Commit graph

1,385,869 commits

Author SHA1 Message Date
Jakub Kicinski
0a65dcf624 net: designate queue counts as "double ops protected" by instance lock
Drivers which opt into instance lock protection of ops should
only call set_real_num_*_queues() under the instance lock.
This means that queue counts are double protected (writes
are under both rtnl_lock and instance lock, readers under
either).

Some readers may still be under the rtnl_lock, however, so for
now we need double protection of writers.

OTOH queue API paths are only under the protection of the instance
lock, so we need to validate that the instance is actually locking
ops, otherwise the input checks we do against queue count are racy.

Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20250324224537.248800-6-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-03-25 10:06:49 -07:00
Jakub Kicinski
4b702f8b72 net: explain "protection types" for the instance lock
Try to define some terminology for which fields are protected
by which lock and how. Some fields are protected by both rtnl_lock
and instance lock which is hard to talk about without having
a "key phrase" to refer to a particular protection scheme.

"ops protected" fields are defined later in the series, one by one.

Add ASSERT_RTNL() to netdev_ops_assert_locked() for drivers
not other instance protection of ops. Hopefully it's not too
confusion that netdev_lock_ops() does not match the lock which
netdev_ops_assert_locked() will assert, exactly. The noun "ops"
is in a different place in the name, so I think it's acceptable...

Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20250324224537.248800-5-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-03-25 10:06:44 -07:00
Jakub Kicinski
e2f81e8f4d net: constify dev pointer in misc instance lock helpers
lockdep asserts and predicates can operate on const pointers.
In the future this will let us add asserts in functions
which operate on const pointers like dev_get_min_mp_channel_count().

Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20250324224537.248800-4-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-03-25 10:04:49 -07:00
Jakub Kicinski
bae2da8261 net: remove netif_set_real_num_rx_queues() helper for when SYSFS=n
Since commit a953be53ce ("net-sysfs: add support for device-specific
rx queue sysfs attributes"), so for at least a decade now it is safe
to call net_rx_queue_update_kobjects() when SYSFS=n. That function
does its own ifdef-inery and will return 0. Remove the unnecessary
stub for netif_set_real_num_rx_queues().

Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20250324224537.248800-3-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-03-25 10:04:49 -07:00
Jakub Kicinski
ba6f418fbf net: bubble up taking netdev instance lock to callers of net_devmem_unbind_dmabuf()
A recent commit added taking the netdev instance lock
in netdev_nl_bind_rx_doit(), but didn't remove it in
net_devmem_unbind_dmabuf() which it calls from an error path.
Always expect the callers of net_devmem_unbind_dmabuf() to
hold the lock. This is consistent with net_devmem_bind_dmabuf().

(Not so) coincidentally this also protects mp_param with the instance
lock, which the rest of this series needs.

Fixes: 1d22d3060b ("net: drop rtnl_lock for queue_mgmt operations")
Reviewed-by: Mina Almasry <almasrymina@google.com>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20250324224537.248800-2-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-03-25 10:04:49 -07:00
Linus Torvalds
0ae2062ee3 A single update for futexes:
Use a precomputed mask for the hash computation instead of computing the
   mask from the size on every invocation.
 -----BEGIN PGP SIGNATURE-----
 
 iQJHBAABCgAxFiEEQp8+kY+LLUocC4bMphj1TA10mKEFAmff5PETHHRnbHhAbGlu
 dXRyb25peC5kZQAKCRCmGPVMDXSYoTCJEADEYDoQGkMCwUoTDJ4xVLohb89lH4KN
 cB38tF0YtHDMoBk2AXzHvDJioqPr5+YRQn3cYVpqQ6ajV0fVay8XDIdbD0GrhSUk
 KSiz6sfIVT8c18bvgK/YrnQL0B5AytY5X+hYd3A5tS7NHlKNs3nHuZetNFrklDUM
 RI7dY2dOU0jjzEB1WFMSIU2gIt/B9994fd+uhsSfKhJgVojOO5VpuREzMRb5nTYr
 sZRvel9JqfvADz1Wr9DpX5JGQ65m1uEXFK8bbdYemJmqqKbdQt0pEWWLC5PJkhNJ
 z9vih+TPJqx7THNgind+KNNO7Yf3V1aKiks+GKVOn0WplhybmlhqqYdfSgzFZOpR
 BymsxPQ9AQRf7O+j23f5Ys1Ty/1mc+ZcKYPCmasoG3MngmMt6nS+7nwiK1lnZEYQ
 hd7EpFKl8pPe3Ga5Cp6iFj64vIna5eMGQ0vFL7DqCY8mnzDnqhZBMk6xrJ/QEgXj
 rpLYKwDXz34EwOKSuWtFQcds4HUDeOxdJQCSbmfpQo4wh08HmcNK+lr3PXub6a7y
 ncEFOD4rmRLXfofQyF/m4MKwv5I/nwkSNDQh/hLNfHrmWxGZPE8J9vFJccQcAsf6
 2vL5G0cNTKjNA5Afsi8bpi8Khlo9hJTW18Sw7DegpEwJxJjww723X+3Ati8qTjtr
 clVhRAfjHPp5WQ==
 =GyIS
 -----END PGP SIGNATURE-----

Merge tag 'locking-futex-2025-03-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull futex update from Thomas Gleixner:
 "A single update for futexes:

  Use a precomputed mask for the hash computation instead of computing
  the mask from the size on every invocation"

* tag 'locking-futex-2025-03-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  futex: Use a hashmask instead of hashsize
2025-03-25 10:00:17 -07:00
Linus Torvalds
0f40464674 Updates for interrupt chip drivers:
- Support for hard indices on RISC-V. The hart index identifies a hart
     (core) within a specific interrupt domain in RISC-V's Priviledged
     Architecture.
 
