The maximum link bpp - determined by the link BW for instance - can be
fractional, handle this properly during computing the link bpp on SST.
This keeps the pipe joiner specific maximum link bpp as a rounded-down
integer value still, changing that to a fractional value is left for
later.
v2: Align the min/max bpp value to the bpp step.
Reviewed-by: Luca Coelho <luciano.coelho@intel.com>
Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://lore.kernel.org/r/20250509180340.554867-8-imre.deak@intel.com
Adjusting the compressed bpp range min/max limits in
intel_dp_dsc_nearest_valid_bpp() is unnecessary:
- The source/sink min/max values are enforced already by the
link_config_limits::min_bpp_x16/max_bpp_x16 values computed early in
intel_dp_compute_config_link_bpp_limits().
- The fixed set of valid bpps are enforced already - for all bpps in the
min .. max range by intel_dp_dsc_valid_compressed_bpp() called from
intel_dp_mtp_tu_compute_config().
The only thing needed is limiting max compressed bpp below the
uncompressed pipe bpp, do that one thing only instead of calling
intel_dp_dsc_nearest_valid_bpp().
Reviewed-by: Luca Coelho <luciano.coelho@intel.com>
Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://lore.kernel.org/r/20250509180340.554867-7-imre.deak@intel.com
Check the BW requirement of a selected compressed bpp against the total
MST link BW early. This didn't cause a problem, since all the BW limits
within the MST topology are checked during the later MST topology BW
check. However it doesn't make sense to defer the total link BW check,
since for resolving a BW limit issue due to this later also (a) requires
selecting a pipe to reduce its bpp, ending up reducing the bpp for
another pipe, which is not ideal (b) requires recomputing the state for
all CRTC/stream's in the topology which may slow down the commit
considerably (especially when using fractional bpps).
Based on the above, check a stream bpp's BW requirement against the MST
link's total BW early.
Ideally drm_dp_atomic_find_time_slots() should check internally the
corresponding PBN/TU slot BW against the total link BW, returning an
error if the check fails, however that change would also affect other
drivers, so leaving this for a follow-up.
v2: Rephrase description of pipe selection/bpp reduction in commit
message. (Ankit)
Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://lore.kernel.org/r/20250509180340.554867-6-imre.deak@intel.com
A follow up change will check a selected bpp's BW requirement in
intel_dp_mtp_tu_compute_config(), however that requires the total link
slot count to be up-to-date. The latter in turn depends on the channel
encoding and hence the link rate used, so it can be set after the
link rate used is selected.
This also allows simplifying mst_stream_update_slots(), do that as well,
moving the function definition before its use.
Cc: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Reviewed-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://lore.kernel.org/r/20250509180340.554867-5-imre.deak@intel.com
Atm TGL supports only a fixed set of valid DSC compressed bpps
(6,8,10,12,15), but this is not taken into account while looking for a
bpp in the minimum..maximum compressed bpp range.
This happened to work only by chance since atm from the above min..max
range it's always the maximum bpp that is selected, which is one of the
above valid bpps (see mst_stream_dsc_compute_link_config() ->
intel_dp_dsc_nearest_valid_bpp()). Before selecting a bpp however, the
bpp's BW requirement should be checked wrt. to the MST total link BW;
after doing that - in a follow-up change - the validity of any bpp in
the min..max range must be ensured before the bpp is selected, do that
here.
Cc: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Reviewed-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://lore.kernel.org/r/20250509180340.554867-4-imre.deak@intel.com
A follow-up change wants to skip invalid bpps in the bpp select loop of
a stream state computation. To allow for that, using the usual
'continue' statement in the loop, change the way the single-bpp range is
handled.
v2: Fix typo in commit message. (Ankit)
Cc: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Luca Coelho <luciano.coelho@intel.com>
Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://lore.kernel.org/r/20250509180340.554867-3-imre.deak@intel.com
Atm, on an MST link in DSC mode
intel_dp_compute_config_link_bpp_limits() calculates the maximum link
bpp limit using the MST root connector's DSC capabilities. That's not
correct in general: the decompression could be performed by a branch
device downstream of the root branch device or the sink itself.
Fix the above by passing to intel_dp_compute_config_link_bpp_limits()
the actual connector being modeset, containing the correct DSC
capabilities.
