Commit graph

1,403,262 commits

Author SHA1 Message Date
Kurt Borja
07ac275981
platform/x86: alienware-wmi-wmax: Add support for manual fan control
All models with the "AWCC" WMAX device support a way of manually
controlling fans.

The PWM duty cycle of a fan can't be controlled directly. Instead the
AWCC interface let's us tune a fan `boost` value, which has the
following empirically discovered, approximate behavior over the PWM
value:

	pwm = pwm_base + (fan_boost / 255) * (pwm_max - pwm_base)

Where the pwm_base is the locked PWM value controlled by the FW and
fan_boost is a value between 0 and 255.

Expose this fan_boost knob as a custom HWMON attribute.

Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Jean Delvare <jdelvare@suse.com>
Cc: linux-hwmon@vger.kernel.org
Reviewed-by: Armin Wolf <W_Armin@gmx.de>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Link: https://lore.kernel.org/r/20250329-hwm-v7-8-a14ea39d8a94@gmail.com
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 13:09:51 +03:00
Karol Wachowski
31660b406d accel/ivpu: Add cmdq_id to job related logs
Add tracking of command queue ID in JOB debug message to improve
debugging capabilities.

Signed-off-by: Karol Wachowski <karol.wachowski@intel.com>
Signed-off-by: Maciej Falkowski <maciej.falkowski@linux.intel.com>
Reviewed-by: Lizhi Hou <lizhi.hou@amd.com>
Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://lore.kernel.org/r/20250401155939.4049467-1-maciej.falkowski@linux.intel.com
2025-04-11 12:07:44 +02:00
Kurt Borja
d699907834
platform/x86: alienware-wmi-wmax: Add HWMON support
All models with the "AWCC" WMAX device support monitoring fan speed and
temperature sensors. Expose this feature through the HWMON interface.

Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Jean Delvare <jdelvare@suse.com>
Cc: linux-hwmon@vger.kernel.org
Reviewed-by: Armin Wolf <W_Armin@gmx.de>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Link: https://lore.kernel.org/r/20250329-hwm-v7-7-a14ea39d8a94@gmail.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 13:04:50 +03:00
Kurt Borja
3dde0ae1eb
platform/x86: alienware-wmi-wmax: Add support for the "custom" thermal profile
All models with the "AWCC" WMAX device support a "custom" thermal
profile. In some models this profile signals user-space that the user
wants to manually control the fans, which are always unlocked. In other
models it actually unlocks manual fan control.

Reviewed-by: Armin Wolf <W_Armin@gmx.de>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Link: https://lore.kernel.org/r/20250329-hwm-v7-6-a14ea39d8a94@gmail.com
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 13:04:48 +03:00
Kurt Borja
32b6372ddd
platform/x86: alienware-wmi-wmax: Improve platform profile probe
Get and store the AWCC system description in alienware_awcc_setup()
instead of awcc_platform_profile_probe() and get the correct offset by
iterating through each member of the system_description.

Then add a debug message for unmatched profiles and replace set_bit()
with it's non-atomic version __set_bit() because the choices bitmap only
belongs to this thread.

In the process also check for a malformed system description by defining
an arbitrary limit of resource count.

Reviewed-by: Armin Wolf <W_Armin@gmx.de>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Link: https://lore.kernel.org/r/20250329-hwm-v7-5-a14ea39d8a94@gmail.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 13:04:46 +03:00
Kurt Borja
77bb2ec557
platform/x86: alienware-wmi-wmax: Modify supported_thermal_profiles[]
Rename supported_thermal_profiles[] -> supported_profiles[] and change
it's type to u8 because it stores AWCC thermal IDs.

Reviewed-by: Armin Wolf <W_Armin@gmx.de>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Link: https://lore.kernel.org/r/20250329-hwm-v7-4-a14ea39d8a94@gmail.com
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 13:04:45 +03:00
Kurt Borja
45983d19f3
platform/x86: alienware-wmi-wmax: Improve internal AWCC API
Inline all AWCC WMI helper methods and directly return the newly
introduced awcc_wmi_command() helper to simplify implementation.

Drop awcc_thermal_control() in favor of awcc_op_activate_profile().

Add awcc_op_get_resource_id(), awcc_op_get_current_profile() and a new
failure code.

