Some ARM Cortex CPUs including A72 have Error Detection And Correction (EDAC)
support on their L1 and L2 caches. That functionality is in implementation
defined registers, so using it is not safe in virtualized environments or when
EL3 already uses these registers. Add a edac-enabled flag which can be
explicitly set when EDAC can be used.
[ bp: Massage commit message. ]
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Vijay Balakrishna <vijayb@linux.microsoft.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/1752714390-27389-3-git-send-email-vijayb@linux.microsoft.com
The driver is designed to support error detection and reporting for
Cortex A72 cores, specifically within their L1 and L2 cache systems.
The errors are detected by reading CPU/L2 memory error syndrome
registers.
Unfortunately there is no robust way to inject errors into the caches,
so this driver doesn't contain any code to actually test it. It has
been tested though with code taken from an older version [1] of this
driver. For reasons stated in thread [1], the error injection code is
not suitable for mainline, so it is removed from the driver.
[1] https://lore.kernel.org/all/1521073067-24348-1-git-send-email-york.sun@nxp.com/#t
[ bp: minor touchups. ]
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Co-developed-by: Vijay Balakrishna <vijayb@linux.microsoft.com>
Signed-off-by: Vijay Balakrishna <vijayb@linux.microsoft.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/1752714390-27389-2-git-send-email-vijayb@linux.microsoft.com
In 32-bit arm, the build fails with:
error[E0308]: mismatched types
--> drivers/gpu/drm/nova/file.rs:42:28
|
42 | getparam.set_value(value);
| --------- ^^^^^ expected `u64`, found `u32`
| |
| arguments to this method are incorrect
|
note: method defined here
--> drivers/gpu/drm/nova/uapi.rs:29:12
|
29 | pub fn set_value(&self, v: u64) {
| ^^^^^^^^^ ------
help: you can convert a `u32` to a `u64`
|
42 | getparam.set_value(value.into());
| +++++++
The reason is that `Getparam::set_value` takes a `u64` (from the UAPI),
but `pci::Device::resource_len()` returns a `resource_size_t`, which is a
`phys_addr_t`, which may be 32- or 64-bit.
Thus add an `into()` call to support the 32-bit case, while allowing the
Clippy lint that complains in the 64-bit case where the type is the same.
Fixes: cdeaeb9dd7 ("drm: nova-drm: add initial driver skeleton")
Signed-off-by: Miguel Ojeda <ojeda@kernel.org>
Reviewed-by: Christian Schrefl <chrisi.schrefl@gmail.com>
Link: https://lore.kernel.org/r/20250724165441.2105632-1-ojeda@kernel.org
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
'-mpreferred-stack-boundary' (the GCC option) and '-mstack-alignment'
(the clang option) have been supported in their respective compilers for
some time, so it is unnecessary to check for support for them via
cc-option. '-mpreferred-stack-boundary=3' had a restriction on
'-mno-sse' until GCC 7.1 but that is irrelevant for most of the kernel,
which includes '-mno-sse'.
Move to simple Kconfig checks to avoid querying the compiler for the
flags that it supports.
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20250814-x86-min-ver-cleanups-v1-2-ff7f19457523@kernel.org
Update call sites in the driver-core files and its related samples
to import `ARef` and `AlwaysRefCounted` from `sync::aref`
instead of `types`.
This aligns with the ongoing effort to move `ARef` and
`AlwaysRefCounted` to sync.
Suggested-by: Benno Lossin <lossin@kernel.org>
Link: https://github.com/Rust-for-Linux/linux/issues/1173
Signed-off-by: Shankari Anand <shankari.ak0208@gmail.com>
Link: https://lore.kernel.org/r/20250814104615.355106-1-shankari.ak0208@gmail.com
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
2290 was found in the field to also require this quirk, as long &
high-bandwidth workloads (e.g. USB ethernet) are consistently able to
crash the controller otherwise.
The same change has been made for a number of SoCs in [1], but QCM2290
somehow escaped the list (even though the very closely related SM6115
was there).