   - Rework of the RISC-V MSI driver.
 
     This moves the driver over to the generic MSI library and solves the
     affinity problem of unmaskable PCI/MSI controllers. Unmaskable PCI/MSI
     controllers are prone to lose interrupts when the MSI message is
     updated to change the affinity because the message write consists of
     three 32-bit subsequent writes, which update address and data. As these
     writes are non-atomic versus the device raising an interrupt, the
     device can observe a half written update and issue an interrupt on the
     wrong vector. This is mitiated by a carefully orchestrated step by step
     update and the observation of an eventually pending interrupt on the
     CPU which issues the update. The algorithm follows the well established
     method of the X86 MSI driver.
 
   - A new driver for the RISC-V Sophgo SG2042 MSI controller
 
   - Overhaul of the Renesas RZQ2L driver.
 
     Simplification of the probe function by using devm_*() mechanisms,
     which avoid the endless list of error prone gotos in the failure paths.
 
   - Expand the Renesas RZV2H driver to support RZ/G3E SoCs
 
   - A workaround for Rockchip 3568002 erratum in the GIC-V3 driver to
     ensure that the addressing is limited to the lower 32-bit of the
     physical address space.
 
   - Add support for the Allwinner AS23 NMI controller
 
   - Expand the IMX irqsteer driver to handle up to 960 input interrupts
 
   - The usual small updates, cleanups and device tree changes.
 -----BEGIN PGP SIGNATURE-----
 
 iQJHBAABCgAxFiEEQp8+kY+LLUocC4bMphj1TA10mKEFAmff454THHRnbHhAbGlu
 dXRyb25peC5kZQAKCRCmGPVMDXSYoZqoD/4kdHzbxfLpf7vC3NnG8NWwTq5FpbSx
 6grQC9hWNMAs4n2IFjJRFLrjeX3AcdAQXL/BWuM0LfW9tQDQaVmqlSIlB/bn69KB
 7HyAR6ozbOgnHKGAqFUXSLf+4pq+6q3mOgGKIF289dy14HFu4ta0DqKgkPZeQnVs
 R/J8i7REUnn+YuxzSt5eOqyDPyt2EHJosSUABSWQZBlrM9jy1W7f6NqDFwawiVsa
 +tv4U/bz91vjzVxwTIgt7nJK+b2HVYdxoZYuKJwPaTsj26ANPp6ltjRTeOmZhb5h
 uKgw+OyzDnk6q+tjGcRqrqwl291VKxCvnRiqHFfu3CERdmI9qvpN9IRcEJqIbkcN
 cakekhAyt7OO7sEPcql5vBL97e9hpb7EcH78gYxwHf8Dy0rFZUvSC5v+L6VRFnJS
 XcKA1L+f9B6u5qxnBtLan9IW08HYNdvmPq6AuVjk+ndKioPUFqB2q6AtXpuA3Rmu
 Y3XH/wh/q5wk0pgeByxQW6swsfpMN3OYK3mpLx475wFh2NKzcdGlwGhDFhiw8DKX
 m1AESy3UZatj1a0qGaFS/M+mm9KGrDYIMrje832Wf4Yf1LGmTsDkd3/V99oazSsq
 Jm4qhDASXChJXd0imQICX9hPw0aHTlLYNs54obUXVULH4HivQKIgWhUXrjG0dBDL
 +tttjuv5FJxr3A==
 =jPHa
 -----END PGP SIGNATURE-----

Merge tag 'irq-drivers-2025-03-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull irq driver updates from Thomas Gleixner:

 - Support for hard indices on RISC-V. The hart index identifies a hart
   (core) within a specific interrupt domain in RISC-V's Priviledged
   Architecture.

 - Rework of the RISC-V MSI driver

   This moves the driver over to the generic MSI library and solves the
   affinity problem of unmaskable PCI/MSI controllers. Unmaskable
   PCI/MSI controllers are prone to lose interrupts when the MSI message
   is updated to change the affinity because the message write consists
   of three 32-bit subsequent writes, which update address and data. As
   these writes are non-atomic versus the device raising an interrupt,
   the device can observe a half written update and issue an interrupt
   on the wrong vector. This is mitiated by a carefully orchestrated
   step by step update and the observation of an eventually pending
   interrupt on the CPU which issues the update. The algorithm follows
   the well established method of the X86 MSI driver.

 - A new driver for the RISC-V Sophgo SG2042 MSI controller

 - Overhaul of the Renesas RZQ2L driver

   Simplification of the probe function by using devm_*() mechanisms,
   which avoid the endless list of error prone gotos in the failure
   paths.