Reviewed-by: Armin Wolf <W_Armin@gmx.de>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Link: https://lore.kernel.org/r/20250329-hwm-v7-3-a14ea39d8a94@gmail.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 13:04:41 +03:00
Kurt Borja
a000da9dbc
platform/x86: alienware-wmi-wmax: Improve ID processing
Rename AWCC_SENSOR_ID_MASK to AWCC_SENSOR_ID_FLAG and reorder the ID
processing defines in a more logical manner. Then replace their use in
bitwise operations with FIELD_GET().

The latter also involves dropping the AWCC_SENSOR_ID_FLAG check inside
is_awcc_thermal_mode() in favor of extracting the first byte out of IDs
obtained with AWCC_OP_GET_RESOURCE_ID. This is also a requirement to add
support for Alienware Aurora desktops.

While at it, also rename is_awcc_thermal_mode() to
is_awcc_thermal_profile_id().

Reviewed-by: Armin Wolf <W_Armin@gmx.de>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Link: https://lore.kernel.org/r/20250329-hwm-v7-2-a14ea39d8a94@gmail.com
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 13:04:38 +03:00
Kurt Borja
8a1a0fb55f
platform/x86: alienware-wmi-wmax: Rename thermal related symbols
The "thermal" features of the WMAX WMI device are only present on the
host device if the ACPI _UID is "AWCC". Replace WMAX prefixes with
"AWCC" to reflect this relationship.

Thermal profiles with WMAX_PROFILE_BASIC prefix are also renamed to
WMAX_PROFILE_LEGACY because they are only supported in older versions
of this WMI device.

Finally, shorten enum defines for AWCC operations from WMAX_OPERATION_*
to AWCC_OP_*.

Reviewed-by: Armin Wolf <W_Armin@gmx.de>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Kurt Borja <kuurtb@gmail.com>
Link: https://lore.kernel.org/r/20250329-hwm-v7-1-a14ea39d8a94@gmail.com
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 13:04:35 +03:00
Andrzej Kacprowski
1524c28b99 accel/ivpu: Show NPU frequency in sysfs
Add sysfs files that show maximum and current
frequency of the NPU's data processing unit.
New sysfs entries:
- npu_max_frequency_mhz
- npu_current_frequency_mhz

Signed-off-by: Andrzej Kacprowski <Andrzej.Kacprowski@intel.com>
Signed-off-by: Maciej Falkowski <maciej.falkowski@linux.intel.com>
Reviewed-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://lore.kernel.org/r/20250401155912.4049340-3-maciej.falkowski@linux.intel.com
2025-04-11 12:01:20 +02:00
Andrzej Kacprowski
6c2b75404d accel/ivpu: Fix the NPU's DPU frequency calculation
Fix the frequency returned to the user space by
the DRM_IVPU_PARAM_CORE_CLOCK_RATE GET_PARAM IOCTL.
The kernel driver returned CPU frequency for MTL and bare
PLL frequency for LNL - this was inconsistent and incorrect
for both platforms. With this fix the driver returns maximum
frequency of the NPU data processing unit (DPU) for all HW
generations. This is what user space always expected.

Also do not set CPU frequency in boot params - the firmware
does not use frequency passed from the driver, it was only
used by the early pre-production firmware.
With that we can remove CPU frequency calculation code.

Show NPU frequency in FREQ_CHANGE interrupt when frequency
tracking is enabled.

Fixes: 8a27ad81f7 ("accel/ivpu: Split IP and buttress code")
Cc: stable@vger.kernel.org # v6.11+
Signed-off-by: Andrzej Kacprowski <Andrzej.Kacprowski@intel.com>
Signed-off-by: Maciej Falkowski <maciej.falkowski@linux.intel.com>
Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://lore.kernel.org/r/20250401155912.4049340-2-maciej.falkowski@linux.intel.com
2025-04-11 12:01:19 +02:00
Mario Limonciello
fcf27a6a92
platform/x86: amd: pmf: Fix STT limits
On some platforms it has been observed that STT limits are not being
applied properly causing poor performance as power limits are set too low.

STT limits that are sent to the platform are supposed to be in Q8.8
format.  Convert them before sending.

