Use attack vector controls to determine if L1TF mitigation is required.
Disable SMT if cross-thread protection is desired.
Signed-off-by: David Kaplan <david.kaplan@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20250707183316.1349127-17-david.kaplan@amd.com
Use attack vector controls to determine if retbleed mitigation is
required.
Disable SMT if cross-thread protection is desired and STIBP is not
available.
Signed-off-by: David Kaplan <david.kaplan@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20250707183316.1349127-13-david.kaplan@amd.com
Use attack vector controls to determine if MDS mitigation is required.
The global mitigations=off command now simply disables all attack vectors
so explicit checking of mitigations=off is no longer needed.
If cross-thread attack mitigations are required, disable SMT.
Signed-off-by: David Kaplan <david.kaplan@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20250707183316.1349127-6-david.kaplan@amd.com
Add a function which defines which vulnerabilities should be mitigated
based on the selected attack vector controls. The selections here are
based on the individual characteristics of each vulnerability.
Signed-off-by: David Kaplan <david.kaplan@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20250707183316.1349127-5-david.kaplan@amd.com
ARCH_HAS_CPU_ATTACK_VECTORS should be set for architectures which implement
the new attack-vector based controls for CPU mitigations. If an arch does
not support attack-vector based controls then an attempt to use them
results in a warning.
Signed-off-by: David Kaplan <david.kaplan@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20250707183316.1349127-4-david.kaplan@amd.com
Define 4 new attack vectors that are used for controlling CPU speculation
mitigations. These may be individually disabled as part of the
mitigations= command line. Attack vector controls are combined with global
options like 'auto' or 'auto,nosmt' like 'mitigations=auto,no_user_kernel'.
The global options come first in the mitigations= string.
Cross-thread mitigations can either remain enabled fully, including
potentially disabling SMT ('auto,nosmt'), remain enabled except for
disabling SMT ('auto'), or entirely disabled through the new
'no_cross_thread' attack vector option.
The default settings for these attack vectors are consistent with existing
kernel defaults, other than the automatic disabling of VM-based attack
vectors if KVM support is not present.
Signed-off-by: David Kaplan <david.kaplan@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20250707183316.1349127-3-david.kaplan@amd.com
The think-lmi driver uses the firwmare_attributes_class. But this class
is registered after think-lmi, causing the "think-lmi" directory in
"/sys/class/firmware-attributes" to be missing when the driver is
compiled as builtin.
Fixes: 5592240380 ("platform/x86: think-lmi: Directly use firmware_attributes_class")
Signed-off-by: Torsten Hilbrich <torsten.hilbrich@secunet.com>
Link: https://lore.kernel.org/r/7dce5f7f-c348-4350-ac53-d14a8e1e8034@secunet.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Document the 5 new attack vector command line options, how they
interact with existing vulnerability controls, and recommendations on when
they can be disabled.
Note that while mitigating against untrusted userspace requires both
user-to-kernel and user-to-user protection, these are kept separate. The
kernel can control what code executes inside of it and that may affect the
risk associated with vulnerabilities especially if new kernel mitigations
are implemented. The same isn't typically true of userspace.
In other words, the risk associated with user-to-user or guest-to-guest
attacks is unlikely to change over time. While the risk associated with
user-to-kernel or guest-to-host attacks may change. Therefore, these
controls are separated.
Signed-off-by: David Kaplan <david.kaplan@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20250709155731.3279419-1-david.kaplan@amd.com
- small fix relevant to arm64 server and custom CMA configuration
(Feng Tang)
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQSrngzkoBtlA8uaaJ+Jp1EFxbsSRAUCaHCzdQAKCRCJp1EFxbsS
RMrMAQDghOwKZqYuC27kJt5T7lgG47YCNE5em1v8WsTSvwQAugEA4AlWIpqQ34eI
Es6ObfMt8Q9gArubFZ0ZDFtmZq9NpA0=
=+z0i
-----END PGP SIGNATURE-----
Merge tag 'dma-mapping-6.16-2025-07-11' of git://git.kernel.org/pub/scm/linux/kernel/git/mszyprowski/linux
Pull dma-mapping fix from Marek Szyprowski:
- small fix relevant to arm64 server and custom CMA configuration (Feng
Tang)
* tag 'dma-mapping-6.16-2025-07-11' of git://git.kernel.org/pub/scm/linux/kernel/git/mszyprowski/linux:
dma-contiguous: hornor the cma address limit setup by user
Make sure to drop the OF node reference taken when creating the Gaokun
auxiliary devices when the devices are later released.
