net/mlx5e: Add MACsec RX steering rules")
Signed-off-by: Emeel Hakim <[email protected]>
Reviewed-by: Raed Salem <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5e: MACsec, fix mlx5e_macsec_update_rxsa bail condition and functionality

Fix update Rx SA wrong bail condition, naturally update functionality
needs to check that something changed otherwise bailout currently the
active state check does just the opposite, furthermore unlike deactivate
path which remove the macsec rules to deactivate the offload, the
activation path does not include the counter part installation of the
macsec rules.

Fix by using correct bailout condition and when Rx SA changes state to
active then add the relevant macsec rules.

While at it, refine function name to reflect more precisely its role.

Fixes: aae3454e4d4c ("net/mlx5e: Add MACsec offload Rx command support")
Signed-off-by: Raed Salem <[email protected]>
Reviewed-by: Emeel Hakim <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5e: MACsec, fix update Rx secure channel active field

The main functionality for this operation is to update the
active state of the Rx security channel (SC) if the new
active setting is different from the current active state
of this Rx SC, however the relevant active state check is
done post updating the current active state to match the
new active state, effectively blocks any offload state
update for the Rx SC in question.

Fix by delay the assignment to be post the relevant check.

Fixes: aae3454e4d4c ("net/mlx5e: Add MACsec offload Rx command support")
Signed-off-by: Raed Salem <[email protected]>
Reviewed-by: Emeel Hakim <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5e: MACsec, fix memory leak when MACsec device is deleted

When the MACsec netdevice is deleted, all related Rx/Tx HW/SW
states should be released/deallocated, however currently part
of the Rx security channel association data is not cleaned
properly, hence the memory leaks.

Fix by make sure all related Rx Sc resources are cleaned/freed,
while at it improve code by grouping release SC context in a
function so it can be used in both delete MACsec device and
delete Rx SC operations.

Fixes: 5a39816a75e5 ("net/mlx5e: Add MACsec offload SecY support")
Signed-off-by: Raed Salem <[email protected]>
Reviewed-by: Emeel Hakim <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5e: MACsec, fix RX data path 16 RX security channel limit

Currently the data path metadata flow id mask wrongly limits the
number of different RX security channels (SC) to 16, whereas in
adding RX SC the limit is "2^16 - 1" this cause an overlap in
metadata flow id once more than 16 RX SCs is added, this corrupts
MACsec RX offloaded flow handling.

Fix by using the correct mask, while at it improve code to use this
mask when adding the Rx rule and improve visibility of such errors
by adding debug massage.

Fixes: b7c9400cbc48 ("net/mlx5e: Implement MACsec Rx data path using MACsec skb_metadata_dst")
Signed-off-by: Raed Salem <[email protected]>
Reviewed-by: Emeel Hakim <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5e: Use kvfree() in mlx5e_accel_fs_tcp_create()

'accel_tcp' is allocated by kvzalloc(), which should freed by kvfree().

Fixes: f52f2faee581 ("net/mlx5e: Introduce flow steering API")
Signed-off-by: YueHaibing <[email protected]>
Reviewed-by: Tariq Toukan <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5e: Fix a couple error codes

If kvzalloc() fails then return -ENOMEM.  Don't return success.

Fixes: 3b20949cb21b ("net/mlx5e: Add MACsec RX steering rules")
Fixes: e467b283ffd5 ("net/mlx5e: Add MACsec TX steering rules")
Signed-off-by: Dan Carpenter <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5e: Fix use-after-free when reverting termination table

When having multiple dests with termination tables and second one
or afterwards fails the driver reverts usage of term tables but
doesn't reset the assignment in attr->dests[num_vport_dests].termtbl
which case a use-after-free when releasing the rule.
Fix by resetting the assignment of termtbl to null.

Fixes: 10caabdaad5a ("net/mlx5e: Use termination table for VLAN push actions")
Signed-off-by: Roi Dayan <[email protected]>
Reviewed-by: Maor Dickman <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5: Fix uninitialized variable bug in outlen_write()

If sscanf() return 0, outlen is uninitialized and used in kzalloc(),
this is unexpected. We should return -EINVAL if the string is invalid.

Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters")
Signed-off-by: YueHaibing <[email protected]>
Reviewed-by: Leon Romanovsky <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5: E-switch, Fix duplicate lag creation

If creating bond first and then enabling sriov in switchdev mode,
will hit the following syndrome:

mlx5_core 0000:08:00.0: mlx5_cmd_out_err:778:(pid 25543): CREATE_LAG(0x840) op_mod(0x0) failed, status bad parameter(0x3), syndrome (0x7d49cb), err(-22)

The reason is because the offending patch removes eswitch mode
none. In vf lag, the checking of eswitch mode none is replaced
by checking if sriov is enabled. But when driver enables sriov,
it triggers the bond workqueue task first and then setting sriov
number in pci_enable_sriov(). So the check fails.

Fix it by checking if sriov is enabled using eswitch internal
counter that is set before triggering the bond workqueue task.

Fixes: f019679ea5f2 ("net/mlx5: E-switch, Remove dependency between sriov and eswitch mode")
Signed-off-by: Chris Mi <[email protected]>
Reviewed-by: Roi Dayan <[email protected]>
Reviewed-by: Mark Bloch <[email protected]>
Reviewed-by: Vlad Buslov <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5: E-switch, Destroy legacy fdb table when needed

The cited commit removes eswitch mode none. But when disabling
sriov in legacy mode or changing from switchdev to legacy mode
without sriov enabled, the legacy fdb table is not destroyed.

It is not the right behavior. Destroy legacy fdb table in above
two caes.

Fixes: f019679ea5f2 ("net/mlx5: E-switch, Remove dependency between sriov and eswitch mode")
Signed-off-by: Chris Mi <[email protected]>
Reviewed-by: Roi Dayan <[email protected]>
Reviewed-by: Eli Cohen <[email protected]>
Reviewed-by: Mark Bloch <[email protected]>
Reviewed-by: Vlad Buslov <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net/mlx5: DR, Fix uninitialized var warning

Smatch warns this:

drivers/net/ethernet/mellanox/mlx5/core/steering/dr_table.c:81
 mlx5dr_table_set_miss_action() error: uninitialized symbol 'ret'.

Initializing ret with -EOPNOTSUPP and fix missing action case.

Fixes: 7838e1725394 ("net/mlx5: DR, Expose steering table functionality")
Signed-off-by: YueHaibing <[email protected]>
Reviewed-by: Roi Dayan <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>

net: wwan: t7xx: Fix the ACPI memory leak

The ACPI buffer memory (buffer.pointer) should be freed as the
buffer is not used after acpi_evaluate_object(), free it to
prevent memory leak.