Reported-by: Yijun Shen <Yijun.Shen@dell.com>
Fixes: 7c45534afa ("platform/x86/amd/pmf: Add support for PMF Policy Binary")
Cc: stable@vger.kernel.org
Tested-by: Yijun Shen <Yijun_Shen@Dell.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Acked-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Link: https://lore.kernel.org/r/20250407181915.1482450-1-superm1@kernel.org
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 12:58:33 +03:00
David Thompson
b129005ddf
mlxbf-bootctl: use sysfs_emit_at() in secure_boot_fuse_state_show()
A warning is seen when running the latest kernel on a BlueField SOC:
[251.512704] ------------[ cut here ]------------
[251.512711] invalid sysfs_emit: buf:0000000003aa32ae
[251.512720] WARNING: CPU: 1 PID: 705264 at fs/sysfs/file.c:767 sysfs_emit+0xac/0xc8

The warning is triggered because the mlxbf-bootctl driver invokes
"sysfs_emit()" with a buffer pointer that is not aligned to the
start of the page. The driver should instead use "sysfs_emit_at()"
to support non-zero offsets into the destination buffer.

Fixes: 9886f575de ("platform/mellanox: mlxbf-bootctl: use sysfs_emit() instead of sprintf()")
Signed-off-by: David Thompson <davthompson@nvidia.com>
Link: https://lore.kernel.org/r/20250407132558.2418719-1-davthompson@nvidia.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 12:56:52 +03:00
Hans de Goede
59df54c67b
platform/x86: x86-android-tablets: Add Vexia Edu Atla 10 tablet 5V data
The Vexia EDU ATLA 10 tablet comes in 2 different versions with
significantly different mainboards. The only outward difference is that
the charging barrel on one is marked 5V and the other is marked 9V.

Both are x86 ACPI tablets which ships with Android x86 as factory OS.
with a DSDT which contains a bunch of I2C devices which are not actually
there, causing various resource conflicts. Enumeration of these is skipped
through the acpi_quirk_skip_i2c_client_enumeration().

Extend the existing support for the 9V version by adding support for
manually instantiating the I2C devices which are actually present on
the 5V version by adding the necessary device info to
the x86-android-tablets module.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20250407092017.273124-2-hdegoede@redhat.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 12:55:32 +03:00
Hans de Goede
3343b086c7
platform/x86: x86-android-tablets: Add "9v" to Vexia EDU ATLA 10 tablet symbols
The Vexia EDU ATLA 10 tablet comes in 2 different versions with
significantly different mainboards. The only outward difference is that
the charging barrel on one is marked 5V and the other is marked 9V.

Both need to be handled by the x86-android-tablets code. Add 9v to
the symbols for the existing support for the 9V Vexia EDU ATLA 10 tablet
symbols to prepare for adding support for the 5V version.

All this patch does is s/vexia_edu_atla10_info/vexia_edu_atla10_9v_info/.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20250407092017.273124-1-hdegoede@redhat.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 12:55:31 +03:00
Denis Arefev
6c683c6887
asus-laptop: Fix an uninitialized variable
The value returned by acpi_evaluate_integer() is not checked,
but the result is not always successful, so it is necessary to
add a check of the returned value.

If the result remains negative during three iterations of the loop,
then the uninitialized variable 'val' will be used in the clamp_val()
macro, so it must be initialized with the current value of the 'curr'
variable.

In this case, the algorithm should be less noisy.

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

Fixes: b23910c219 ("asus-laptop: Pegatron Lucid accelerometer")
Cc: stable@vger.kernel.org
Signed-off-by: Denis Arefev <arefev@swemel.ru>
Link: https://lore.kernel.org/r/20250403122603.18172-1-arefev@swemel.ru
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
2025-04-11 12:52:57 +03:00
Karol Wachowski
082a29e20a accel/ivpu: Update FW Boot API to version 3.28.3
Update firmware Boot API to 3.28.3 version and
adjust driver to API changes for preemption buffers.

Use new preemption buffer size fields from FW header added to
firmware boot API for preemption buffers allocations,
if those new fields are zeroed, use old values instead.

Signed-off-by: Karol Wachowski <karol.wachowski@intel.com>
Signed-off-by: Maciej Falkowski <maciej.falkowski@linux.intel.com>
Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://lore.kernel.org/r/20250401155817.4049220-1-maciej.falkowski@linux.intel.com
2025-04-11 11:50:31 +02:00
Maciej Falkowski
683e9fa1c8 accel/ivpu: Flush pending jobs of device's workqueues
Use flush_work() instead of cancel_work_sync() for driver IRQ
workqueues to guarantee that remaining pending work
will be handled.