Upon a controller crash, the log would read:
xhci-hcd.12.auto: xHCI host not responding to stop endpoint command
xhci-hcd.12.auto: xHCI host controller not responding, assume dead
xhci-hcd.12.auto: HC died; cleaning up
Add snps,parkmode-disable-ss-quirk to the DWC3 instance in order to
prevent the aforementioned breakage.
[1] https://lore.kernel.org/all/20240704152848.3380602-1-quic_kriskura@quicinc.com/
Cc: stable@vger.kernel.org
Reported-by: Rob Clark <robin.clark@oss.qualcomm.com>
Fixes: a64a0192b7 ("arm64: dts: qcom: Add initial QCM2290 device tree")
Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Link: https://lore.kernel.org/r/20250708-topic-2290_usb-v1-1-661e70a63339@oss.qualcomm.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
Based on a bisect, it appears that commit 7ee9887703 ("timers:
Implement the hierarchical pull model") has somehow inadvertently
broken BPF selftest test_tcpnotify_user. The error that is being
generated by this test is as follows:
FAILED: Wrong stats Expected 10 calls, got 8
It looks like the change allows timer functions to be run on CPUs
different from the one they are armed on. The test had pinned itself
to CPU 0, and in the past the retransmit attempts also occurred on CPU
0. The test had set the max_entries attribute for
BPF_MAP_TYPE_PERF_EVENT_ARRAY to 2 and was calling
bpf_perf_event_output() with BPF_F_CURRENT_CPU, so the entry was
likely to be in range. With the change to allow timers to run on other
CPUs, the current CPU tasked with performing the retransmit might be
bumped and in turn fall out of range, as the event will be filtered
out via __bpf_perf_event_output() using:
if (unlikely(index >= array->map.max_entries))
return -E2BIG;
A possible change would be to explicitly set the max_entries attribute
for perf_event_map in test_tcpnotify_kern.c to a value that's at least
as large as the number of CPUs. As it turns out however, if the field
is left unset, then the libbpf will determine the number of CPUs available
on the underlying system and update the max_entries attribute accordingly
in map_set_def_max_entries().
A further problem with the test is that it has a thread that continues
running up until the program exits. The main thread cleans up some
LIBBPF data structures, while the other thread continues to use them,
which inevitably will trigger a SIGSEGV. This can be dealt with by
telling the thread to run for as long as necessary and doing a
pthread_join on it before exiting the program.
Finally, I don't think binding the process to CPU 0 is meaningful for
this test any more, so get rid of that.
Fixes: 435f90a338 ("selftests/bpf: add a test case for sock_ops perf-event notification")
Signed-off-by: Matt Bobrowski <mattbobrowski@google.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/aJ8kHhwgATmA3rLf@google.com
Signed-off-by: Carlos Maiolino <cem@kernel.org>
-----BEGIN PGP SIGNATURE-----
iJUEABMJAB0WIQSmtYVZ/MfVMGUq1GNcsMJ8RxYuYwUCaJ8nJQAKCRBcsMJ8RxYu
Y91GAX4q+aKhXBJzzEYGaCGEajMNHlClPI9Ac5AlhSkzW/XdZdS8FhIgJMNegTst
yOKZ32kBgPoBtiVDbRs7h0USqmcd94f9IwmoLa+0miUKyHHJaZIR97of9F/P1o3q
Lx2dLf5hwQ==
=e0dN
-----END PGP SIGNATURE-----
Merge tag 'xfs-fixes-6.17-rc2' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux
Pull xfs fixes from Carlos Maiolino:
- Fix an assert trigger introduced during the merge window
- Prevent atomic writes to be used with DAX
- Prevent users from using the max_atomic_write mount option without
reflink, as atomic writes > 1block are not supported without reflink
- Fix a null-pointer-deref in a tracepoint
* tag 'xfs-fixes-6.17-rc2' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
xfs: split xfs_zone_record_blocks
xfs: fix scrub trace with null pointer in quotacheck
xfs: reject max_atomic_write mount option for no reflink
xfs: disallow atomic writes on DAX
fs/dax: Reject IOCB_ATOMIC in dax_iomap_rw()
xfs: remove XFS_IBULK_SAME_AG
xfs: fully decouple XFS_IBULK* flags from XFS_IWALK* flags
xfs: fix frozen file system assert in xfs_trans_alloc
While trying to fill a random RSS key, the size of the pointer
is being used rather than the actual size of the RSS key.