Fixes: 13e920d93e37 ("net: wwan: t7xx: Add core components")
Signed-off-by: Hanjun Guo <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>

octeontx2-pf: Add check for devm_kcalloc

As the devm_kcalloc may return NULL pointer,
it should be better to add check for the return
value, as same as the others.

Fixes: e8e095b3b370 ("octeontx2-af: cn10k: Bandwidth profiles config support")
Signed-off-by: Jiasheng Jiang <[email protected]>
Reviewed-by: Maciej Fijalkowski <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Paolo Abeni <[email protected]>

net: enetc: preserve TX ring priority across reconfiguration

In the blamed commit, a rudimentary reallocation procedure for RX buffer
descriptors was implemented, for the situation when their format changes
between normal (no PTP) and extended (PTP).

enetc_hwtstamp_set() calls enetc_close() and enetc_open() in a sequence,
and this sequence loses information which was previously configured in
the TX BDR Mode Register, specifically via the enetc_set_bdr_prio() call.
The TX ring priority is configured by tc-mqprio and tc-taprio, and
affects important things for TSN such as the TX time of packets. The
issue manifests itself most visibly by the fact that isochron --txtime
reports premature packet transmissions when PTP is first enabled on an
enetc interface.

Save the TX ring priority in a new field in struct enetc_bdr (occupies a
2 byte hole on arm64) in order to make this survive a ring reconfiguration.

Fixes: 434cebabd3a2 ("enetc: Add dynamic allocation of extended Rx BD rings")
Signed-off-by: Vladimir Oltean <[email protected]>
Reviewed-by: Alexander Lobakin <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>

Merge branch 'bonding-fix-bond-recovery-in-mode-2'

Jonathan Toppins says:

====================
bonding: fix bond recovery in mode 2

When a bond is configured with a non-zero updelay and in mode 2 the bond
never recovers after all slaves lose link. The first patch adds
selftests that demonstrate the issue and the second patch fixes the
issue by ignoring the updelay when there are no usable slaves.
====================

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>

bonding: fix link recovery in mode 2 when updelay is nonzero

Before this change when a bond in mode 2 lost link, all of its slaves
lost link, the bonding device would never recover even after the
expiration of updelay. This change removes the updelay when the bond
currently has no usable links. Conforming to bonding.txt section 13.1
paragraph 4.

Fixes: 41f891004063 ("bonding: ignore updelay param when there is no active slave")
Signed-off-by: Jonathan Toppins <[email protected]>
Acked-by: Jay Vosburgh <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

selftests: bonding: up/down delay w/ slave link flapping

Verify when a bond is configured with {up,down}delay and the link state
of slave members flaps if there are no remaining members up the bond
should immediately select a member to bring up. (from bonding.txt
section 13.1 paragraph 4)

Suggested-by: Liang Li <[email protected]>
Signed-off-by: Jonathan Toppins <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

net: marvell: prestera: add missing unregister_netdev() in prestera_port_create()

If prestera_port_sfp_bind() fails, unregister_netdev() should be called
in error handling path.

Compile tested only.

Fixes: 52323ef75414 ("net: marvell: prestera: add phylink support")
Signed-off-by: Zhang Changzhong <[email protected]>
Reviewed-by: Maciej Fijalkowski <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>

ethtool: avoiding integer overflow in ethtool_phys_id()

The value of an arithmetic expression "n * id.data" is subject
to possible overflow due to a failure to cast operands to a larger data
type before performing arithmetic. Used macro for multiplication instead
operator for avoiding overflow.

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

Signed-off-by: Maxim Korotkov <[email protected]>
Reviewed-by: Alexander Lobakin <[email protected]>
Reviewed-by: Andrew Lunn <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>

Merge branch 'nfc-st-nci-restructure-validating-logic-in-evt_transaction'

Martin Faltesek says:

====================
nfc: st-nci: Restructure validating logic in EVT_TRANSACTION

These are the same 3 patches that were applied in st21nfca here:
https://lore.kernel.org/netdev/20220607025729.1673212[email protected]
with a couple minor differences.

st-nci has nearly identical code to that of st21nfca for EVT_TRANSACTION,
except that there are two extra validation checks that are not present
in the st-nci code.

The 3/3 patch as coded for st21nfca pulls those checks in, bringing both
drivers into parity.
====================

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>

nfc: st-nci: fix incorrect sizing calculations in EVT_TRANSACTION

The transaction buffer is allocated by using the size of the packet buf,
and subtracting two which seems intended to remove the two tags which are
not present in the target structure. This calculation leads to under
counting memory because of differences between the packet contents and the
target structure. The aid_len field is a u8 in the packet, but a u32 in
the structure, resulting in at least 3 bytes always being under counted.
Further, the aid data is a variable length field in the packet, but fixed
in the structure, so if this field is less than the max, the difference is
added to the under counting.

To fix, perform validation checks progressively to safely reach the
next field, to determine the size of both buffers and verify both tags.
Once all validation checks pass, allocate the buffer and copy the data.
This eliminates freeing memory on the error path, as validation checks are
moved ahead of memory allocation.

Reported-by: Denis Efremov <[email protected]>
Reviewed-by: Guenter Roeck <[email protected]>
Fixes: 5d1ceb7f5e56 ("NFC: st21nfcb: Add HCI transaction event support")
Signed-off-by: Martin Faltesek <[email protected]>
Reviewed-by: Krzysztof Kozlowski <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

nfc: st-nci: fix memory leaks in EVT_TRANSACTION

Error path does not free previously allocated memory. Add devm_kfree() to
the failure path.

Reported-by: Denis Efremov <[email protected]>
Reviewed-by: Guenter Roeck <[email protected]>
Fixes: 5d1ceb7f5e56 ("NFC: st21nfcb: Add HCI transaction event support")
Signed-off-by: Martin Faltesek <[email protected]>
Reviewed-by: Krzysztof Kozlowski <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

nfc: st-nci: fix incorrect validating logic in EVT_TRANSACTION

The first validation check for EVT_TRANSACTION has two different checks
tied together with logical AND. One is a check for minimum packet length,
and the other is for a valid aid_tag. If either condition is true (fails),
then an error should be triggered. The fix is to change && to ||.

Reported-by: Denis Efremov <[email protected]>
Reviewed-by: Guenter Roeck <[email protected]>
Fixes: 5d1ceb7f5e56 ("NFC: st21nfcb: Add HCI transaction event support")
Signed-off-by: Martin Faltesek <[email protected]>
Reviewed-by: Krzysztof Kozlowski <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

ublk_drv: don't forward io commands in reserve order

Either ublk_can_use_task_work() is true or not, io commands are
forwarded to ublk server in reverse order, since llist_add() is
always to add one element to the head of the list.