Cc: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Fixes: 1c5b72daff ("drm/i915/dp: Set the DSC link limits in intel_dp_compute_config_link_bpp_limits")
Reviewed-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
Reviewed-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://lore.kernel.org/r/20250509180340.554867-2-imre.deak@intel.com
- New board support: TQMa8XxS, TQMa95xxSA, TQMa93xx, MBa91xxCA,
i.MX943 EVK, Nitrogen8M Plus ENC Carrier, Toradex SMARC i.MX8MP,
Libra-i.MX 8M Plus FPSC board
- A couple of imx8mp-tqma8mpql-mba8mp-ras314 board updates that support
Raspberry Pi Camera V2 and LVDS using device tree overlay
- A series from Adam Ford that updates i.MX8M Beacon boards for RTC
capacitive load, HDMI audio, Ethernet PHY, etc.
- A set of changes from Daniel Baluta that enables i.MX8MP DSP node
for rproc usage
- A few changes from Francesco Dolcini that add EEPROM compatible
fallback for imx8mp-verdin board, add fan PWM configuation for
imx8mp-toradex-smarc board
- A series from Frank Li to enable PCIe EP support all i.MX8 devices
using device tree overlay
- A change from Laurentiu Mihalcea to enable Sound Open Firmware (SOF)
support on imx95-19x19-evk board
- A few changes from Markus Niebel to disable MDIO Open Drain for
imx93-tqma9352 devices
- A couple of changes from Max Krummenacher to enable PCIe and SATA
support for i.MX8 Apalis and Colibri boards
- A series from Primoz Fiser to enable various devices/functions for
i.MX93 phycore boards
- A patch set from Xu Yang to add USB2.0 support for i.MX95 EVK boards
-----BEGIN PGP SIGNATURE-----
iQFIBAABCgAyFiEEFmJXigPl4LoGSz08UFdYWoewfM4FAmghzEsUHHNoYXduZ3Vv
QGtlcm5lbC5vcmcACgkQUFdYWoewfM7o7gf+K6yNLsEH7fJ/L1faCYfElURJXg+f
X3nfCJwlC7rHAa7yFTb8U2Fu/1Tl4Xm8EtBr8RYIiPq3i+K80jRk/Ns9g/YJPIAT
XrPFVXI9OFWKrYazZVcMvxIDSDS6oB+HjugKb4iJGqktDDvvC+hVc1dzjvyTua39
N0sU3vOUDMPVfmzUFuedfH181M8MLe2gK3dM8t+5c8asVSogP6XElZG+nLF/2djm
/ZuBTUePrMLLnVWTSNzbmYZcFxMoK34Bl87yoin47S1zALL9gfxS/VMXFXiJsQ8Q
c0be/siJQlwSCzmkvrDzuU7Nw5mEboNHM1e7bDEMfdtrZfkA3b7z0bTIVQ==
=8sq6
-----END PGP SIGNATURE-----
gpgsig -----BEGIN PGP SIGNATURE-----
iQIzBAABCgAdFiEEiK/NIGsWEZVxh/FrYKtH/8kJUicFAmgh56YACgkQYKtH/8kJ
Uic7eg//VbLKZM1GQH52y0LRQCpklQRG4l3PjiOJyMAsFEf0GTTEPgNiRNiSv5WS
kXlsojsS0QulOZdRb2v84fOFmffaG7pq0o4dgN4+iWSbGUvWLb+PM46CMi39DeOD
lqgRlAVdKkOLxvDnIAEq7ce6/C4EIoIYul4D9CXLF7Jnf/NnGpB5lv8M+Sh1nv5D
xurEzUlbNuSBbxkw0mgoeiqXaGO3sFeILEV6F67np1Z6QvgeJ6nNDrX5Dd7FlwwV
eqMWpXMwGHLFok0RYWlBCpFO//hrjQi4T9y+weKEnzI1TRIiDWxH1kUWdNMpGQ5l
EcbpT42zaQamnqgbC25K+dqpSyZLLWSXZb/DkbaFvDY++riXRIEU9qOrWxBkQ73V
AWjldmzI8h1mKJKPkvHeo8gY4lmm9aaQSlzELyC8zYo8HJkcsyTQCG31uc/50S0z
765SoGd4eFTyC1aqwd01sJTqIJpi4nTAObWdCf7ptjaBFrQjliUgjTwbRY0Y8IFr
erk86QJJw9ti2RdVz4ygKjiwpin7ROBvgzNkjh8Ud/DKYhKX6BAMXKcz38MOYoV/
jg5V2EJt2U02XzdP2C8pSe1TNlaibMDgvR+KSs/8hlIvokhX/HNm01ET2y/EONbi
7M0s7sYN4Ii8X39bG5BOh/wRCZo9sLfsdHA2buOWwlQ5NgvyHWM=
=A7Fy
-----END PGP SIGNATURE-----
Merge tag 'imx-dt64-6.16' of https://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into soc/dt
i.MX arm64 device tree change for 6.16:
- New board support: TQMa8XxS, TQMa95xxSA, TQMa93xx, MBa91xxCA,
i.MX943 EVK, Nitrogen8M Plus ENC Carrier, Toradex SMARC i.MX8MP,
Libra-i.MX 8M Plus FPSC board
- A couple of imx8mp-tqma8mpql-mba8mp-ras314 board updates that support
Raspberry Pi Camera V2 and LVDS using device tree overlay
- A series from Adam Ford that updates i.MX8M Beacon boards for RTC
capacitive load, HDMI audio, Ethernet PHY, etc.