Fix by passing an appropriate value of the RSS key.
This issue was reported by static coverity analyser.
Fixes: eb4898fde1 ("net: libwx: add wangxun vf common api")
Signed-off-by: Chandra Mohan Sundar <chandramohan.explore@gmail.com>
Link: https://patch.msgid.link/20250814163014.613004-1-chandramohan.explore@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Replace memset(0) followed by strscpy() with strscpy_pad() to improve
netdev_boot_setup_add(). This avoids zeroing the memory before copying
the string and ensures the destination buffer is only written to once,
simplifying the code and improving efficiency.
Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
Link: https://patch.msgid.link/20250814180514.251000-2-thorsten.blum@linux.dev
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Tariq Toukan says:
====================
net/mlx5: Support disabling host PFs
This small series by Daniel adds support for disabling host PFs.
If device is capable and configured, the driver won't access vports of
disabled host functions.
====================
Link: https://patch.msgid.link/1755112796-467444-1-git-send-email-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Some devices support disabling the physical function on the host. When
this is configured the vports for the host functions do not exist.
This patch checks if host functions are enabled before trying to access
their vports.
Signed-off-by: Daniel Jurgens <danielj@nvidia.com>
Reviewed-by: William Tu <witu@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://patch.msgid.link/1755112796-467444-3-git-send-email-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The host PF can be disabled, query firmware to check if the host PF of
this function exists.
Signed-off-by: Daniel Jurgens <danielj@nvidia.com>
Reviewed-by: William Tu <witu@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://patch.msgid.link/1755112796-467444-2-git-send-email-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
A struct mii_bus doesn't need a parent, so we can simplify the code and
remove using a faux device. Only difference is the following in sysfs
under /sys/class/mdio_bus:
old: fixed-0 -> '../../devices/faux/Fixed MDIO bus/mdio_bus/fixed-0'
new: fixed-0 -> ../../devices/virtual/mdio_bus/fixed-0
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://patch.msgid.link/e9426bb9-f228-4b99-bc09-a80a958b5a93@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
vhca id is already cached in the vport structure no need to query on
every mlx5 layer, use the mlx5_vport_get_vhca_id, where possible.
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Alexei Lazar <alazar@nvidia.com>
Reviewed-by: Feng Liu <feliu@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Dynamically created vports require vhca id as input to set/query other
vport hca cap, when FW is capable and the vhca id of a vport is valid
use it instead of the local function id.
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Adithya Jayachandran <ajayachandra@nvidia.com>
Reviewed-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Feng Liu <feliu@nvidia.com>
Reviewed-by: William Tu <witu@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
We need vhca_id to set up the vhca_id to vport mapping for every vport,
for that we query the firmware in mlx5_esw_vport_vhca_id_set, and it is
redundant since in esw_vport_setup, we already query hca caps which has
the vhca_id, cache it there and save 2 extra fw queries per vport.
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Alexei Lazar <alazar@nvidia.com>
Reviewed-by: Feng Liu <feliu@nvidia.com>
Next patches will implement the discovery and creation of adjacent
functions vports, this patch introduces the hardware structures
definitions needed for the driver implementation.
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Parav Pandit <parav@nvidia.com>
Reviewed-by: Jack Morgenstein <jackm@nvidia.com>
Signed-off-by: Alexei Lazar <alazar@nvidia.com>
When converting a Box<T> into a void pointer, the allocator might
guarantee a higher alignment than the type itself does, and in that case
it is guaranteed that the void pointer has that higher alignment.
This is quite useful when combined with the XArray, which you can only
create using a ForeignOwnable whose FOREIGN_ALIGN is at least 4. This
means that you can now always use a Box<T> with the XArray no matter the
alignment of T.