Even though block layer doesn't guarantee request dispatch order,
requests should be sent to hardware in the sequence order generated
from io scheduler, which usually considers the request's LBA, and
order is often important for HDD.

So forward io commands in the sequence made from io scheduler by
aligning task work with current io_uring command's batch handling,
and it has been observed that both can get similar performance data
if IORING_SETUP_COOP_TASKRUN is set from ublk server.

Reported-by: Andreas Hindborg <[email protected]>
Cc: Damien Le Moal <[email protected]>
Signed-off-by: Ming Lei <[email protected]>
Reviewed-by: Damien Le Moal <[email protected]>
Reviewed-by: ZiyangZhang <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>

Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec

Steffen Klassert says:

====================
ipsec 2022-11-23

1) Fix "disable_policy" on ipv4 early demuxP Packets after
   the initial packet in a flow might be incorectly dropped
   on early demux if there are no matching policies.
   From Eyal Birger.

2) Fix a kernel warning in case XFRM encap type is not
   available. From Eyal Birger.

3) Fix ESN wrap around for GSO to avoid a double usage of a
    sequence number. From Christian Langrock.

4) Fix a send_acquire race with pfkey_register.
   From Herbert Xu.

5) Fix a list corruption panic in __xfrm_state_delete().
   Thomas Jarosch.

6) Fix an unchecked return value in xfrm6_init().
   Chen Zhongjin.

* 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec:
  xfrm: Fix ignored return value in xfrm6_init()
  xfrm: Fix oops in __xfrm_state_delete()
  af_key: Fix send_acquire race with pfkey_register
  xfrm: replay: Fix ESN wrap around for GSO
  xfrm: lwtunnel: squelch kernel warning in case XFRM encap type is not available
  xfrm: fix "disable_policy" on ipv4 early demux
====================

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>

Merge git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf

Pablo Neira Ayuso says:

====================
Netfilter fixes for net

1) Fix regression in ipset hash:ip with IPv4 range, from Vishwanath Pai.
   This is fixing up a bug introduced in the 6.0 release.

2) The "netfilter: ipset: enforce documented limit to prevent allocating
   huge memory" patch contained a wrong condition which makes impossible to
   add up to 64 clashing elements to a hash:net,iface type of set while it
   is the documented feature of the set type. The patch fixes the condition
   and thus makes possible to add the elements while keeps preventing
   allocating huge memory, from Jozsef Kadlecsik. This has been broken
   for several releases.

3) Missing locking when updating the flow block list which might lead
   a reader to crash. This has been broken since the introduction of the
   flowtable hardware offload support.

* git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
  netfilter: flowtable_offload: add missing locking
  netfilter: ipset: restore allowing 64 clashing elements in hash:net,iface
  netfilter: ipset: regression in ip_set_hash_ip.c
====================

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>

selftests/bpf: Add selftests for bpf_task_from_pid()

Add some selftest testcases that validate the expected behavior of the
bpf_task_from_pid() kfunc that was added in the prior patch.

Signed-off-by: David Vernet <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Alexei Starovoitov <[email protected]>

bpf: Add bpf_task_from_pid() kfunc

Callers can currently store tasks as kptrs using bpf_task_acquire(),
bpf_task_kptr_get(), and bpf_task_release(). These are useful if a
caller already has a struct task_struct *, but there may be some callers
who only have a pid, and want to look up the associated struct
task_struct * from that to e.g. find task->comm.

This patch therefore adds a new bpf_task_from_pid() kfunc which allows
BPF programs to get a struct task_struct * kptr from a pid.

Signed-off-by: David Vernet <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Alexei Starovoitov <[email protected]>

Documentation: networking: Update generic_netlink_howto URL

The documentation refers to invalid web page under www.linuxfoundation.org
The patch refers to a working URL under wiki.linuxfoundation.org

Signed-off-by: Nir Levy <[email protected]>
Link: https://lore.kernel.org/all/[email protected]/
Signed-off-by: Jakub Kicinski <[email protected]>

bpf: Unify and simplify btf_func_proto_check error handling

Replace 'err = x; break;' with 'return x;'.

Suggested-by: Daniel Borkmann <[email protected]>
Signed-off-by: Stanislav Fomichev <[email protected]>
Signed-off-by: Daniel Borkmann <[email protected]>
Link: https://lore.kernel.org/bpf/[email protected]

bpf: Update bpf_{g,s}etsockopt() documentation

* append missing optnames to the end
* simplify bpf_getsockopt()'s doc

Signed-off-by: Ji Rongfeng <[email protected]>
Link: https://lore.kernel.org/r/DU0P192MB15479B86200B1216EC90E162D6099@DU0P192MB1547.EURP192.PROD.OUTLOOK.COM
Signed-off-by: Martin KaFai Lau <[email protected]>

scripts: add rust in scripts/Makefile.package

Add rust argument at TAR_CONTENT in
scripts/Makefile.package script with alphabetical order.

Signed-off-by: Paran Lee <[email protected]>
Signed-off-by: Masahiro Yamada <[email protected]>

kbuild: fix "cat: .version: No such file or directory"

Since commit 2df8220cc511 ("kbuild: build init/built-in.a just once"),
the .version file is not touched at all when KBUILD_BUILD_VERSION is
given.

If KBUILD_BUILD_VERSION is specified and the .version file is missing
(for example right after 'make mrproper'), "No such file or director"
is shown. Even if the .version exists, it is irrelevant to the version
of the current build.

  $ make -j$(nproc) KBUILD_BUILD_VERSION=100 mrproper defconfig all
    [ snip ]
    BUILD   arch/x86/boot/bzImage
  cat: .version: No such file or directory
  Kernel: arch/x86/boot/bzImage is ready  (#)

Show KBUILD_BUILD_VERSION if it is given.

Fixes: 2df8220cc511 ("kbuild: build init/built-in.a just once")
Signed-off-by: Masahiro Yamada <[email protected]>
Reviewed-by: Nicolas Schier <[email protected]>

tracing: Free buffers when a used dynamic event is removed

After 65536 dynamic events have been added and removed, the "type" field
of the event then uses the first type number that is available (not
currently used by other events). A type number is the identifier of the
binary blobs in the tracing ring buffer (known as events) to map them to
logic that can parse the binary blob.

The issue is that if a dynamic event (like a kprobe event) is traced and
is in the ring buffer, and then that event is removed (because it is
dynamic, which means it can be created and destroyed), if another dynamic
event is created that has the same number that new event's logic on
parsing the binary blob will be used.