 - Expand the Renesas RZV2H driver to support RZ/G3E SoCs

 - A workaround for Rockchip 3568002 erratum in the GIC-V3 driver to
   ensure that the addressing is limited to the lower 32-bit of the
   physical address space.

 - Add support for the Allwinner AS23 NMI controller

 - Expand the IMX irqsteer driver to handle up to 960 input interrupts

 - The usual small updates, cleanups and device tree changes

* tag 'irq-drivers-2025-03-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (40 commits)
  irqchip/imx-irqsteer: Support up to 960 input interrupts
  irqchip/sunxi-nmi: Support Allwinner A523 NMI controller
  dt-bindings: irq: sun7i-nmi: Document the Allwinner A523 NMI controller
  irqchip/davinci-cp-intc: Remove public header
  irqchip/renesas-rzv2h: Add RZ/G3E support
  irqchip/renesas-rzv2h: Update macros ICU_TSSR_TSSEL_{MASK,PREP}
  irqchip/renesas-rzv2h: Update TSSR_TIEN macro
  irqchip/renesas-rzv2h: Add field_width to struct rzv2h_hw_info
  irqchip/renesas-rzv2h: Add max_tssel to struct rzv2h_hw_info
  irqchip/renesas-rzv2h: Add struct rzv2h_hw_info with t_offs variable
  irqchip/renesas-rzv2h: Use devm_pm_runtime_enable()
  irqchip/renesas-rzv2h: Use devm_reset_control_get_exclusive_deasserted()
  irqchip/renesas-rzv2h: Simplify rzv2h_icu_init()
  irqchip/renesas-rzv2h: Drop irqchip from struct rzv2h_icu_priv
  irqchip/renesas-rzv2h: Fix wrong variable usage in rzv2h_tint_set_type()
  dt-bindings: interrupt-controller: renesas,rzv2h-icu: Document RZ/G3E SoC
  riscv: sophgo: dts: Add msi controller for SG2042
  irqchip: Add the Sophgo SG2042 MSI interrupt controller
  dt-bindings: interrupt-controller: Add Sophgo SG2042 MSI
  arm64: dts: rockchip: rk356x: Move PCIe MSI to use GIC ITS instead of MBI
  ...
2025-03-25 09:54:36 -07:00
Kiran K
3b5715aeb8 t blameBluetooth: btintel: Fix leading white space
Remove the unwanted leading whitespace.

Fixes: 6ed8304738 ("Bluetooth: btintel_pcie: Setup buffers for firmware traces")
Fixes: bb3569ac36 ("Bluetooth: btintel: Add DSBR support for ScP")
Signed-off-by: Kiran K <kiran.k@intel.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:52:00 -04:00
Vijay Satija
e036afb1fe Bluetooth: btintel: Add support to configure TX power
BRDS - Bluetooth Regulatory Domain Specific absorption rate

Bluetooth has regulatory limitations which prohibit or allow usage of certain
bands or channels as well as limiting Tx power. The Tx power values can be
configured in ACPI table. This patch reads from ACPI entry configures the
controller accordingly.

Signed-off-by: Kiran K <kiran.k@intel.com>
Signed-off-by: Vijay Satija <vijay.satija@intel.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:51:42 -04:00
Sean Wang
6ac4233afb Bluetooth: btmtksdio: Prevent enabling interrupts after IRQ handler removal
Ensure interrupts are not re-enabled when the IRQ handler has already been
removed. This prevents unexpected IRQ handler execution due to stale or
unhandled interrupts.

Modify btmtksdio_txrx_work to check if bdev->func->irq_handler exists
before calling sdio_writel to enable interrupts.

Co-developed-by: Pedro Tsai <pedro.tsai@mediatek.com>
Signed-off-by: Pedro Tsai <pedro.tsai@mediatek.com>
Co-developed-by: Felix Freimann <felix.freimann@mediatek.com>
Signed-off-by: Felix Freimann <felix.freimann@mediatek.com>
Signed-off-by: Sean Wang <sean.wang@mediatek.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:51:26 -04:00
Hao Qin
33634e2ab7 Bluetooth: btmtk: Remove the resetting step before downloading the fw
Remove the resetting step before downloading the fw, as it may cause
other usb devices to fail to initialise when connected during boot
on kernels 6.11 and newer.

Signed-off-by: Hao Qin <hao.qin@mediatek.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:51:10 -04:00
Pauli Virtanen
bdbcd52871 Bluetooth: SCO: add TX timestamping
Support TX timestamping in SCO sockets.
Not available for hdevs without SCO_FLOWCTL.

Support MSG_ERRQUEUE in SCO recvmsg.

Signed-off-by: Pauli Virtanen <pav@iki.fi>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:50:54 -04:00
Jann Horn
ece69af2ed rwonce: handle KCSAN like KASAN in read_word_at_a_time()
read_word_at_a_time() is allowed to read out of bounds by straddling the
end of an allocation (and the caller is expected to then mask off
out-of-bounds data). This works as long as the caller guarantees that the
access won't hit a pagefault (either by ensuring that addr is aligned or by
explicitly checking where the next page boundary is).