Reviewed-by: Benno Lossin <lossin@kernel.org>
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
Acked-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Reviewed-by: Andreas Hindborg <a.hindborg@kernel.org>
Link: https://lore.kernel.org/r/20250811-align-min-allocator-v2-2-3386cc94f4fc@google.com
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
The kernel's allocators sometimes provide a higher alignment than the
end-user requested, so add a new constant on the Allocator trait to let
the allocator specify what its minimum guaranteed alignment is.
This allows the ForeignOwnable trait to provide a more accurate value of
FOREIGN_ALIGN when using a pointer type such as Box, which will be
useful with certain collections such as XArray that store its own data
in the low bits of pointers.
Reviewed-by: Benno Lossin <lossin@kernel.org>
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
Acked-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Reviewed-by: Andreas Hindborg <a.hindborg@kernel.org>
Link: https://lore.kernel.org/r/20250811-align-min-allocator-v2-1-3386cc94f4fc@google.com
[ Add helper for ARCH_KMALLOC_MINALIGN; remove cast to usize. - Danilo ]
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
vcpu_write_sys_reg()'s signature is not totally obvious, and it
is rather easy to write something that looks correct, except that...
Oh wait...
Swap addr and FAR_EL2 to restore some sanity in the nested SEA
department.
Fixes: 9aba641b9e ("KVM: arm64: nv: Respect exception routing rules for SEAs")
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20250813163747.2591317-1-maz@kernel.org
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Replace kfree() with kvfree() for memory allocated by kvmalloc().
Compile-tested only.
Cc: stable@vger.kernel.org
Fixes: 8a8b1ec526 ("drm/nouveau/gsp: split rpc handling out on its own")
Signed-off-by: Qianfeng Rong <rongqianfeng@vivo.com>
Reviewed-by: Timur Tabi <ttabi@nvidia.com>
Acked-by: Zhi Wang <zhiw@nvidia.com>
Link: https://lore.kernel.org/r/20250813125412.96178-1-rongqianfeng@vivo.com
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Add bindings to obtain a PCI device's resource start address, bus/
device function, revision ID and subsystem device and vendor IDs.
These will be used by the nova-core GPU driver which is currently in
development.
Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
Cc: Danilo Krummrich <dakr@kernel.org>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Krzysztof Wilczyński <kwilczynski@kernel.org>
Cc: Miguel Ojeda <ojeda@kernel.org>
Cc: Alex Gaynor <alex.gaynor@gmail.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Gary Guo <gary@garyguo.net>
Cc: Björn Roy Baron <bjorn3_gh@protonmail.com>
Cc: Benno Lossin <lossin@kernel.org>
Cc: Andreas Hindborg <a.hindborg@kernel.org>
Cc: Alice Ryhl <aliceryhl@google.com>
Cc: Trevor Gross <tmgross@umich.edu>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Rafael J. Wysocki <rafael@kernel.org>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Alexandre Courbot <acourbot@nvidia.com>
Cc: linux-pci@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Alistair Popple <apopple@nvidia.com>
Link: https://lore.kernel.org/r/20250730013417.640593-2-apopple@nvidia.com
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Update PCI FFI callback signatures to use `c_` from the prelude,
instead of accessing it via `kernel::ffi::`.
Signed-off-by: Abhinav Ananthu <abhinav.ogl@gmail.com>
Reviewed-by: Benno Lossin <lossin@kernel.org>
Link: https://lore.kernel.org/r/20250812033101.5257-1-abhinav.ogl@gmail.com
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Update auxiliary FFI callback signatures to reference the `c_` types
provided by the kernel prelude, rather than accessing them via
`kernel::ffi::`.