To show how this can be an issue, the following can crash the kernel:

 # cd /sys/kernel/tracing
 # for i in `seq 65536`; do
     echo 'p:kprobes/foo do_sys_openat2 $arg1:u32' > kprobe_events
 # done

For every iteration of the above, the writing to the kprobe_events will
remove the old event and create a new one (with the same format) and
increase the type number to the next available on until the type number
reaches over 65535 which is the max number for the 16 bit type. After it
reaches that number, the logic to allocate a new number simply looks for
the next available number. When an dynamic event is removed, that number
is then available to be reused by the next dynamic event created. That is,
once the above reaches the max number, the number assigned to the event in
that loop will remain the same.

Now that means deleting one dynamic event and created another will reuse
the previous events type number. This is where bad things can happen.
After the above loop finishes, the kprobes/foo event which reads the
do_sys_openat2 function call's first parameter as an integer.

 # echo 1 > kprobes/foo/enable
 # cat /etc/passwd > /dev/null
 # cat trace
             cat-2211    [005] ....  2007.849603: foo: (do_sys_openat2+0x0/0x130) arg1=4294967196
             cat-2211    [005] ....  2007.849620: foo: (do_sys_openat2+0x0/0x130) arg1=4294967196
             cat-2211    [005] ....  2007.849838: foo: (do_sys_openat2+0x0/0x130) arg1=4294967196
             cat-2211    [005] ....  2007.849880: foo: (do_sys_openat2+0x0/0x130) arg1=4294967196
 # echo 0 > kprobes/foo/enable

Now if we delete the kprobe and create a new one that reads a string:

 # echo 'p:kprobes/foo do_sys_openat2 +0($arg2):string' > kprobe_events

And now we can the trace:

 # cat trace
        sendmail-1942    [002] .....   530.136320: foo: (do_sys_openat2+0x0/0x240) arg1=             cat-2046    [004] .....   530.930817: foo: (do_sys_openat2+0x0/0x240) arg1="������������������������������������������������������������������������������������������������"
             cat-2046    [004] .....   530.930961: foo: (do_sys_openat2+0x0/0x240) arg1="������������������������������������������������������������������������������������������������"
             cat-2046    [004] .....   530.934278: foo: (do_sys_openat2+0x0/0x240) arg1="������������������������������������������������������������������������������������������������"
             cat-2046    [004] .....   530.934563: foo: (do_sys_openat2+0x0/0x240) arg1="������������������������������������������������������������������������������������������������"
            bash-1515    [007] .....   534.299093: foo: (do_sys_openat2+0x0/0x240) arg1="kkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkk���������@��4Z����;Y�����U

And dmesg has:

==================================================================
BUG: KASAN: use-after-free in string+0xd4/0x1c0
Read of size 1 at addr ffff88805fdbbfa0 by task cat/2049

 CPU: 0 PID: 2049 Comm: cat Not tainted 6.1.0-rc6-test+ #641
 Hardware name: Hewlett-Packard HP Compaq Pro 6300 SFF/339A, BIOS K01 v03.03 07/14/2016
 Call Trace:
  <TASK>
  dump_stack_lvl+0x5b/0x77
  print_report+0x17f/0x47b
  kasan_report+0xad/0x130
  string+0xd4/0x1c0
  vsnprintf+0x500/0x840
  seq_buf_vprintf+0x62/0xc0
  trace_seq_printf+0x10e/0x1e0
  print_type_string+0x90/0xa0
  print_kprobe_event+0x16b/0x290
  print_trace_line+0x451/0x8e0
  s_show+0x72/0x1f0
  seq_read_iter+0x58e/0x750
  seq_read+0x115/0x160
  vfs_read+0x11d/0x460
  ksys_read+0xa9/0x130
  do_syscall_64+0x3a/0x90
  entry_SYSCALL_64_after_hwframe+0x63/0xcd
 RIP: 0033:0x7fc2e972ade2
 Code: c0 e9 b2 fe ff ff 50 48 8d 3d b2 3f 0a 00 e8 05 f0 01 00 0f 1f 44 00 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 0f 05 <48> 3d 00 f0 ff ff 77 56 c3 0f 1f 44 00 00 48 83 ec 28 48 89 54 24
 RSP: 002b:00007ffc64e687c8 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
 RAX: ffffffffffffffda RBX: 0000000000020000 RCX: 00007fc2e972ade2
 RDX: 0000000000020000 RSI: 00007fc2e980d000 RDI: 0000000000000003
 RBP: 00007fc2e980d000 R08: 00007fc2e980c010 R09: 0000000000000000
 R10: 0000000000000022 R11: 0000000000000246 R12: 0000000000020f00
 R13: 0000000000000003 R14: 0000000000020000 R15: 0000000000020000
  </TASK>

 The buggy address belongs to the physical page:
 page:ffffea00017f6ec0 refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x5fdbb
 flags: 0xfffffc0000000(node=0|zone=1|lastcpupid=0x1fffff)
 raw: 000fffffc0000000 0000000000000000 ffffea00017f6ec8 0000000000000000
 raw: 0000000000000000 0000000000000000 00000000ffffffff 0000000000000000
 page dumped because: kasan: bad access detected

 Memory state around the buggy address:
  ffff88805fdbbe80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
  ffff88805fdbbf00: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
 >ffff88805fdbbf80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
                                ^
  ffff88805fdbc000: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
  ffff88805fdbc080: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
 ==================================================================

This was found when Zheng Yejian sent a patch to convert the event type
number assignment to use IDA, which gives the next available number, and
this bug showed up in the fuzz testing by Yujie Liu and the kernel test
robot. But after further analysis, I found that this behavior is the same
as when the event type numbers go past the 16bit max (and the above shows
that).

As modules have a similar issue, but is dealt with by setting a
"WAS_ENABLED" flag when a module event is enabled, and when the module is
freed, if any of its events were enabled, the ring buffer that holds that
event is also cleared, to prevent reading stale events. The same can be
done for dynamic events.

If any dynamic event that is being removed was enabled, then make sure the
buffers they were enabled in are now cleared.

Link: https://lkml.kernel.org/r/[email protected]
Link: https://lore.kernel.org/all/[email protected]/
Cc: [email protected]
Cc: Andrew Morton <[email protected]>
Depends-on: e18eb8783ec49 ("tracing: Add tracing_reset_all_online_cpus_unlocked() function")
Depends-on: 5448d44c38557 ("tracing: Add unified dynamic event framework")
Depends-on: 6212dd29683ee ("tracing/kprobes: Use dyn_event framework for kprobe events")
Depends-on: 065e63f951432 ("tracing: Only have rmmod clear buffers that its events were active in")
Depends-on: 575380da8b469 ("tracing: Only clear trace buffer on module unload if event was traced")
Fixes: 77b44d1b7c283 ("tracing/kprobes: Rename Kprobe-tracer to kprobe-event")
Reported-by: Zheng Yejian <[email protected]>
Reported-by: Yujie Liu <[email protected]>
Reported-by: kernel test robot <[email protected]>
Acked-by: Masami Hiramatsu (Google) <[email protected]>
Signed-off-by: Steven Rostedt (Google) <[email protected]>

tracing: Add tracing_reset_all_online_cpus_unlocked() function

Currently the tracing_reset_all_online_cpus() requires the
trace_types_lock held. But only one caller of this function actually has
that lock held before calling it, and the other just takes the lock so
that it can call it. More users of this function is needed where the lock
is not held.