This resolves two issues that were encountered where a driver was left
in an incorrect state as the bottom-half was canceled:

1. Cancelling context-abort of a job that is still executing and
   is causing translation faults which is going to cause additional TDRs

2. Cancelling bottom-half of a DCT (duty-cycle throttling) request
   which will cause a device to not be adjusted to an external frequency
   request.

Fixes: bc3e5f48b7 ("accel/ivpu: Use workqueue for IRQ handling")
Signed-off-by: Maciej Falkowski <maciej.falkowski@linux.intel.com>
Reviewed-by: Lizhi Hou <lizhi.hou@amd.com>
Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://lore.kernel.org/r/20250401155755.4049156-1-maciej.falkowski@linux.intel.com
2025-04-11 11:42:55 +02:00
Jason Andryuk
164a9f712f x86/xen: Fix __xen_hypercall_setfunc()
Hypercall detection is failing with xen_hypercall_intel() chosen even on
an AMD processor.  Looking at the disassembly, the call to
xen_get_vendor() was removed.

The check for boot_cpu_has(X86_FEATURE_CPUID) was used as a proxy for
the x86_vendor having been set.

When CONFIG_X86_REQUIRED_FEATURE_CPUID=y (the default value), DCE eliminates
the call to xen_get_vendor().  An uninitialized value 0 means
X86_VENDOR_INTEL, so the Intel function is always returned.

Remove the if and always call xen_get_vendor() to avoid this issue.

Fixes: 3d37d9396e ("x86/cpufeatures: Add {REQUIRED,DISABLED} feature configs")
Suggested-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Jason Andryuk <jason.andryuk@amd.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Juergen Gross <jgross@suse.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: "Xin Li (Intel)" <xin@zytor.com>
Link: https://lore.kernel.org/r/20250410193106.16353-1-jason.andryuk@amd.com
2025-04-11 11:39:50 +02:00
Ahmed S. Darwish
62e5652739 x86/cacheinfo: Standardize header files and CPUID references
Reference header files using their canonical form <linux/cacheinfo.h>.

Standardize on CPUID(0xN), instead of CPUID(N), for all standard leaves.
This removes ambiguity and aligns them with their extended counterparts
like CPUID(0x8000001d).

References: 0dd09e215a ("x86/cacheinfo: Apply maintainer-tip coding style fixes")
Signed-off-by: Ahmed S. Darwish <darwi@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Andrew Cooper <andrew.cooper3@citrix.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: John Ogness <john.ogness@linutronix.de>
Link: https://lore.kernel.org/r/20250411070401.1358760-3-darwi@linutronix.de
2025-04-11 11:14:55 +02:00
Ahmed S. Darwish
718f9038ac x86/cpuid: Remove obsolete CPUID(0x2) iteration macro
The CPUID(0x2) cache descriptors iterator at <cpuid/leaf_0x2_api.h>:

    for_each_leaf_0x2_desc()

has no more call sites.  Remove it.

Signed-off-by: Ahmed S. Darwish <darwi@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20250411070401.1358760-2-darwi@linutronix.de
2025-04-11 11:14:55 +02:00
Ingo Molnar
9f13acb240 Linux 6.15-rc1
-----BEGIN PGP SIGNATURE-----
 
 iQFSBAABCAA8FiEEq68RxlopcLEwq+PEeb4+QwBBGIYFAmfy3/YeHHRvcnZhbGRz
 QGxpbnV4LWZvdW5kYXRpb24ub3JnAAoJEHm+PkMAQRiG/ygIAItY5dzf5fVnVEPy
 UrF+EzIaWGWRw3N+41AyT5X7z77FPX7E0cA6MD4KxfWW/OYzeAoeZSyrM2xIsEh3
 26qiohvJjpHjfHdzvKmxNItvW8+xBv3km00U/CWWqJo89JsIVnJtrSBHOut2/gNp
 f6sGoOrrR4GXXz8JX3yG/pmizr23lN81ZkVdz0ayYEK4uY92hSsBspvyFWcdffgF
 o8NCtR+JVGac8xm+f3VPSLyunLMXsh8NWETumMHP6tHQif36I3BQqeU8DgXCgjEK
 pfZ8gEyRtXIKbEt+qniUetT+2Cwu/lAN2GjTu0LqIe9Ro3HzjtotwQdk5h6kC+Lc
 BogxIs8=
 =bf5G
 -----END PGP SIGNATURE-----