- A set of changes from Daniel Baluta that enables i.MX8MP DSP node
for rproc usage
- A few changes from Francesco Dolcini that add EEPROM compatible
fallback for imx8mp-verdin board, add fan PWM configuation for
imx8mp-toradex-smarc board
- A series from Frank Li to enable PCIe EP support all i.MX8 devices
using device tree overlay
- A change from Laurentiu Mihalcea to enable Sound Open Firmware (SOF)
support on imx95-19x19-evk board
- A few changes from Markus Niebel to disable MDIO Open Drain for
imx93-tqma9352 devices
- A couple of changes from Max Krummenacher to enable PCIe and SATA
support for i.MX8 Apalis and Colibri boards
- A series from Primoz Fiser to enable various devices/functions for
i.MX93 phycore boards
- A patch set from Xu Yang to add USB2.0 support for i.MX95 EVK boards
* tag 'imx-dt64-6.16' of https://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux: (71 commits)
arm64: dts: freescale: Add PHYTEC phyBOARD-Nash-i.MX93 support
arm64: dts: freescale: imx8mp-toradex-smarc: use generic gpio node name
arm64: dts: freescale: imx8mp-toradex-smarc: add gpio expander
arm64: dts: freescale: imx8mp-toradex-smarc: add embedded controller
arm64: dts: freescale: imx8mp-toradex-smarc: add fan PWM configuration
arm64: dts: imx93-tqma9352-mba91xxca: disable Open Drain for MDIO
arm64: dt: imx95: Add TQMa95xxSA
arm64: dts: imx: Align wifi node name with bindings
arm64: dts: freescale: add initial device tree for TQMa8XxS
arm64: dts: imx8mp-tqma8mpql-mba8mp-ras314: Add Raspberry Pi Camera V2 overlay
arm64: dts: imx8mp-tqma8mpql-mba8mp-ras314: Add LVDS device tree overlay
arm64: dts: freescale: Add minimal dts support for imx943 evk
arm64: dts: freescale: Add basic dtsi for imx943
arm64: dts: imx8-colibri: Add PCIe support
arm64: dts: freescale: imx93-phyboard-segin: Order node alphabetically
arm64: dts: freescale: imx93-phyboard-segin: Add EQOS Ethernet
arm64: dts: freescale: imx93-phyboard-segin: Add I2S audio
arm64: dts: freescale: imx93-phyboard-segin: Add USB support
arm64: dts: freescale: imx93-phyboard-segin: Add CAN support
arm64: dts: freescale: imx93-phyboard-segin: Add RTC support
...