Add a tracing_reset_all_online_cpus_unlocked() function for the one use
case that calls it without being held, and also add a lockdep_assert to
make sure it is held when called.

Then have tracing_reset_all_online_cpus() take the lock internally, such
that callers do not need to worry about taking it.

Link: https://lkml.kernel.org/r/[email protected]
Cc: Masami Hiramatsu <[email protected]>
Cc: Andrew Morton <[email protected]>
Cc: Zheng Yejian <[email protected]>
Signed-off-by: Steven Rostedt (Google) <[email protected]>

tracing: Fix race where histograms can be called before the event

commit 94eedf3dded5 ("tracing: Fix race where eprobes can be called before
the event") fixed an issue where if an event is soft disabled, and the
trigger is being added, there's a small window where the event sees that
there's a trigger but does not see that it requires reading the event yet,
and then calls the trigger with the record == NULL.

This could be solved with adding memory barriers in the hot path, or to
make sure that all the triggers requiring a record check for NULL. The
latter was chosen.

Commit 94eedf3dded5 set the eprobe trigger handle to check for NULL, but
the same needs to be done with histograms.

Link: https://lore.kernel.org/linux-trace-kernel/[email protected]/
Link: https://lore.kernel.org/linux-trace-kernel/[email protected]
Cc: Tom Zanussi <[email protected]>
Cc: [email protected]
Fixes: 7491e2c442781 ("tracing: Add a probe that attaches to trace events")
Reported-by: Masami Hiramatsu (Google) <[email protected]>
Acked-by: Masami Hiramatsu (Google) <[email protected]>
Signed-off-by: Steven Rostedt (Google) <[email protected]>

docs/bpf: Fix sphinx warnings in BPF map docs

Fix duplicate C declaration warnings when using sphinx >= 3.1.

Reported-by: Akira Yokosawa <[email protected]>
Signed-off-by: Donald Hunter <[email protected]>
Signed-off-by: Daniel Borkmann <[email protected]>
Reviewed-by: Akira Yokosawa <[email protected]>
Link: https://lore.kernel.org/bpf/[email protected]
Link: https://lore.kernel.org/bpf/[email protected]

Merge branch 'kvm-dwmw2-fixes' into HEAD

This brings in a few important fixes for Xen emulation.
While nobody should be enabling it, the bug effectively
allows userspace to read arbitrary memory.

Signed-off-by: Paolo Bonzini <[email protected]>

KVM: Update gfn_to_pfn_cache khva when it moves within the same page

In the case where a GPC is refreshed to a different location within the
same page, we didn't bother to update it. Mostly we don't need to, but
since the ->khva field also includes the offset within the page, that
does have to be updated.

Fixes: 3ba2c95ea180 ("KVM: Do not incorporate page offset into gfn=>pfn cache user address")
Signed-off-by: David Woodhouse <[email protected]>
Reviewed-by: Paul Durrant <[email protected]>
Reviewed-by: Sean Christopherson <[email protected]>
Cc: [email protected]
Signed-off-by: Paolo Bonzini <[email protected]>

KVM: x86/xen: Only do in-kernel acceleration of hypercalls for guest CPL0

There are almost no hypercalls which are valid from CPL > 0, and definitely
none which are handled by the kernel.

Fixes: 2fd6df2f2b47 ("KVM: x86/xen: intercept EVTCHNOP_send from guests")
Reported-by: Michal Luczaj <[email protected]>
Signed-off-by: David Woodhouse <[email protected]>
Reviewed-by: Sean Christopherson <[email protected]>
Cc: [email protected]
Signed-off-by: Paolo Bonzini <[email protected]>

KVM: x86/xen: Validate port number in SCHEDOP_poll

We shouldn't allow guests to poll on arbitrary port numbers off the end
of the event channel table.

Fixes: 1a65105a5aba ("KVM: x86/xen: handle PV spinlocks slowpath")
[dwmw2: my bug though; the original version did check the validity as a
 side-effect of an idr_find() which I ripped out in refactoring.]
Reported-by: Michal Luczaj <[email protected]>
Signed-off-by: David Woodhouse <[email protected]>
Reviewed-by: Sean Christopherson <[email protected]>
Cc: [email protected]
Signed-off-by: Paolo Bonzini <[email protected]>

KVM: x86/mmu: Fix race condition in direct_page_fault

make_mmu_pages_available() must be called with mmu_lock held for write.
However, if the TDP MMU is used, it will be called with mmu_lock held for
read.
This function does nothing unless shadow pages are used, so there is no
race unless nested TDP is used.
Since nested TDP uses shadow pages, old shadow pages may be zapped by this
function even when the TDP MMU is enabled.
Since shadow pages are never allocated by kvm_tdp_mmu_map(), a race
condition can be avoided by not calling make_mmu_pages_available() if the
TDP MMU is currently in use.

I encountered this when repeatedly starting and stopping nested VM.
It can be artificially caused by allocating a large number of nested TDP
SPTEs.

For example, the following BUG and general protection fault are caused in
the host kernel.