Merge tag 'v6.15-rc1' into x86/cpu, to refresh the branch with upstream changes

Signed-off-by: Ingo Molnar <mingo@kernel.org>
2025-04-11 11:13:27 +02:00
Nikolay Borisov
23a76739d6 x86/alternatives: Make smp_text_poke_batch_process() subsume smp_text_poke_batch_finish()
Simplify the alternatives interface some more by moving the
poke_batch_finish check into poke_batch_process and renaming the latter.
The net effect is one less function name to consider when reading the
code.

Signed-off-by: Nikolay Borisov <nik.borisov@suse.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-54-mingo@kernel.org
2025-04-11 11:01:35 +02:00
Ingo Molnar
4f9534719e x86/alternatives: Add comment about noinstr expectations
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-53-mingo@kernel.org
2025-04-11 11:01:35 +02:00
Ingo Molnar
023f42dd59 x86/alternatives: Rename 'apply_relocation()' to 'text_poke_apply_relocation()'
Join the text_poke_*() API namespace.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-52-mingo@kernel.org
2025-04-11 11:01:35 +02:00
Ingo Molnar
dac0d75427 x86/alternatives: Update the comments in smp_text_poke_batch_process()
- Capitalize 'INT3' consistently,

 - make it clear that 'sync cores' means an SMP sync to all CPUs,

 - fix typos and spelling.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-51-mingo@kernel.org
2025-04-11 11:01:35 +02:00
Ingo Molnar
2c373ca064 x86/alternatives: Remove 'smp_text_poke_batch_flush()'
It only has a single user left, merge it into smp_text_poke_batch_add()
and remove the helper function.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-50-mingo@kernel.org
2025-04-11 11:01:35 +02:00
Ingo Molnar
b1bb39185d x86/alternatives: Move declarations of vmlinux.lds.S defined section symbols to <asm/alternative.h>
Move it from the middle of a .c file next to the similar declarations
of __alt_instructions[] et al.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-49-mingo@kernel.org
2025-04-11 11:01:35 +02:00
Ingo Molnar
db5c68c88c x86/alternatives: Simplify the #include section
We accumulated lots of unnecessary header inclusions over the years,
trim them.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-48-mingo@kernel.org
2025-04-11 11:01:35 +02:00
Ingo Molnar
3c8454dfc9 x86/alternatives: Rename 'POKE_MAX_OPCODE_SIZE' to 'TEXT_POKE_MAX_OPCODE_SIZE'
Join the TEXT_POKE_ namespace.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-47-mingo@kernel.org
2025-04-11 11:01:35 +02:00
Ingo Molnar
8036fbe5a5 x86/alternatives: Rename 'TP_ARRAY_NR_ENTRIES_MAX' to 'TEXT_POKE_ARRAY_MAX'
Standardize on TEXT_POKE_ namespace for CPP constants too.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-46-mingo@kernel.org
2025-04-11 11:01:35 +02:00
Ingo Molnar
22b9662313 x86/alternatives: Standardize on 'tpl' local variable names for 'struct smp_text_poke_loc *'
There's no toilet paper in this code.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-45-mingo@kernel.org
2025-04-11 11:01:35 +02:00
Ingo Molnar
3e6f47573e x86/alternatives: Simplify and clean up patch_cmp()
- No need to cast over to 'struct smp_text_poke_loc *', void * is just fine
  for a binary search,

- Use the canonical (a, b) input parameter nomenclature of cmp_func_t
  functions and rename the input parameters from (tp, elt) to
  (tpl_a, tpl_b).

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-44-mingo@kernel.org
2025-04-11 11:01:35 +02:00
Ingo Molnar
6af9540379 x86/alternatives: Constify text_poke_addr()
This will also allow the simplification of patch_cmp().

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-43-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
0e67e587e2 x86/alternatives: Simplify text_poke_addr_ordered()
- Use direct 'void *' pointer comparison, there's no
   need to force the type to 'unsigned long'.