Such out-of-bounds data could include things like KASAN redzones, adjacent
allocations that are concurrently written to, or simply an adjacent struct
field that is concurrently updated. KCSAN should ignore racy reads of OOB
data that is not actually used, just like KASAN, so (similar to the code
above) change read_word_at_a_time() to use __no_sanitize_or_inline instead
of __no_kasan_or_inline, and explicitly inform KCSAN that we're reading
the first byte.

We do have an instrument_read() helper that calls into both KASAN and
KCSAN, but I'm instead open-coding that here to avoid having to pull the
entire instrumented.h header into rwonce.h.

Also, since this read can be racy by design, we should technically do
READ_ONCE(), so add that.

Fixes: dfd402a4c4 ("kcsan: Add Kernel Concurrency Sanitizer infrastructure")
Signed-off-by: Jann Horn <jannh@google.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Marco Elver <elver@google.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2025-03-25 17:50:38 +01:00
Pauli Virtanen
11770f41b8 Bluetooth: L2CAP: add TX timestamping
Support TX timestamping in L2CAP sockets.

Support MSG_ERRQUEUE recvmsg.

For other than SOCK_STREAM L2CAP sockets, if a packet from sendmsg() is
fragmented, only the first ACL fragment is timestamped.

For SOCK_STREAM L2CAP sockets, use the bytestream convention and
timestamp the last fragment and count bytes in tskey.

Timestamps are not generated in the Enhanced Retransmission mode, as
meaning of COMPLETION stamp is unclear if L2CAP layer retransmits.

Signed-off-by: Pauli Virtanen <pav@iki.fi>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:50:35 -04:00
Pauli Virtanen
d415ba2882 Bluetooth: ISO: add TX timestamping
Add BT_SCM_ERROR socket CMSG type.

Support TX timestamping in ISO sockets.

Support MSG_ERRQUEUE in ISO recvmsg.

If a packet from sendmsg() is fragmented, only the first ACL fragment is
timestamped.

Signed-off-by: Pauli Virtanen <pav@iki.fi>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:50:07 -04:00
Pauli Virtanen
134f4b39df Bluetooth: add support for skb TX SND/COMPLETION timestamping
Support enabling TX timestamping for some skbs, and track them until
packet completion. Generate software SCM_TSTAMP_COMPLETION when getting
completion report from hardware.

Generate software SCM_TSTAMP_SND before sending to driver. Sending from
driver requires changes in the driver API, and drivers mostly are going
to send the skb immediately.

Make the default situation with no COMPLETION TX timestamping more
efficient by only counting packets in the queue when there is nothing to
track.  When there is something to track, we need to make clones, since
the driver may modify sent skbs.

The tx_q queue length is bounded by the hdev flow control, which will
not send new packets before it has got completion reports for old ones.

Signed-off-by: Pauli Virtanen <pav@iki.fi>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:49:38 -04:00
Pauli Virtanen
983e0e4e87 net-timestamp: COMPLETION timestamp on packet tx completion
Add SOF_TIMESTAMPING_TX_COMPLETION, for requesting a software timestamp
when hardware reports a packet completed.

Completion tstamp is useful for Bluetooth, as hardware timestamps do not
exist in the HCI specification except for ISO packets, and the hardware
has a queue where packets may wait.  In this case the software SND
timestamp only reflects the kernel-side part of the total latency
(usually small) and queue length (usually 0 unless HW buffers
congested), whereas the completion report time is more informative of
the true latency.

It may also be useful in other cases where HW TX timestamps cannot be
obtained and user wants to estimate an upper bound to when the TX
probably happened.

Signed-off-by: Pauli Virtanen <pav@iki.fi>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:48:05 -04:00
Luiz Augusto von Dentz
b257e02ecc HCI: coredump: Log devcd dumps into the monitor
This logs the devcd dumps with hci_recv_diag so they appear in the
monitor traces with proper timestamps which can then be used to relate
the HCI traffic that caused the dump:

= Vendor Diagnostic (len 174)
        42 6c 75 65 74 6f 6f 74 68 20 64 65 76 63 6f 72  Bluetooth devcor
        65 64 75 6d 70 0a 53 74 61 74 65 3a 20 32 0a 00  edump.State: 2..
        43 6f 6e 74 72 6f 6c 6c 65 72 20 4e 61 6d 65 3a  Controller Name:
        20 76 68 63 69 5f 63 74 72 6c 0a 46 69 72 6d 77   vhci_ctrl.Firmw
        61 72 65 20 56 65 72 73 69 6f 6e 3a 20 76 68 63  are Version: vhc
        69 5f 66 77 0a 44 72 69 76 65 72 3a 20 76 68 63  i_fw.Driver: vhc
        69 5f 64 72 76 0a 56 65 6e 64 6f 72 3a 20 76 68  i_drv.Vendor: vh
        63 69 0a 2d 2d 2d 20 53 74 61 72 74 20 64 75 6d  ci.--- Start dum
        70 20 2d 2d 2d 0a 74 65 73 74 20 64 61 74 61 00  p ---.test data.
        00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
        00 00 00 00 00 00 00 00 00 00 00 00 00 00        ..............

Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:47:49 -04:00
Wentao Guan
e8c00f5433 Bluetooth: HCI: Add definition of hci_rp_remote_name_req_cancel
Return Parameters is not only status, also bdaddr:

BLUETOOTH CORE SPECIFICATION Version 5.4 | Vol 4, Part E
page 1870:
BLUETOOTH CORE SPECIFICATION Version 5.0 | Vol 2, Part E
page 802:

Return parameters:
  Status:
  Size: 1 octet
  BD_ADDR:
  Size: 6 octets

Note that it also fixes the warning:
"Bluetooth: hci0: unexpected cc 0x041a length: 7 > 1"

Fixes: c8992cffbe ("Bluetooth: hci_event: Use of a function table to handle Command Complete")
Signed-off-by: Wentao Guan <guanwentao@uniontech.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:47:23 -04:00
Luiz Augusto von Dentz
5b86e2a575 Bluetooth: hci_vhci: Mark Sync Flow Control as supported
This sets HCI_QUIRK_SYNC_FLOWCTL_SUPPORTED which indicates that
controllers created by vhci driver support Sync Flow Control.

Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:47:07 -04:00
Luiz Augusto von Dentz
1321845352 Bluetooth: hci_core: Enable buffer flow control for SCO/eSCO
This enables buffer flow control for SCO/eSCO
(see: Bluetooth Core 6.0 spec: 6.22. Synchronous Flow Control Enable),
recently this has caused the following problem and is actually a nice
addition for the likes of Socket TX complete:

< HCI Command: Read Buffer Size (0x04|0x0005) plen 0
> HCI Event: Command Complete (0x0e) plen 11
      Read Buffer Size (0x04|0x0005) ncmd 1
        Status: Success (0x00)
        ACL MTU: 1021 ACL max packet: 5
        SCO MTU: 240  SCO max packet: 8
...
< SCO Data TX: Handle 257 flags 0x00 dlen 120
< SCO Data TX: Handle 257 flags 0x00 dlen 120
< SCO Data TX: Handle 257 flags 0x00 dlen 120
< SCO Data TX: Handle 257 flags 0x00 dlen 120
< SCO Data TX: Handle 257 flags 0x00 dlen 120
< SCO Data TX: Handle 257 flags 0x00 dlen 120
< SCO Data TX: Handle 257 flags 0x00 dlen 120
< SCO Data TX: Handle 257 flags 0x00 dlen 120
< SCO Data TX: Handle 257 flags 0x00 dlen 120
> HCI Event: Hardware Error (0x10) plen 1
        Code: 0x0a

To fix the code will now attempt to enable buffer flow control when
HCI_QUIRK_SYNC_FLOWCTL_SUPPORTED is set by the driver:

< HCI Command: Write Sync Fl.. (0x03|0x002f) plen 1
        Flow control: Enabled (0x01)
> HCI Event: Command Complete (0x0e) plen 4
      Write Sync Flow Control Enable (0x03|0x002f) ncmd 1
        Status: Success (0x00)

On success then HCI_SCO_FLOWCTL would be set which indicates sco_cnt
shall be used for flow contro.

Fixes: 7fedd3bb6b ("Bluetooth: Prioritize SCO traffic")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Tested-by: Pauli Virtanen <pav@iki.fi>
2025-03-25 12:46:40 -04:00
Luiz Augusto von Dentz
42c6c7a0cf Bluetooth: btintel_pci: Fix build warning
This fixes the following warning:

drivers/bluetooth/btintel_pcie.c:695:20: warning: unused function 'btintel_pcie_in_rom' [-Wunused-function]
  695 | static inline bool btintel_pcie_in_rom(struct btintel_pcie_data *data)
      |                    ^~~~~~~~~~~~~~~~~~~

Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:46:24 -04:00
Kiran K
15f6f62114 Bluetooth: btintel_pcie: Trigger device coredump on hardware exception
Driver dumps device core dump on firmware exception.

Signed-off-by: Kiran K <kiran.k@intel.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:46:08 -04:00
Neeraj Sanjay Kale
9148ac0a16 Bluetooth: btnxpuart: Add support to set BD address
This adds support for setting BD address during hci registration. NXP
FW does not allow vendor commands unless it receives a reset command
after FW download and initialization done.

As a workaround, the .set_bdaddr callback function will first send the
HCI reset command, followed by the actual vendor command to set BD
address.

The driver checks for the local-bd-address property in device tree, and
if preset, it sets the HCI_QUIRK_USE_BDADDR_PROPERTY quirk.

With this quirk set, the driver's set_bdaddr callback function is called
after FW download is complete and before HCI initialization, which sends
the hci reset and 3f 22 commands. During initialization, kernel reads
the newly set BD address from the controller.

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Signed-off-by: Johan Korsnes <johan.korsnes@remarkable.no>
Signed-off-by: Kristian Krohn <kristian.krohn@remarkable.no>
Tested-by: Neeraj Sanjay Kale <neeraj.sanjaykale@nxp.com>
Signed-off-by: Neeraj Sanjay Kale <neeraj.sanjaykale@nxp.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:45:52 -04:00
Neeraj Sanjay Kale
b13b6d669c dt-bindings: net: bluetooth: nxp: Add support to set BD address
Allow user to set custom BD address for NXP chipsets.