pte_list_remove: 00000000cd54fc10 many->many
------------[ cut here ]------------
kernel BUG at arch/x86/kvm/mmu/mmu.c:963!
invalid opcode: 0000 [#1] PREEMPT SMP NOPTI
RIP: 0010:pte_list_remove.cold+0x16/0x48 [kvm]
Call Trace:
 <TASK>
 drop_spte+0xe0/0x180 [kvm]
 mmu_page_zap_pte+0x4f/0x140 [kvm]
 __kvm_mmu_prepare_zap_page+0x62/0x3e0 [kvm]
 kvm_mmu_zap_oldest_mmu_pages+0x7d/0xf0 [kvm]
 direct_page_fault+0x3cb/0x9b0 [kvm]
 kvm_tdp_page_fault+0x2c/0xa0 [kvm]
 kvm_mmu_page_fault+0x207/0x930 [kvm]
 npf_interception+0x47/0xb0 [kvm_amd]
 svm_invoke_exit_handler+0x13c/0x1a0 [kvm_amd]
 svm_handle_exit+0xfc/0x2c0 [kvm_amd]
 kvm_arch_vcpu_ioctl_run+0xa79/0x1780 [kvm]
 kvm_vcpu_ioctl+0x29b/0x6f0 [kvm]
 __x64_sys_ioctl+0x95/0xd0
 do_syscall_64+0x5c/0x90

general protection fault, probably for non-canonical address
0xdead000000000122: 0000 [#1] PREEMPT SMP NOPTI
RIP: 0010:kvm_mmu_commit_zap_page.part.0+0x4b/0xe0 [kvm]
Call Trace:
 <TASK>
 kvm_mmu_zap_oldest_mmu_pages+0xae/0xf0 [kvm]
 direct_page_fault+0x3cb/0x9b0 [kvm]
 kvm_tdp_page_fault+0x2c/0xa0 [kvm]
 kvm_mmu_page_fault+0x207/0x930 [kvm]
 npf_interception+0x47/0xb0 [kvm_amd]

CVE: CVE-2022-45869
Fixes: a2855afc7ee8 ("KVM: x86/mmu: Allow parallel page faults for the TDP MMU")
Signed-off-by: Kazuki Takiguchi <[email protected]>
Cc: [email protected]
Signed-off-by: Paolo Bonzini <[email protected]>

selftests/bpf: Add reproducer for decl_tag in func_proto argument

It should trigger a WARN_ON_ONCE in btf_type_id_size:

  RIP: 0010:btf_type_id_size+0x8bd/0x940 kernel/bpf/btf.c:1952
  btf_func_proto_check kernel/bpf/btf.c:4506 [inline]
  btf_check_all_types kernel/bpf/btf.c:4734 [inline]
  btf_parse_type_sec+0x1175/0x1980 kernel/bpf/btf.c:4763
  btf_parse kernel/bpf/btf.c:5042 [inline]
  btf_new_fd+0x65a/0xb00 kernel/bpf/btf.c:6709
  bpf_btf_load+0x6f/0x90 kernel/bpf/syscall.c:4342
  __sys_bpf+0x50a/0x6c0 kernel/bpf/syscall.c:5034
  __do_sys_bpf kernel/bpf/syscall.c:5093 [inline]
  __se_sys_bpf kernel/bpf/syscall.c:5091 [inline]
  __x64_sys_bpf+0x7c/0x90 kernel/bpf/syscall.c:5091
  do_syscall_64+0x54/0x70 arch/x86/entry/common.c:48

Signed-off-by: Stanislav Fomichev <[email protected]>
Signed-off-by: Daniel Borkmann <[email protected]>
Acked-by: Yonghong Song <[email protected]>
Link: https://lore.kernel.org/bpf/[email protected]

bpf: Prevent decl_tag from being referenced in func_proto arg

Syzkaller managed to hit another decl_tag issue:

  btf_func_proto_check kernel/bpf/btf.c:4506 [inline]
  btf_check_all_types kernel/bpf/btf.c:4734 [inline]
  btf_parse_type_sec+0x1175/0x1980 kernel/bpf/btf.c:4763
  btf_parse kernel/bpf/btf.c:5042 [inline]
  btf_new_fd+0x65a/0xb00 kernel/bpf/btf.c:6709
  bpf_btf_load+0x6f/0x90 kernel/bpf/syscall.c:4342
  __sys_bpf+0x50a/0x6c0 kernel/bpf/syscall.c:5034
  __do_sys_bpf kernel/bpf/syscall.c:5093 [inline]
  __se_sys_bpf kernel/bpf/syscall.c:5091 [inline]
  __x64_sys_bpf+0x7c/0x90 kernel/bpf/syscall.c:5091
  do_syscall_64+0x54/0x70 arch/x86/entry/common.c:48

This seems similar to commit ea68376c8bed ("bpf: prevent decl_tag from being
referenced in func_proto") but for the argument.

Reported-by: [email protected]
Signed-off-by: Stanislav Fomichev <[email protected]>
Signed-off-by: Daniel Borkmann <[email protected]>
Acked-by: Yonghong Song <[email protected]>
Link: https://lore.kernel.org/bpf/[email protected]

Merge tag 'pci-v6.1-fixes-3' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci

Pull pci fixes from Bjorn Helgaas:

 - Update MAINTAINERS to add Manivannan Sadhasivam as Qcom PCIe RC
   maintainer (replacing Stanimir Varbanov) and include DT PCI bindings
   in the "PCI native host bridge and endpoint drivers" entry.

* tag 'pci-v6.1-fixes-3' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
  MAINTAINERS: Include PCI bindings in host bridge entry
  MAINTAINERS: Add Manivannan Sadhasivam as Qcom PCIe RC maintainer

docs/bpf: Document BPF_MAP_TYPE_BLOOM_FILTER

Add documentation for BPF_MAP_TYPE_BLOOM_FILTER including kernel
BPF helper usage, userspace usage and examples.

Signed-off-by: Donald Hunter <[email protected]>
Signed-off-by: Daniel Borkmann <[email protected]>
Acked-by: Joanne Koong <[email protected]>
Link: https://lore.kernel.org/bpf/[email protected]

docs/bpf: Fix sphinx warnings for devmap

Sphinx version >=3.1 warns about duplicate function declarations in the
DEVMAP documentation. This is because the function name is the same for
kernel and user space BPF progs but the parameters and return types
they take is what differs. This patch moves from using the ``c:function::``
directive to using the ``code-block:: c`` directive. The patches also fix
the indentation for the text associated with the "new" code block delcarations.
The missing support of c:namespace-push:: and c:namespace-pop:: directives by
helper scripts for kernel documentation prevents using the ``c:function::``
directive with proper namespacing.

Signed-off-by: Maryam Tahhan <[email protected]>
Signed-off-by: Daniel Borkmann <[email protected]>
Link: https://lore.kernel.org/bpf/[email protected]

docs/bpf: Fix sphinx warnings for cpumap

Sphinx version >=3.1 warns about duplicate function declarations in the
CPUMAP documentation. This is because the function name is the same for
kernel and user space BPF progs but the parameters and return types
they take is what differs. This patch moves from using the ``c:function::``
directive to using the ``code-block:: c`` directive. The patches also fix
the indentation for the text associated with the "new" code block delcarations.
The missing support of c:namespace-push:: and c:namespace-pop:: directives by
helper scripts for kernel documentation prevents using the ``c:function::``
directive with proper namespacing.

Signed-off-by: Maryam Tahhan <[email protected]>
Signed-off-by: Daniel Borkmann <[email protected]>
Link: https://lore.kernel.org/bpf/[email protected]

docs/bpf: Add table of BPF program types to libbpf docs

Extend the libbpf documentation with a table of program types,
attach points and ELF section names.