Link: https://lore.kernel.org/r/20250512103858.50501-4-shawnguo@kernel.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
- A series from Alexander Stein that updates ls1021a-tqmals1021a device
tree mostly for display support inlcuding HDMI, LVDS and CDTech panel
- A change from Dario Binacchi to use pad config defines for i.MX23/28
device trees
- A i.MX7D change from Efe Can İçöz to include min and max voltage in
opp-microvolt
- A few patches from Fabio Estevam to fix dt-schema warnings
- A couple of changes from Krzysztof Kozlowski to align NAND and WIFI
node name with binings
- A clean-up change from Rob Herring to drop redundant CPU
"clock-latency"
-----BEGIN PGP SIGNATURE-----
iQFIBAABCgAyFiEEFmJXigPl4LoGSz08UFdYWoewfM4FAmghwGUUHHNoYXduZ3Vv
QGtlcm5lbC5vcmcACgkQUFdYWoewfM6e6AgAudNxNY08NmcooFSWVc/b50X86M1n
zJUBlrgMQW6/C+JcjfqFSiPUa2tVjRXKQ0w3vyBfK5tXdPmWjxBK+It91fPCPHt6
U95diIs5RyrKuueTG4VBk1IEDQ7CO9CQueYezeAz7/ysv5kpDHLlLgFq0XZ1VrNK
RFfQjuU9ZWG5seV2Ji5D8FXT3JksTgXHSVZtV/qcjN2hDzs0wu0GwP4ShbuadZoW
TECM8GJbdHCv0A79litox8mkCK8+qmqXyU6SR4SXGVrIiuSWQkziB4VzoC6UbKfa
fWXpGmuCYZW5ohFxc4cKI/Bl+Fk0b6mbVfumM5tTwmUhtjOPHiP6Zsby3Q==
=OpE5
-----END PGP SIGNATURE-----
gpgsig -----BEGIN PGP SIGNATURE-----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=4zTX
-----END PGP SIGNATURE-----
Merge tag 'imx-dt-6.16' of https://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into soc/dt
i.MX ARM device tree change for 6.16:
- A series from Alexander Stein that updates ls1021a-tqmals1021a device
tree mostly for display support inlcuding HDMI, LVDS and CDTech panel
- A change from Dario Binacchi to use pad config defines for i.MX23/28
device trees
- A i.MX7D change from Efe Can İçöz to include min and max voltage in
opp-microvolt
- A few patches from Fabio Estevam to fix dt-schema warnings
- A couple of changes from Krzysztof Kozlowski to align NAND and WIFI
node name with binings
- A clean-up change from Rob Herring to drop redundant CPU
"clock-latency"
* tag 'imx-dt-6.16' of https://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
ARM: dts: mxs: use padconfig macros
ARM: dts: imx7d: update opp-table voltages
ARM: dts: nxp: Align wifi node name with bindings
ARM: dts: imx6q-apalis: remove pcie-switch node
ARM: dts: ls1021a-tqmals1021a: change sound card model name
ARM: dts: ls1021a-tqmals1021a: Add overlay for CDTech DC44 RGB display
ARM: dts: ls1021a-tqmals1021a: Add overlay for CDTech FC21 RGB display
ARM: dts: ls1021a-tqmals1021a: Add LVDS overlay for Tianma TM070JVGH33
ARM: dts: ls1021a-tqmals1021a: Add HDMI overlay
ARM: dts: ls1021a-tqmals1021a: Add vcc-supply for spi-nor
ARM: dts: ls1021a-tqmals1021a: Fix license
ARM: dts: imx: Drop redundant CPU "clock-latency"
ARM: dts: imx51-digi-connectcore-som: Fix MMA7455 compatible
ARM: dts: nxp: Align NAND controller node name with bindings
ARM: dts: imx: Fix the iim compatible string
ARM: dts: imx31/imx6: Use flash as the NOR node name
Link: https://lore.kernel.org/r/20250512103858.50501-3-shawnguo@kernel.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
- A series from Peng Fan to dump full 128-bits UID for i.MX8MP
-----BEGIN PGP SIGNATURE-----
iQFIBAABCgAyFiEEFmJXigPl4LoGSz08UFdYWoewfM4FAmghimgUHHNoYXduZ3Vv
QGtlcm5lbC5vcmcACgkQUFdYWoewfM5aJggAlDCcQsA4z24IPWb8FRKWb4r48sSv
4Jpn1lkr42rIkGO/VwuBuPRbRWCz+zBulT9YDfLanSNUj5BCjhfHg+EUCjoy/KWz
pdXNeRCQyQs9M4n0X6tyyRtNDhp/z9e7K5ZironWcJ8zKIvhtFqbzjh6fywne1Gl
zc8rcrdgh44FsvRjrOudv1fQo7aVcURDnFAZMKT6m6VYtMa7dwb2zm+UETxvrOH/
6GQst0q6bqG+SLljTjVGplUG0cV5gWvsAbO2ilTTq56bSXGaKNeZPJa2kNq5XCsA
cb/OEJMBsVcM7YgpKr/Vl3/uqIamCKdMCt8QofZkj7muUVJrp6QxDsiafg==
=J+MT
-----END PGP SIGNATURE-----
gpgsig -----BEGIN PGP SIGNATURE-----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=yyBZ
-----END PGP SIGNATURE-----
Merge tag 'imx-drivers-6.16' of https://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into soc/drivers
i.MX drivers change for 6.16:
- A series from Peng Fan to dump full 128-bits UID for i.MX8MP
* tag 'imx-drivers-6.16' of https://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
soc: imx8m: Dump higher 64bits UID
soc: imx8m: Introduce soc_uid hook
soc: imx8m: Cleanup with adding imx8m_soc_[un]prepare
Link: https://lore.kernel.org/r/20250512103858.50501-1-shawnguo@kernel.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
When restarting a CPU powered by the PCA9450 power management IC, it
is beneficial to use the PCA9450 to power cycle the CPU and all its
connected peripherals to start up in a known state. The PCA9450 features
a cold start procedure initiated by an I2C command.