Signed-off-by: Neeraj Sanjay Kale <neeraj.sanjaykale@nxp.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:45:37 -04:00
Neeraj Sanjay Kale
998e447f44 Bluetooth: btnxpuart: Add support for HCI coredump feature
This adds support for Bluetooth Coredump feature to BTNXPUART driver to
collect FW dumps on demand, or in case FW goes in a bad state.

To trigger manual FW dump, following command can be used:
echo 1 > /sys/class/bluetooth/hci0/device/coredump

Once FW dump is complete, it can be written to a file:
cat /sys/class/bluetooth/hci0/devcoredump/data > fw_dump

While FW dump is in progress, any HCI command will return -EBUSY.

After FW dump is complete, driver will give HCI_NXP_IND_RESET command
which soft-resets the chip, allowing FW re-download.

Signed-off-by: Neeraj Sanjay Kale <neeraj.sanjaykale@nxp.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:45:21 -04:00
Neeraj Sanjay Kale
6fca6781d1 Bluetooth: btnxpuart: Move vendor specific initialization to .post_init
This moves change baudrate and power save vendor commands from
nxp_setup() to nxp_post_init().

This also moves the baudrate restore logic from nxp_serdev_remove() to
nxp_shutdown() which ensure baudrate is restored even when HCI dev is
down, preventing baudrate mismatch between host and controller when
device is probed again next time.

In case of removal when the hdev is up and running, we have to call the
shutdown procedure explicitly before unregistering the hdev.

Signed-off-by: Neeraj Sanjay Kale <neeraj.sanjaykale@nxp.com>
Co-developed-by: Loic Poulain <loic.poulain@linaro.org>
Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:45:05 -04:00
Kiran K
07e6bddb54 Bluetooth: btintel_pcie: Add support for device coredump
1. Driver registers device coredump callback
2. Dumps firmware traces as part of coredump

Co-developed-by: Vijay Satija <vijay.satija@intel.com>
Signed-off-by: Vijay Satija <vijay.satija@intel.com>
Signed-off-by: Kiran K <kiran.k@intel.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:44:48 -04:00
Pedro Nishiyama
1f04b0e5e3 Bluetooth: btusb: Fix regression in the initialization of fake Bluetooth controllers
Set HCI_READ_VOICE_SETTING and HCI_READ_PAGE_SCAN_TYPE as broken.

Once the min/max length of the commands began to be asserted, these fake
controllers can no longer be initialized because they return a smaller
report for these commands.

This affects various fake controllers reusing the 0A12:0001 VID/PID.

Fixes: c8992cffbe ("Bluetooth: hci_event: Use of a function table to handle Command Complete")
Signed-off-by: Pedro Nishiyama <nishiyama.pedro@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:44:32 -04:00
Pedro Nishiyama
14d17c78a4 Bluetooth: Disable SCO support if READ_VOICE_SETTING is unsupported/broken
A SCO connection without the proper voice_setting can cause
the controller to lock up.

Signed-off-by: Pedro Nishiyama <nishiyama.pedro@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:44:15 -04:00
Pedro Nishiyama
127881334e Bluetooth: Add quirk for broken READ_PAGE_SCAN_TYPE
Some fake controllers cannot be initialized because they return a smaller
report than expected for READ_PAGE_SCAN_TYPE.

Signed-off-by: Pedro Nishiyama <nishiyama.pedro@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:43:51 -04:00
Pedro Nishiyama
ff26b2dd65 Bluetooth: Add quirk for broken READ_VOICE_SETTING
Some fake controllers cannot be initialized because they return a smaller
report than expected for READ_VOICE_SETTING.

Signed-off-by: Pedro Nishiyama <nishiyama.pedro@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:43:25 -04:00
Kiran K
b9465e6670 Bluetooth: btintel_pcie: Read hardware exception data
On hardware error, controller writes hardware error event and optional
vendor specific hci events in device memory in TLV format and raises
MSIX interrupt. Driver reads the device memory and passes the events to
the stack for further processing.

Co-developed-by: Vijay Satija <vijay.satija@intel.com>
Signed-off-by: Vijay Satija <vijay.satija@intel.com>
Signed-off-by: Kiran K <kiran.k@intel.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:43:08 -04:00
Kiran K
6ed8304738 Bluetooth: btintel_pcie: Setup buffers for firmware traces
This patch allocates the host memory which is used by controller to dump
the firmware traces. The memory needs to be shared with controller via
context information.

Co-developed-by: Vijay Satija <vijay.satija@intel.com>
Signed-off-by: Vijay Satija <vijay.satija@intel.com>
Signed-off-by: Kiran K <kiran.k@intel.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2025-03-25 12:42:51 -04:00
Dmitry Baryshkov
d5712c511c Bluetooth: qca: add WCN3950 support
WCN3950 is another example of the WCN39xx BT/WiFI family of chips. It
requires different firmware files and has different current
requirements, so add it as a separate SoC type.