Signed-off-by: Donald Hunter <[email protected]>
Signed-off-by: Andrii Nakryiko <[email protected]>
Reviewed-by: Bagas Sanjaya <[email protected]>
Acked-by: David Vernet <[email protected]>
Link: https://lore.kernel.org/bpf/[email protected]

bpf: Fix a BTF_ID_LIST bug with CONFIG_DEBUG_INFO_BTF not set

With CONFIG_DEBUG_INFO_BTF not set, we hit the following compilation error,
  /.../kernel/bpf/verifier.c:8196:23: error: array index 6 is past the end of the array
  (that has type 'u32[5]' (aka 'unsigned int[5]')) [-Werror,-Warray-bounds]
        if (meta->func_id == special_kfunc_list[KF_bpf_cast_to_kern_ctx])
                             ^                  ~~~~~~~~~~~~~~~~~~~~~~~
  /.../kernel/bpf/verifier.c:8174:1: note: array 'special_kfunc_list' declared here
  BTF_ID_LIST(special_kfunc_list)
  ^
  /.../include/linux/btf_ids.h:207:27: note: expanded from macro 'BTF_ID_LIST'
  #define BTF_ID_LIST(name) static u32 __maybe_unused name[5];
                            ^
  /.../kernel/bpf/verifier.c:8443:19: error: array index 5 is past the end of the array
  (that has type 'u32[5]' (aka 'unsigned int[5]')) [-Werror,-Warray-bounds]
                 btf_id == special_kfunc_list[KF_bpf_list_pop_back];
                           ^                  ~~~~~~~~~~~~~~~~~~~~
  /.../kernel/bpf/verifier.c:8174:1: note: array 'special_kfunc_list' declared here
  BTF_ID_LIST(special_kfunc_list)
  ^
  /.../include/linux/btf_ids.h:207:27: note: expanded from macro 'BTF_ID_LIST'
  #define BTF_ID_LIST(name) static u32 __maybe_unused name[5];
  ...

Fix the problem by increase the size of BTF_ID_LIST to 16 to avoid compilation error
and also prevent potentially unintended issue due to out-of-bound access.

Reported-by: kernel test robot <[email protected]>
Reported-by: Dan Carpenter <[email protected]>
Reported-by: Nathan Chancellor <[email protected]>
Signed-off-by: Yonghong Song <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Alexei Starovoitov <[email protected]>

Merge branch 'net-complete-conversion-to-i2c_probe_new'

Jakub Kicinski says:

====================
net: Complete conversion to i2c_probe_new

Reposting for Uwe the networking slice of his mega-series:
https://lore.kernel.org/all/20221118224540[email protected]/
so that our build bot can confirm the obvious.

fix mlx5 -> mlxsw while at it.
====================

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>

nfc: st21nfca: i2c: Convert to i2c's .probe_new()

The probe function doesn't make use of the i2c_device_id * parameter so it
can be trivially converted.

Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

nfc: st-nci: Convert to i2c's .probe_new()

The probe function doesn't make use of the i2c_device_id * parameter so it
can be trivially converted.

Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

nfc: s3fwrn5: Convert to i2c's .probe_new()

The probe function doesn't make use of the i2c_device_id * parameter so it
can be trivially converted.

Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

nfc: pn544: Convert to i2c's .probe_new()

The probe function doesn't make use of the i2c_device_id * parameter so it
can be trivially converted.

Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

nfc: pn533: Convert to i2c's .probe_new()

The probe function doesn't make use of the i2c_device_id * parameter so it
can be trivially converted.

Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

NFC: nxp-nci: Convert to i2c's .probe_new()

The probe function doesn't make use of the i2c_device_id * parameter so it
can be trivially converted.

Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

nfc: mrvl: Convert to i2c's .probe_new()

The probe function doesn't make use of the i2c_device_id * parameter so it
can be trivially converted.

Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

nfc: microread: Convert to i2c's .probe_new()

The probe function doesn't make use of the i2c_device_id * parameter so it
can be trivially converted.

Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

net/mlxsw: Convert to i2c's .probe_new()

.probe_new() doesn't get the i2c_device_id * parameter, so determine
that explicitly in the probe function.

Signed-off-by: Uwe Kleine-König <[email protected]>
Reviewed-by: Ido Schimmel <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

net: dsa: xrs700x: Convert to i2c's .probe_new()

The probe function doesn't make use of the i2c_device_id * parameter so it
can be trivially converted.

Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

net: dsa: microchip: ksz9477: Convert to i2c's .probe_new()

The probe function doesn't make use of the i2c_device_id * parameter so it
can be trivially converted.

Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

net: dsa: lan9303: Convert to i2c's .probe_new()

The probe function doesn't make use of the i2c_device_id * parameter so it
can be trivially converted.

Signed-off-by: Uwe Kleine-König <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>

selftests/bpf: Mount debugfs in setns_by_fd

Jiri reports broken test_progs after recent commit 68f8e3d4b916
("selftests/bpf: Make sure zero-len skbs aren't redirectable").
Apparently we don't remount debugfs when we switch back networking namespace.
Let's explicitly mount /sys/kernel/debug.

0: https://lore.kernel.org/bpf/63b85917-a2ea-8e35-620c-808560910819@meta.com/T/#ma66ca9c92e99eee0a25e40f422489b26ee0171c1

Fixes: a30338840fa5 ("selftests/bpf: Move open_netns() and close_netns() into network_helpers.c")
Reported-by: Jiri Olsa <[email protected]>
Signed-off-by: Stanislav Fomichev <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Alexei Starovoitov <[email protected]>

bpf: Don't use idx variable when registering kfunc dtors

In commit fda01efc6160 ("bpf: Enable cgroups to be used as kptrs"), I
added an 'int idx' variable to kfunc_init() which was meant to
dynamically set the index of the btf id entries of the
'generic_dtor_ids' array. This was done to make the code slightly less
brittle as the struct cgroup * kptr kfuncs such as bpf_cgroup_aquire()
are compiled out if CONFIG_CGROUPS is not defined. This, however, causes
an lkp build warning:

>> kernel/bpf/helpers.c:2005:40: warning: multiple unsequenced
   modifications to 'idx' [-Wunsequenced]
.btf_id       = generic_dtor_ids[idx++],

Fix the warning by just hard-coding the indices.

Fixes: fda01efc6160 ("bpf: Enable cgroups to be used as kptrs")
Reported-by: kernel test robot <[email protected]>
Signed-off-by: David Vernet <[email protected]>
Acked-by: Yonghong Song <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Alexei Starovoitov <[email protected]>

NFSD: Fix reads with a non-zero offset that don't end on a page boundary

This was found when virtual machines with nfs-mounted qcow2 disks
failed to boot properly.