Signed-off-by: Abhinav Ananthu <abhinav.ogl@gmail.com>
Link: https://lore.kernel.org/r/20250812075109.4099-1-abhinav.ogl@gmail.com
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
- hci_conn: Fix running bis_cleanup for hci_conn->type PA_LINK
- hci_conn: Fix not cleaning up Broadcaster/Broadcast Source
- hci_core: Fix using {cis,bis}_capable for current settings
- hci_core: Fix using ll_privacy_capable for current settings
- hci_core: Fix not accounting for BIS/CIS/PA links separately
- hci_conn: do return error from hci_enhanced_setup_sync()
- hci_event: fix MTU for BN == 0 in CIS Established
- hci_sync: Fix scan state after PA Sync has been established
- hci_sync: Avoid adding default advertising on startup
- hci_sync: Prevent unintended PA sync when SID is 0xFF
- ISO: Fix getname not returning broadcast fields
- btmtk: Fix wait_on_bit_timeout interruption during shutdown
- btnxpuart: Uses threaded IRQ for host wakeup handling
-----BEGIN PGP SIGNATURE-----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=ZtN+
-----END PGP SIGNATURE-----
Merge tag 'for-net-2025-08-15' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth
Luiz Augusto von Dentz says:
====================
bluetooth pull request for net:
- hci_conn: Fix running bis_cleanup for hci_conn->type PA_LINK
- hci_conn: Fix not cleaning up Broadcaster/Broadcast Source
- hci_core: Fix using {cis,bis}_capable for current settings
- hci_core: Fix using ll_privacy_capable for current settings
- hci_core: Fix not accounting for BIS/CIS/PA links separately
- hci_conn: do return error from hci_enhanced_setup_sync()
- hci_event: fix MTU for BN == 0 in CIS Established
- hci_sync: Fix scan state after PA Sync has been established
- hci_sync: Avoid adding default advertising on startup
- hci_sync: Prevent unintended PA sync when SID is 0xFF
- ISO: Fix getname not returning broadcast fields
- btmtk: Fix wait_on_bit_timeout interruption during shutdown
- btnxpuart: Uses threaded IRQ for host wakeup handling
* tag 'for-net-2025-08-15' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth:
Bluetooth: hci_core: Fix not accounting for BIS/CIS/PA links separately
Bluetooth: btnxpuart: Uses threaded IRQ for host wakeup handling
Bluetooth: hci_conn: do return error from hci_enhanced_setup_sync()
Bluetooth: hci_event: fix MTU for BN == 0 in CIS Established
Bluetooth: hci_sync: Prevent unintended PA sync when SID is 0xFF
Bluetooth: hci_core: Fix using ll_privacy_capable for current settings
Bluetooth: hci_core: Fix using {cis,bis}_capable for current settings
Bluetooth: btmtk: Fix wait_on_bit_timeout interruption during shutdown
Bluetooth: hci_conn: Fix not cleaning up Broadcaster/Broadcast Source
Bluetooth: hci_conn: Fix running bis_cleanup for hci_conn->type PA_LINK
Bluetooth: ISO: Fix getname not returning broadcast fields
Bluetooth: hci_sync: Fix scan state after PA Sync has been established
Bluetooth: hci_sync: Avoid adding default advertising on startup
====================
Link: https://patch.msgid.link/20250815142229.253052-1-luiz.dentz@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Petr Machata says:
====================
mlxsw: spectrum: Forward packets with an IPv4 link-local source IP
By default, Spectrum devices do not forward IPv4 packets with a link-local
source IP (i.e., 169.254.0.0/16). This behavior does not align with the
kernel which does forward them. Fix the issue and add a selftest.
====================
Link: https://patch.msgid.link/cover.1755174341.git.petrm@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
By default, the device does not forward IPv4 packets with a link-local
source IP (i.e., 169.254.0.0/16). This behavior does not align with the
kernel which does forward them.
Fix by instructing the device to forward such packets instead of
dropping them.
Fixes: ca360db4b8 ("mlxsw: spectrum: Disable DIP_LINK_LOCAL check in hardware pipeline")
Reported-by: Zoey Mertes <zoey@cloudflare.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Link: https://patch.msgid.link/6721e6b2c96feb80269e72ce8d0b426e2f32d99c.1755174341.git.petrm@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Skipped tests reported by kselftest.h use a different format than KTAP,
there is no explicit test name. Normally the test name is part of the
free-form string after the SKIP keyword:
ok 3 # SKIP test: some reason
Extend the parser to handle those correctly. Use the free-form string as
test name instead.