Fixes: 7636f090d0 ("platform: arm64: add Huawei Matebook E Go EC driver")
Cc: Pengyu Luo <mitltlatltl@gmail.com>
Signed-off-by: Johan Hovold <johan@kernel.org>
Reviewed-by: Pengyu Luo <mitltlatltl@gmail.com>
Link: https://lore.kernel.org/r/20250708085358.15657-1-johan@kernel.org
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Registering the "other mode" notifier fails if that is disabled:
x86_64-linux-ld: drivers/platform/x86/lenovo/wmi-gamezone.o: in function `lwmi_gz_probe':
wmi-gamezone.c:(.text+0x336): undefined reference to `devm_lwmi_om_register_notifier'
This could be fixed by adding a stub helper, but a Kconfig 'select'
seems simpler here.
Fixes: 22024ac536 ("platform/x86: Add Lenovo Gamezone WMI Driver")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20250709151734.1268435-1-arnd@kernel.org
[ij: retained the other selects as wmi-gamezone is using them directly.]
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
After we do the modification on the host side, ensure we write the
result back to VRAM and not the other way around, otherwise the
modification will be lost if treated like a read.
Fixes: 270172f64b ("drm/xe: Update xe_ttm_access_memory to use GPU for non-visible access")
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://lore.kernel.org/r/20250710134128.800756-2-matthew.auld@intel.com
Firmware would assert on undefined bits in DSM-originated DWs.
With this change, Firmware introduces a fail-safe mechanism
and removes the assert behavior. This ensures robustness when
handling raw DSM table data.
Signed-off-by: Pagadala Yesu Anjaneyulu <pagadala.yesu.anjaneyulu@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20250710212632.eee871df03c9.I2be2eaa16437e84aa1be0c6c95ec334034ce7e50@changeid
In the current code, if there was a rekey, we remove all the existing keys
from mac80211, then re-add the ones that the FW sent with
ieee80211_gtk_rekey_add, (newer FW will send also the existing GTKs/BIGTKs)
and then update the sequence number.
Instead of removing and re-adding the existing keys for no good reason,
we can just update the sequence of all keys, also of the ones that are
going to be replaced, and update again after the replace.
This change is required because ieee80211_gtk_rekey_add is going to be
changed to lookup the cipher from the old key instead of receiving it as an
argument, and for this it will need the old key(s), so we can't remove all
keys.
Note that with this change, in case that a key that existed before wowlan
is replaced, mac80211 will now call the driver to remove the old key and
add the new one (as opposed the previous behaviour, in which the key was
removed by the driver itself).
Of course we don't want to run the set_key callbacks in this case, so just
return early.
Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20250710212632.f0b5e19f77f5.I958f4926f168cdad6d4d7720ebde2f5e812b297d@changeid
In the current code, if there was a rekey, we remove all the existing keys
from mac80211, then re-add the ones that the FW sent with
ieee80211_gtk_rekey_add, (newer FW will send also the existing GTKs/BIGTKs)
and then update the sequence number.
Instead of removing and re-adding the existing keys for no good reason,
we can just update the sequence of all keys, also of the ones that are
going to be replaced, and update again after the replace.
This change is required because ieee80211_gtk_rekey_add is going to be
changed to lookup the cipher from the old key instead of receiving it as an
argument, and for this it will need the old key(s), so we can't remove all
keys.
Note that with this change, in case that a key that existed before wowlan
is replaced, mac80211 will now call the driver to remove the old key and
add the new one (as opposed the previous behaviour, in which the key was
removed by the driver itself).
Of course we don't want to run the set_key callbacks in this case, so just
return early.
Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20250710212632.10091484e38e.I45daf089189f606f3879ca4538fb46303d761710@changeid
For older APIs, gtk_seq[0].key_id will be set to -1, while other code
assumes that it will always be a valid gtk key idx.
Since the FW does provide the idx for those APIs, simply set it.
Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20250710212632.2a5a5e4d7b19.Ib795751119070495c0c95526cd36bd7f87594f56@changeid
The default capabilities were set to much more than the hardware
currently is intended to support, and then masked off for only
the GL MAC type. However, this was due to some miscommunication
and is incorrect, it should've been masked off for all current
and planned MACs/RFs. Instead of doing this removing and adding
of capabilities, simply list the ones that currently apply to
all devices. If this changes for a new device type we can change
the code, but at least the driver won't erroneously advertise
some capabilities that aren't actually implemented in hardware.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20250710212632.69b9dac7d906.I7885499bc80453d5a84285ec927f5e89f9adfb21@changeid
Add iwl_trans_suppress_cmd_error_once() function to be called by the op
modes instead of directly checking the trans status bits.
This hides the trans internal implementation details from callers.
Signed-off-by: Rotem Kerem <rotem.kerem@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20250710212632.e061ae241fd3.Ie2043c4e237196ebcfe195006d3d76371de48a55@changeid
When a GO is started, the P2P device mac is updated to indicate
that frames for the P2P device mac should be filtered in while
the GO is active. However, this configuration is done after the GO is
already started so it doesn't take effect. Fix it by updating the
P2P device mac before adding the broadcast station, which actually
starts the GO.
Signed-off-by: Avraham Stern <avraham.stern@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20250710212632.463a0ad545f9.I85a25484d787b65f6a27e794285911e319df0b2d@changeid
The scan request validation function uses bitwise and instead
of logical and. Fix it.
Signed-off-by: Avraham Stern <avraham.stern@intel.com>
Reviewed-by: Daniel Gabay <daniel.gabay@intel.com>
Reviewed-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20250710212632.ec7d665f56a4.I416816b491fafa5d3efdf0a4be78356eedf2bd95@changeid
Add a device-tree for the Asus Portable AiO P1801-T, which is a NVIDIA
Tegra30-based 2-in-1 detachable tablet, originally running Android.
The tablet was also sold together with a PC docking station as the
Transformer AiO P1801.
Tested-by: Andreas Westman Dorcsak <hedmoo@yahoo.com> # P1801-T with dock
Co-developed-by: Maxim Schwalm <maxim.schwalm@gmail.com>
Signed-off-by: Maxim Schwalm <maxim.schwalm@gmail.com>
Signed-off-by: Svyatoslav Ryhel <clamor95@gmail.com>
Link: https://lore.kernel.org/r/20250616073947.13675-3-clamor95@gmail.com
Signed-off-by: Thierry Reding <treding@nvidia.com>
When -Wformat-security is enabled, this new code triggers it:
drivers/platform/x86/intel/pmt/discovery.c: In function 'pmt_features_discovery':
drivers/platform/x86/intel/pmt/discovery.c:505:36: error: format not a string literal and no format arguments [-Werror=format-security]
505 | pmt_feature_names[feature->id]);
Fixes: d9a0788093 ("platform/x86/intel/pmt: Add PMT Discovery driver")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20250711072718.2748415-1-arnd@kernel.org
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Add support to enable Memory Controller (MC) and External Memory
Controller (EMC) drivers for Tegra264. The nodes for MC and EMC are
mostly the same as Tegra234 but differ in number of channels and
interrupt numbers.
The patch also adds the bandwidth manager definitions required for
Tegra264 and uses them to populate the memory client table. All of
these are needed to properly enable memory interconnect (ICC) support.
Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20250709222147.3758356-3-thierry.reding@gmail.com
Signed-off-by: Thierry Reding <treding@nvidia.com>
The BPMP firmware on Tegra264 defines a set of IDs for clock and reset
resources. These are not enumerations but provided by hardware, and 0 is
a reserved value, hence the numbering starts at 1.
Acked-by: Rob Herring (Arm) <robh@kernel.org>
Signed-off-by: Thierry Reding <treding@nvidia.com>