The firmware for these chips has been recently added to the
linux-firmware repository and will be a part of the upcoming release:
- qca/cmbtfw12.tlv
- qca/cmbtfw13.tlv
- qca/cmnv12.bin
- qca/cmnv13.bin
- qca/cmnv13s.bin
- qca/cmnv13t.bin

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:42:35 -04:00
Dmitry Baryshkov
1cc41b5092 Bluetooth: qca: simplify WCN399x NVM loading
The WCN399x code has two separate cases for loading the NVM data. In
preparation to adding support for WCN3950, which also requires similar
quirk, split the "variant" to be specified explicitly and merge two
snprintfs into a single one.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:42:19 -04:00
Dmitry Baryshkov
d0b391ef3b dt-bindings: net: bluetooth: qualcomm: document WCN3950
WCN3950 is another member of the WiFi/BT WCN39xx family of the chips. It
requires different firmware, so document it as a new compat string.

Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:42:05 -04:00
Loic Poulain
d154b0ff89 dt-bindings: net: bluetooth: nxp: Add wakeup pin properties
NXP bluetooth controller may have GPIO pins used and routed for `WAKE_IN`
and `WAKE_OUT`, such pin info must be known so that the driver is can
configure the controller's firmware accordingly.

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:41:51 -04:00
Loic Poulain
873b6754af bluetooth: btnxpuart: Support for controller wakeup gpio config
When using the out-of-band WAKE_IN and WAKE_OUT pins, we have to tell
the firmware which pins to use (from controller point of view). This
allows to report remote wakeup support when WAKE_OUT(c2h) is configured.

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Reviewed-by: Neeraj Sanjay Kale <neeraj.sanjaykale@nxp.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:41:34 -04:00
Janaki Ramaiah Thota
852cfdc7a5 Bluetooth: hci_qca: use the power sequencer for wcn6750
Older boards are having entry "enable-gpios" in dts, we can safely assume
latest boards which are supporting PMU node enrty will support power
sequencer.

Signed-off-by: Janaki Ramaiah Thota <quic_janathot@quicinc.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:41:19 -04:00
Jiande Lu
a88643b7e4 Bluetooth: btusb: Add 2 HWIDs for MT7922
Add below HWIDs for MediaTek MT7922 USB Bluetooth chip.
VID 0x0489, PID 0xe152
VID 0x0489, PID 0xe153

Patch has been tested successfully and controller is recognized
device pair successfully.

MT7922 module bring up message as below.
Bluetooth: Core ver 2.22
Bluetooth: HCI device and connection manager initialized
Bluetooth: HCI socket layer initialized
Bluetooth: L2CAP socket layer initialized
Bluetooth: SCO socket layer initialized
Bluetooth: hci0: HW/SW Version: 0x008a008a, Build Time: 20241106163512
Bluetooth: hci0: Device setup in 2284925 usecs
Bluetooth: hci0: HCI Enhanced Setup Synchronous Connection command is advertised, but not supported.
Bluetooth: hci0: AOSP extensions version v1.00
Bluetooth: BNEP (Ethernet Emulation) ver 1.3
Bluetooth: BNEP filters: protocol multicast
Bluetooth: BNEP socket layer initialized
Bluetooth: MGMT ver 1.22
Bluetooth: RFCOMM TTY layer initialized
Bluetooth: RFCOMM socket layer initialized
Bluetooth: RFCOMM ver 1.11

Signed-off-by: Jiande Lu <jiande.lu@mediatek.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:41:03 -04:00
Easwar Hariharan
c9d84da18d Bluetooth: L2CAP: convert timeouts to secs_to_jiffies()
Commit b35108a51c ("jiffies: Define secs_to_jiffies()") introduced
secs_to_jiffies().  As the value here is a multiple of 1000, use
secs_to_jiffies() instead of msecs_to_jiffies() for readability.

Signed-off-by: Easwar Hariharan <eahariha@linux.microsoft.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:40:46 -04:00
Easwar Hariharan
3f0a819e8c Bluetooth: SMP: convert timeouts to secs_to_jiffies()
Commit b35108a51c ("jiffies: Define secs_to_jiffies()") introduced
secs_to_jiffies().  As the value here is a multiple of 1000, use
secs_to_jiffies() instead of msecs_to_jiffies() for readability.

Signed-off-by: Easwar Hariharan <eahariha@linux.microsoft.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:40:31 -04:00
Easwar Hariharan
e3e627e6b2 Bluetooth: MGMT: convert timeouts to secs_to_jiffies()
Commit b35108a51c ("jiffies: Define secs_to_jiffies()") introduced
secs_to_jiffies().  As the value here is a multiple of 1000, use
secs_to_jiffies() instead of msecs_to_jiffies to avoid the multiplication.

This is converted using scripts/coccinelle/misc/secs_to_jiffies.cocci with
the following Coccinelle rules:

@depends on patch@
expression E;
@@

-msecs_to_jiffies(E * 1000)
+secs_to_jiffies(E)

-msecs_to_jiffies(E * MSEC_PER_SEC)
+secs_to_jiffies(E)

Signed-off-by: Easwar Hariharan <eahariha@linux.microsoft.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:40:15 -04:00
Easwar Hariharan
a9d5131e22 Bluetooth: hci_vhci: convert timeouts to secs_to_jiffies()
Commit b35108a51c ("jiffies: Define secs_to_jiffies()") introduced
secs_to_jiffies().  As the value here is a multiple of 1000, use
secs_to_jiffies() instead of msecs_to_jiffies to avoid the multiplication.