Link: https://lore.kernel.org/r/20250813-kunit-kselftesth-skip-v1-1-57ae3de4f109@linutronix.de
Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
For kbuild to properly clean up these build artifacts in the subdirectory,
even after CONFIG_KUNIT changed do disabled, the directory needs to be
processed always.
Pushing the special logic for hook.o into the kunit Makefile also makes the
logic easier to understand.
Link: https://lore.kernel.org/r/20250813-kunit-always-descend-v1-1-7bbd387ff13b@linutronix.de
Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Since commit 67b746f94f ("net: bridge: mcast: make sure querier
port/address updates are consistent"), the argument 'querier' is unused,
just get rid of it.
Signed-off-by: Wang Liang <wangliang74@huawei.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Link: https://patch.msgid.link/20250814042355.1720755-1-wangliang74@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When a poison is consumed on the guest before the guest receives the host's poison creation msg, a corner case may occur to have poison_handler complete processing earlier than it should to cause the guest to hang waiting for the req_bad_pages reply during a VF FLR, resulting in the VM becoming inaccessible in stress tests.
To fix this issue, this patch refactored the mailbox sequence by seperating the bad_page_work into two parts req_bad_pages_work and handle_bad_pages_work.
Old sequence:
1.Stop data exchange work
2.Guest sends MB_REQ_RAS_BAD_PAGES to host and keep polling for IDH_RAS_BAD_PAGES_READY
3.If the IDH_RAS_BAD_PAGES_READY arrives within timeout limit, re-init the data exchange region for updated bad page info
else timeout with error message
New sequence:
req_bad_pages_work:
1.Stop data exhange work
2.Guest sends MB_REQ_RAS_BAD_PAGES to host
Once Guest receives IDH_RAS_BAD_PAGES_READY event
handle_bad_pages_work:
3.re-init the data exchange region for updated bad page info
Signed-off-by: Chenglei Xie <Chenglei.Xie@amd.com>
Reviewed-by: Shravan Kumar Gande <Shravankumar.Gande@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The duplicate judgment of wptr_va could be removed to simplify the logic
Signed-off-by: Qiang Liu <liuqiang@kylinos.cn>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The function dc_stream_set_cursor_attributes() currently dereferences
the `stream` pointer and nested members `stream->ctx->dc->current_state`
without checking for NULL.
All callers of these functions, such as in
`dcn30_apply_idle_power_optimizations()` and
`amdgpu_dm_plane_handle_cursor_update()`, already perform NULL checks
before calling these functions.
Fixes below:
drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc_stream.c:336 dc_stream_program_cursor_attributes()
error: we previously assumed 'stream' could be null (see line 334)
drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc_stream.c
327 bool dc_stream_program_cursor_attributes(
328 struct dc_stream_state *stream,
329 const struct dc_cursor_attributes *attributes)
330 {
331 struct dc *dc;
332 bool reset_idle_optimizations = false;
333
334 dc = stream ? stream->ctx->dc : NULL;
^^^^^^
The old code assumed stream could be NULL.
335
--> 336 if (dc_stream_set_cursor_attributes(stream, attributes)) {
^^^^^^
The refactor added an unchecked dereference.
drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc_stream.c
313 bool dc_stream_set_cursor_attributes(
314 struct dc_stream_state *stream,
315 const struct dc_cursor_attributes *attributes)
316 {
317 bool result = false;
318
319 if (dc_stream_check_cursor_attributes(stream, stream->ctx->dc->current_state, attributes)) {
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Here.
This function used to check for if stream as NULL and return false at
the start. Probably we should add that back.