Add a restart handler so that the PCA9450 is used to restart the CPU.
The restart handler sends command 0x14 to the SW_RST register,
initiating a cold reset (Power recycle all regulators except LDO1, LDO2
and CLK_32K_OUT)
As the PCA9450 is a PMIC specific for the i.MX8M family CPU, the restart
handler priority is set just slightly higher than imx2_wdt and the PSCI
restart handler. This makes sure this restart handler takes precedence.
Signed-off-by: Paul Geurts <paul.geurts@prodrive-technologies.com>
Link: https://patch.msgid.link/20250505115936.1946891-1-paul.geurts@prodrive-technologies.com
Signed-off-by: Mark Brown <broonie@kernel.org>
We have below 2 functions, but these are very similar
(A) snd_soc_unregister_component_by_driver()
(B) snd_soc_unregister_component()
(A) void snd_soc_unregister_component_by_driver(...)
{
...
(a) mutex_lock(&client_mutex);
^ (X) component = snd_soc_lookup_component_nolocked(dev, component_driver->name);
| if (!component) ^^^^^^^^^^^^^^^^^^^^^^
| goto out;
(b)
| snd_soc_del_component_unlocked(component);
v
out:
(c) mutex_unlock(&client_mutex);
}
(B) void snd_soc_unregister_component_by_driver(...)
{
(a) mutex_lock(&client_mutex);
^ while (1) {
| (X) struct snd_soc_component *component = snd_soc_lookup_component_nolocked(dev, NULL);
| ^^^^
(b) if (!component)
| break;
|
| snd_soc_del_component_unlocked(component);
v }
(c) mutex_unlock(&client_mutex);
}
Both are calling lock (a), find component and remove it (b), and
unlock (c). The big diff is whether use driver name for lookup() or
not (X).
Merge these into snd_soc_unregister_component_by_driver() (B), and
snd_soc_unregister_component_by_driver() (A) can be macro.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://patch.msgid.link/87h61qy2vn.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
There is a chance that obj->dma_buf would be NULL by the time
virtgpu_dma_buf_free_obj() is called. This can happen for imported
prime objects, when drm_gem_object_exported_dma_buf_free() gets
called on them before drm_gem_object_free(). This is because
drm_gem_object_exported_dma_buf_free() explicitly sets
obj->dma_buf to NULL.
Therefore, fix this issue by storing the dma_buf pointer in the
virtio_gpu_object instance and using it in virtgpu_dma_buf_free_obj.
This stored pointer is guaranteed to be valid until the object is
freed as we took a reference on it in virtgpu_gem_prime_import().
Fixes: 415cb45895 ("drm/virtio: Use dma_buf from GEM object instance")
Cc: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Signed-off-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Link: https://lore.kernel.org/r/20250501232419.180337-1-vivek.kasireddy@intel.com
These are updates from Marek Behún for the cznic platform drivers:
This series adds support for generating ECDSA signatures with hardware
stored private key on Turris Omnia and Turris MOX.
This ability is exposed via the keyctl() syscall.
* 'cznic/platform' of https://git.kernel.org/pub/scm/linux/kernel/git/soc/soc:
platform: cznic: use ffs() instead of __bf_shf()
firmware: turris-mox-rwtm: fix building without CONFIG_KEYS
platform: cznic: fix function parameter names
firmware: turris-mox-rwtm: Add support for ECDSA signatures with HW private key
firmware: turris-mox-rwtm: Drop ECDSA signatures via debugfs
platform: cznic: turris-omnia-mcu: Add support for digital message signing with HW private key
platform: cznic: Add keyctl helpers for Turris platform
platform: cznic: turris-omnia-mcu: Refactor requesting MCU interrupt
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Handle soc servicing events which require the rdma auxiliary device resources to
be cleaned up during a suspend, and re-initialized during a resume.