This is converted using scripts/coccinelle/misc/secs_to_jiffies.cocci with
the following Coccinelle rules:

@depends on patch@
expression E;
@@

-msecs_to_jiffies(E * 1000)
+secs_to_jiffies(E)

-msecs_to_jiffies(E * MSEC_PER_SEC)
+secs_to_jiffies(E)

Signed-off-by: Easwar Hariharan <eahariha@linux.microsoft.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:40:00 -04:00
Arseniy Krasnov
5df5dafc17 Bluetooth: hci_uart: Fix another race during initialization
Do not set 'HCI_UART_PROTO_READY' before call 'hci_uart_register_dev()'.
Possible race is when someone calls 'hci_tty_uart_close()' after this bit
is set, but 'hci_uart_register_dev()' wasn't done. This leads to access
to uninitialized fields. To fix it let's set this bit after device was
registered (as before patch c411c62cc133) and to fix previous problem let's
add one more bit in addition to 'HCI_UART_PROTO_READY' which allows to
perform power up without original bit set (pls see commit c411c62cc133).

Crash backtrace from syzbot report:

RIP: 0010:skb_queue_empty_lockless include/linux/skbuff.h:1887 [inline]
RIP: 0010:skb_queue_purge_reason+0x6d/0x140 net/core/skbuff.c:3936

Call Trace:
 <TASK>
 skb_queue_purge include/linux/skbuff.h:3364 [inline]
 mrvl_close+0x2f/0x90 drivers/bluetooth/hci_mrvl.c:100
 hci_uart_tty_close+0xb6/0x120 drivers/bluetooth/hci_ldisc.c:557
 tty_ldisc_close drivers/tty/tty_ldisc.c:455 [inline]
 tty_ldisc_kill+0x66/0xc0 drivers/tty/tty_ldisc.c:613
 tty_ldisc_release+0xc9/0x120 drivers/tty/tty_ldisc.c:781
 tty_release_struct+0x10/0x80 drivers/tty/tty_io.c:1690
 tty_release+0x4ef/0x640 drivers/tty/tty_io.c:1861
 __fput+0x86/0x2a0 fs/file_table.c:450
 task_work_run+0x82/0xb0 kernel/task_work.c:239
 resume_user_mode_work include/linux/resume_user_mode.h:50 [inline]
 exit_to_user_mode_loop kernel/entry/common.c:114 [inline]
 exit_to_user_mode_prepare include/linux/entry-common.h:329 [inline]
 __syscall_exit_to_user_mode_work kernel/entry/common.c:207 [inline]
 syscall_exit_to_user_mode+0xa3/0x1b0 kernel/entry/common.c:218
 do_syscall_64+0x9a/0x190 arch/x86/entry/common.c:89
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

Signed-off-by: Arseniy Krasnov <avkrasnov@salutedevices.com>
Reported-by: syzbot+683f8cb11b94b1824c77@syzkaller.appspotmail.com
Tested-by: syzbot+683f8cb11b94b1824c77@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/linux-bluetooth/d159c57f-8490-4c26-79da-6ad3612c4a14@salutedevices.com/
Fixes: 366ceff495 ("Bluetooth: hci_uart: fix race during initialization")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:32:15 -04:00
Arseniy Krasnov
366ceff495 Bluetooth: hci_uart: fix race during initialization
'hci_register_dev()' calls power up function, which is executed by
kworker - 'hci_power_on()'. This function does access to bluetooth chip
using callbacks from 'hci_ldisc.c', for example 'hci_uart_send_frame()'.
Now 'hci_uart_send_frame()' checks 'HCI_UART_PROTO_READY' bit set, and
if not - it fails. Problem is that 'HCI_UART_PROTO_READY' is set after
'hci_register_dev()', and there is tiny chance that 'hci_power_on()' will
be executed before setting this bit. In that case HCI init logic fails.

Patch moves setting of 'HCI_UART_PROTO_READY' before calling function
'hci_uart_register_dev()'.

Signed-off-by: Arseniy Krasnov <avkrasnov@salutedevices.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:31:59 -04:00
Kiran K
bb3569ac36 Bluetooth: btintel: Add DSBR support for ScP
Add DSBR support for Scorpious Peak cores.

Refer commit eb9e749c01 ("Bluetooth: btintel: Allow configuring drive
strength of BRI") for details about DSBR.

Signed-off-by: Kiran K <kiran.k@intel.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:31:41 -04:00
Jeremy Clifton
461159ece0 Bluetooth: Fix code style warning
Output of checkpatch shows warning:
drivers/bluetooth/bfusb.c:368: WARNING: braces {} are not necessary
for single statement blocks

Remove braces for single line statement.

Signed-off-by: Jeremy Clifton <deaner92@yahoo.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:31:27 -04:00
Dr. David Alan Gilbert
60bfe8a7dc Bluetooth: MGMT: Remove unused mgmt_*_discovery_complete
mgmt_start_discovery_complete() and mgmt_stop_discovery_complete() last
uses were removed in 2022 by
commit ec2904c259 ("Bluetooth: Remove dead code from hci_request.c")

Remove them.

Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
2025-03-25 12:31:02 -04:00