Fixes: 4465dd0e41 ("drm/amd/display: Refactor SubVP cursor limiting logic")
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Cc: Alex Hung <alex.hung@amd.com>
Cc: Alvin Lee <alvin.lee2@amd.com>
Cc: Ray Wu <ray.wu@amd.com>
Cc: Dillon Varone <dillon.varone@amd.com>
Cc: Aurabindo Pillai <aurabindo.pillai@amd.com>
Cc: Roman Li <roman.li@amd.com>
Cc: ChiaHsuan Chung <chiahsuan.chung@amd.com>
Cc: Harry Wentland <harry.wentland@amd.com>
Cc: Daniel Wheeler <daniel.wheeler@amd.com>
Cc: Tom Chung <chiahsuan.chung@amd.com>
Cc: Wenjing Liu <wenjing.liu@amd.com>
Cc: Jun Lei <Jun.Lei@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Dillon Varone <Dillon.varone@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Various small typos found around.
Signed-off-by: Alexandre Demers <alexandre.f.demers@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Take into account the limits from the vbios. Ported
from the SMU13 code.
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4352
Reviewed-by: Jesse Zhang <Jesse.Zhang@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This change reorganizes VCN reset capability detection by:
1. Moving reset mask configuration from sw_init to new late_init phase
2. Adding vcn_v4_0_3_late_init() to properly check for per-queue reset support
3. Only setting soft full reset mask as fallback when per-queue reset isn't supported
4. Removing TODO comment now that queue reset support is implemented
V2: Removed unrelated changes. Keep amdgpu_get_soft_full_reset_mask in place
and remove TODO comment. (Alex)
v3: set the flags at one place (all in late_init) (Lijo)
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Ruili Ji <ruiliji2@amd.com>
Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
HMM assumes that pages have READ permissions by default. Inside
svm_range_validate_and_map, we add READ permissions then add WRITE
permissions if the VMA isn't read-only. This will conflict with regions
that only have PROT_WRITE or have PROT_NONE. When that happens,
svm_range_restore_work will continue to retry, silently, giving the
impression of a hang if pr_debug isn't enabled to show the retries..
If pages don't have READ permissions, simply unmap them and continue. If
they weren't mapped in the first place, this would be a no-op. Since x86
doesn't support write-only, and PROT_NONE doesn't allow reads or writes
anyways, this will allow the svm range validation to continue without
getting stuck in a loop forever on mappings we can't use with HMM.
Signed-off-by: Kent Russell <kent.russell@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit implements VCN reset capability for SMU v13.0.6 with the following changes:
1. Added new PPSMC message ID (0x5B) for VCN reset in SMU firmware interface
2. Extended SMU capabilities to include VCN_RESET support
3. Implemented VCN reset support check:
- Added smu_v13_0_6_reset_vcn_is_supported() function
4. Updated SMU v13.0.6 PPT functions to include VCN reset operations
v2: clean up debug info (Alex)
v3: remove unsupported message and split smu v13.0.6 changes to a separate patch (Lijo)
v4: simply the function (smu_v13_0_6_reset_vcn_is_supported) (Lijo)
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Ruili Ji <ruiliji2@amd.com>
Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This change introduces infrastructure to check whether VCN reset
is supported by the SMU firmware. Key changes include:
1. Added new functions to query VCN reset support:
- amdgpu_dpm_reset_vcn_is_supported()
- smu_reset_vcn_is_supported()
- pptable_funcs.reset_vcn_is_supported callback
2. Implemented proper locking in the DPM layer with mutex protection
3. Maintained consistency with existing SDMA reset support checks
The new capability allows callers to check for VCN reset support
before attempting the operation, preventing unnecessary attempts
on unsupported platforms.
v2: clean up debug info(Alex)
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Ruili Ji <ruiliji2@amd.com>
Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
If a amdgpu_bo_va is fpriv->prt_va, the bo of this one is always NULL.
So, such kind of amdgpu_bo_va should be updated separately before
amdgpu_vm_handle_moved.
Signed-off-by: Heng Zhou <Heng.Zhou@amd.com>
Reviewed-by: Kasiviswanathan, Harish <Harish.Kasiviswanathan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Currently each kfd compute partition/node is sharing
the same uuid of AID, which doen't meet the CUDA spec
for visible device, so corresponding XCD id for each
partition in smu has been assigned to xcp, and exposed
to kfd topology.
v2: add NULL check (Lijo)
Signed-off-by: Eric Huang <jinhuieric.huang@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