Reported-by: Anders Blomdell <[email protected]>
Suggested-by: Al Viro <[email protected]>
Link: https://bugzilla.redhat.com/show_bug.cgi?id=2142132
Fixes: bfbfb6182ad1 ("nfsd_splice_actor(): handle compound pages")
Signed-off-by: Chuck Lever <[email protected]>

Merge tag 'spi-fix-v6.1-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi

Pull spi fixes from Mark Brown:
 "A few fixes, all device specific.

  The most important ones are for the i.MX driver which had a couple of
  nasty data corruption inducing errors appear after the change to
  support PIO mode in the last merge window (one introduced by the
  change and one latent one which the PIO changes exposed).

  Thanks to Frieder, Fabio, Marc and Marek for jumping on that and
  resolving the issues quickly once they were found"

* tag 'spi-fix-v6.1-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
  spi: spi-imx: spi_imx_transfer_one(): check for DMA transfer first
  spi: tegra210-quad: Fix duplicate resource error
  spi: dw-dma: decrease reference count in dw_spi_dma_init_mfld()
  spi: spi-imx: Fix spi_bus_clk if requested clock is higher than input clock
  spi: mediatek: Fix DEVAPC Violation at KO Remove

Merge tag '9p-for-6.1-rc7' of https://github.com/martinetd/linux

Pull 9p fixes from Dominique Martinet:

 - 9p now uses a variable size for its recv buffer, but every place
   hadn't been updated properly to use it and some buffer overflows have
   been found and needed fixing.

   There's still one place where msize is incorrectly used in a safety
   check (p9_check_errors), but all paths leading to it should already
   be avoiding overflows and that patch took a bit more time to get
   right for zero-copy requests so I'll send it for 6.2

 - yet another race condition in p9_conn_cancel introduced by a fix for
   a syzbot report in the same place. Maybe at some point we'll get it
   right without burning it all down...

* tag '9p-for-6.1-rc7' of https://github.com/martinetd/linux:
  9p/xen: check logical size for buffer size
  9p/fd: Use P9_HDRSZ for header size
  9p/fd: Fix write overflow in p9_read_work
  9p/fd: fix issue of list_del corruption in p9_fd_cancel()

lib/vdso: use "grep -E" instead of "egrep"

The latest version of grep claims the egrep is now obsolete so the build
now contains warnings that look like:
egrep: warning: egrep is obsolescent; using grep -E
fix this up by moving the vdso Makefile to use "grep -E" instead.

Cc: Andy Lutomirski <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Reviewed-by: Vincenzo Frascino <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Greg Kroah-Hartman <[email protected]>

fscache: fix OOB Read in __fscache_acquire_volume

The type of a->key[0] is char in fscache_volume_same().  If the length
of cache volume key is greater than 127, the value of a->key[0] is less
than 0.  In this case, klen becomes much larger than 255 after type
conversion, because the type of klen is size_t.  As a result, memcmp()
is read out of bounds.

This causes a slab-out-of-bounds Read in __fscache_acquire_volume(), as
reported by Syzbot.

Fix this by changing the type of the stored key to "u8 *" rather than
"char *" (it isn't a simple string anyway).  Also put in a check that
the volume name doesn't exceed NAME_MAX.

  BUG: KASAN: slab-out-of-bounds in memcmp+0x16f/0x1c0 lib/string.c:757
  Read of size 8 at addr ffff888016f3aa90 by task syz-executor344/3613
  Call Trace:
   memcmp+0x16f/0x1c0 lib/string.c:757
   memcmp include/linux/fortify-string.h:420 [inline]
   fscache_volume_same fs/fscache/volume.c:133 [inline]
   fscache_hash_volume fs/fscache/volume.c:171 [inline]
   __fscache_acquire_volume+0x76c/0x1080 fs/fscache/volume.c:328
   fscache_acquire_volume include/linux/fscache.h:204 [inline]
   v9fs_cache_session_get_cookie+0x143/0x240 fs/9p/cache.c:34
   v9fs_session_init+0x1166/0x1810 fs/9p/v9fs.c:473
   v9fs_mount+0xba/0xc90 fs/9p/vfs_super.c:126
   legacy_get_tree+0x105/0x220 fs/fs_context.c:610
   vfs_get_tree+0x89/0x2f0 fs/super.c:1530
   do_new_mount fs/namespace.c:3040 [inline]
   path_mount+0x1326/0x1e20 fs/namespace.c:3370
   do_mount fs/namespace.c:3383 [inline]
   __do_sys_mount fs/namespace.c:3591 [inline]
   __se_sys_mount fs/namespace.c:3568 [inline]
   __x64_sys_mount+0x27f/0x300 fs/namespace.c:3568

Fixes: 62ab63352350 ("fscache: Implement volume registration")
Reported-by: [email protected]
Signed-off-by: David Howells <[email protected]>
Reviewed-by: Zhang Peng <[email protected]>
Reviewed-by: Jingbo Xu <[email protected]>
cc: Dominique Martinet <[email protected]>
cc: Jeff Layton <[email protected]>
cc: [email protected]
cc: [email protected]
Link: https://lore.kernel.org/r/[email protected]/
Link: https://lore.kernel.org/r/[email protected]/
Link: https://lore.kernel.org/r/166869954095.3793579.8500020902371015443.stgit@warthog.procyon.org.uk/
Signed-off-by: Linus Torvalds <[email protected]>

s390/dasd: fix possible buffer overflow in copy_pair_show

dasd_copy_relation->entry[] array might be accessed out of bounds if the
loop does not break.

Fixes: a91ff09d39f9 ("s390/dasd: add copy pair setup")
Signed-off-by: Stefan Haberland <[email protected]>
Reviewed-by: Jan Hoeppner <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>

s390/dasd: fix no record found for raw_track_access

For DASD devices in raw_track_access mode only full track images are
read and written.
For this purpose it is not necessary to do search operation in the
locate record extended function. The documentation even states that
this might fail if the searched record is not found on a track.

Currently the driver sets a value of 1 in the search field for the first
record after record zero. This is the default for disks not in
raw_track_access mode but record 1 might be missing on a completely
empty track.

There has not been any problem with this on IBM storage servers but it
might lead to errors with DASD devices on other vendors storage servers.

Fix this by setting the search field to 0. Record zero is always available
even on a completely empty track.

Fixes: e4dbb0f2b5dd ("[S390] dasd: Add support for raw ECKD access.")
Signed-off-by: Stefan Haberland <[email protected]>
Reviewed-by: Jan Hoeppner <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>