Signed-off-by: Shiraz Saleem <shirazsaleem@microsoft.com>
Signed-off-by: Konstantin Taranov <kotaranov@microsoft.com>
Link: https://patch.msgid.link/1746633545-17653-5-git-send-email-kotaranov@linux.microsoft.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
When suspending, save_processor_state() calls mtrr_save_fixed_ranges()
to save fixed-range MTRRs.
On platforms without fixed-range MTRRs like the ACRN hypervisor which
has removed fixed-range MTRR emulation, accessing these MSRs will
trigger an unchecked MSR access error. Make sure fixed-range MTRRs are
supported before access to prevent such error.
Since mtrr_state.have_fixed is only set when MTRRs are present and
enabled, checking the CPU feature flag in mtrr_save_fixed_ranges() is
unnecessary.
Fixes: 3ebad59056 ("[PATCH] x86: Save and restore the fixed-range MTRRs of the BSP when suspending")
Signed-off-by: Jiaqing Zhao <jiaqing.zhao@linux.intel.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20250509170633.3411169-2-jiaqing.zhao@linux.intel.com
Allow mana_ib to be created over ethernet gdma device and
over rnic gdma device. The HW has two devices with different
capabilities and different use-cases. Initialize required
resources depending on the used gdma device.
Signed-off-by: Konstantin Taranov <kotaranov@microsoft.com>
Link: https://patch.msgid.link/1746633545-17653-3-git-send-email-kotaranov@linux.microsoft.com
Reviewed-by: Long Li <longli@microsoft.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Initialize gdma device for rdma inside mana module.
For each gdma device, initialize an auxiliary ib device.
Signed-off-by: Konstantin Taranov <kotaranov@microsoft.com>
Link: https://patch.msgid.link/1746633545-17653-2-git-send-email-kotaranov@linux.microsoft.com
Reviewed-by: Long Li <longli@microsoft.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The health of a given battery is exposed over the Dell DDV WMI
interface using the "BatteryManufactureAceess" WMI method. The
resulting data contains, among other data, the health status of
the battery.
Expose this value to userspace using the power supply extension
interface.
Tested on a Dell Inspiron 3505.
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20250429003606.303870-4-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
The manufacture date of a given battery is exposed over the Dell DDV
WMI interface using the "BatteryManufactureDate" WMI method. The
resulting data contains the manufacture date of the battery encoded
inside a 16-bit value as described in the Smart Battery Data
Specification.
Expose this value to userspace using the power supply extension
interface.
Tested on a Dell Inspiron 3505.
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20250429003606.303870-3-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Since commit db0a507cb2 ("ACPICA: Update integer-to-hex-string
conversions") the battery serial number is no longer distorted,
allowing us to finally implement the battery matching algorithm.
The battery matchign algorithm is necessary when translating between
ACPI batteries and the associated indices used by the WMI interface
based on the battery serial number. Since this serial number can only
be retrieved when the battery is present we cannot perform the initial
translation inside dell_wmi_ddv_add_battery() because the ACPI battery
might be absent at this point in time.
Introduce dell_wmi_ddv_battery_translate() which implements the
battery matching algorithm and replaces dell_wmi_ddv_battery_index().
Also implement a translation cache for caching previous translations
between ACPI batteries and indices. This is necessary because
performing a translation can be very expensive.
Tested on a Dell Inspiron 3505.
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Link: https://lore.kernel.org/r/20250429003606.303870-2-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Some batteries can signal when an internal fuse was blown. In such a
case POWER_SUPPLY_HEALTH_DEAD is too vague for userspace applications
to perform meaningful diagnostics.
Additionally some batteries can also signal when some of their
internal cells are imbalanced. In such a case returning
POWER_SUPPLY_HEALTH_UNSPEC_FAILURE is again too vague for userspace
applications to perform meaningful diagnostics.
Add new health status values for both cases.
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.com>
Link: https://lore.kernel.org/r/20250429003606.303870-1-W_Armin@gmx.de
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Make frequently fetched telemetry available via sysfs. These parameters
do not fit in hwmon sensor model, hence make them available via sysfs.
Create following sysfs files per acpi device node.