 - Remove the 'tp' local variable indirection

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-42-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
6e4955a9d7 x86/alternatives: Rename 'text_poke_sync()' to 'smp_text_poke_sync_each_cpu()'
Unlike sync_core(), text_poke_sync() is a very heavy operation, as
it sends an IPI to every online CPU in the system and waits for
completion.

Reflect this in the name.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-41-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
7fbadb50d9 x86/alternatives: Move text_poke_array completion from smp_text_poke_batch_finish() and smp_text_poke_batch_flush() to smp_text_poke_batch_process()
Simplifies the code and improves code generation a bit:

   text	   data	    bss	    dec	    hex	filename
  14769	   1017	   4112	  19898	   4dba	alternative.o.before
  14742	   1017	   4112	  19871	   4d9f	alternative.o.after

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-40-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
cca3473956 x86/alternatives: Add documentation for smp_text_poke_batch_add()
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-39-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
9647ce4652 x86/alternatives: Document 'smp_text_poke_single()'
Extend the documentation to better describe its purpose.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-38-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
8a6a1b4e0e x86/alternatives: Remove the mixed-patching restriction on smp_text_poke_single()
At this point smp_text_poke_single(addr, opcode, len, emulate) is equivalent to:

	smp_text_poke_batch_add(addr, opcode, len, emulate);
	smp_text_poke_batch_finish();

So remove the restriction on mixing single-instruction patching
with multi-instruction patching.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-37-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
0e351aec2b x86/alternatives: Move the text_poke_array manipulation into text_poke_int3_loc_init() and rename it to __smp_text_poke_batch_add()
This simplifies the code and code generation a bit:

   text	   data	    bss	    dec	    hex	filename
  14802	   1029	   4112	  19943	   4de7	alternative.o.before
  14784	   1029	   4112	  19925	   4dd5	alternative.o.after

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-36-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
74e8e2bf95 x86/alternatives: Simplify smp_text_poke_batch_process()
This function is now using the text_poke_array state exclusively,
make that explicit by removing the redundant input parameters.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-34-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
8e35752f0c x86/alternatives: Simplify smp_text_poke_int3_handler()
Remove the 'desc' local variable indirection and use
text_poke_array directly.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-33-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
b6a25841c1 x86/alternatives: Simplify try_get_text_poke_array()
There's no need to return a pointer on success - it's always
the same pointer.

Return a bool instead.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-32-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
3916eec516 x86/alternatives: Rename 'put_desc()' to 'put_text_poke_array()'
Just like with try_get_text_poke_array(), this name better reflects
what the underlying code is doing, there's no 'descriptor'
indirection anymore.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-31-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
46f3d9d329 x86/alternatives: Rename 'try_get_desc()' to 'try_get_text_poke_array()'
This better reflects what the underlying code is doing,
there's no 'descriptor' indirection anymore.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-30-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
0494b16b9c x86/alternatives: Remove the tp_vec indirection
At this point we are always working out of an uptodate
text_poke_array, there's no need for smp_text_poke_int3_handler()
to read via the int3_vec indirection - remove it.

This simplifies the code:

   1 file changed, 5 insertions(+), 15 deletions(-)

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-29-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
6e7dc03aee x86/alternatives: Introduce 'struct smp_text_poke_array' and move tp_vec and tp_vec_nr to it
struct text_poke_array is an equivalent structure to these global variables:

	static struct smp_text_poke_loc tp_vec[TP_VEC_MAX];
	static int tp_vec_nr;

Note that we intentionally mirror much of the naming of
'struct text_poke_int3_vec', which will further highlight
the unecessary layering going on in this code, and will
ease its removal.

No change in functionality.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-28-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
37725b64a9 x86/alternatives: Assert input parameters in smp_text_poke_batch_process()
At this point the 'tp' input parameter must always be the
global 'tp_vec' array, and 'nr_entries' must always be equal
to 'tp_vec_nr'.

Assert these conditions - which will allow the removal of
a layer of indirection between these values.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-27-mingo@kernel.org
2025-04-11 11:01:34 +02:00
Ingo Molnar
476ad071c6 x86/alternatives: Assert that smp_text_poke_int3_handler() can only ever handle 'tp_vec[]' based requests
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juergen Gross <jgross@suse.com>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250411054105.2341982-26-mingo@kernel.org
2025-04-11 11:01:34 +02:00