* c0_residency_input
* prochot_status
* smu_fw_version
* protocol_version
* ddr_max_bw(GB/s)
* ddr_utilised_bw_input(GB/s)
* ddr_utilised_bw_perc_input(%)
* mclk_input(MHz)
* fclk_input(MHz)
* clk_fmax(MHz)
* clk_fmin(MHz)
* cclk_freq_limit_input(MHz)
* pwr_current_active_freq_limit(MHz)
* pwr_current_active_freq_limit_source
Signed-off-by: Suma Hegde <suma.hegde@amd.com>
Reviewed-by: Naveen Krishna Chatradhi <naveenkrishna.chatradhi@amd.com>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Link: https://lore.kernel.org/r/20250506101542.200811-3-suma.hegde@amd.com
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
The use of the ternary operator on rv is redundant, rv is
either the initialized value of 0 or a negative error return
code, so it can never be greater than zero, and hence the
zero assignment in ternary operator is redundant. Just return
rv instead.
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Link: https://patch.msgid.link/20250507131834.253823-1-colin.i.king@gmail.com
Acked-by: Bernard Metzler <bmt@zurich.ibm.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Expose power reading and power limits via hwmon power sensors.
Signed-off-by: Suma Hegde <suma.hegde@amd.com>
Reviewed-by: Naveen Krishna Chatradhi <naveenkrishna.chatradhi@amd.com>
Link: https://lore.kernel.org/r/20250506101542.200811-2-suma.hegde@amd.com
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Create separate functions for the implicit ODP initialization
which is different from the explicit ODP initialization.
Tested-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Reuse newly added DMA API to cache IOVA and only link/unlink pages
in fast path for UMEM ODP flow.
Tested-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
As a preparation to remove dma_list, store access mask in PFN pointer
and not in dma_addr_t.
Tested-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
HMM callers use PFN list to populate range while calling
to hmm_range_fault(), the conversion from PFN to DMA address
is done by the callers with help of another DMA list. However,
it is wasteful on any modern platform and by doing the right
logic, that DMA list can be avoided.
Provide generic logic to manage these lists and gave an interface
to map/unmap PFNs to DMA addresses, without requiring from the callers
to be an experts in DMA core API.
Tested-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Introduce new sticky flag (HMM_PFN_DMA_MAPPED), which isn't overwritten
by HMM range fault. Such flag allows users to tag specific PFNs with
information if this specific PFN was already DMA mapped.
Tested-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Amir Goldstein <amir73il@gmail.com> says:
This adds a test for fanotify mount ns notifications inside userns [1].
While working on the test I ended up making lots of cleanups to reduce
build dependency on make headers_install.
These patches got rid of the dependency for my kvm setup for the
affected filesystems tests.
Building with TOOLS_INCLUDES dir was recommended by John Hubbard [2].
NOTE #1: these patches are based on a merge of vfs-6.16.mount
(changes wrappers.h) into v6.15-rc5 (changes mount-notify_test.c),
so if this cleanup is acceptable, we should probably setup a selftests
branch for 6.16, so that it can be used to test the fanotify patches.
NOTE #2: some of the defines in wrappers.h are left for overlayfs and
mount_setattr tests, which were not converted to use TOOLS_INCLUDES.
I did not want to mess with those tests.
* patches from https://lore.kernel.org/20250509133240.529330-1-amir73il@gmail.com:
selftests/fs/mount-notify: add a test variant running inside userns
selftests/filesystems: create setup_userns() helper
selftests/filesystems: create get_unique_mnt_id() helper
selftests/fs/mount-notify: build with tools include dir
selftests/mount_settattr: remove duplicate syscall definitions
selftests/pidfd: move syscall definitions into wrappers.h
selftests/fs/statmount: build with tools include dir
selftests/filesystems: move wrapper.h out of overlayfs subdir
Link: https://lore.kernel.org/20250509133240.529330-1-amir73il@gmail.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
unshare userns in addition to mntns and verify that:
1. watching tmpfs mounted inside userns is allowed with any mark type
2. watching orig root with filesystem mark type is not allowed
3. watching mntns of orig userns is not allowed
4. watching mntns in userns where fanotify_init was called is allowed
mount events are only tested with the last case of mntns mark.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Link: https://lore.kernel.org/20250509133240.529330-9-amir73il@gmail.com
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Add helper to utils.c and use it in statmount userns tests.
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Link: https://lore.kernel.org/20250509133240.529330-8-amir73il@gmail.com
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Add helper to utils.c and use it in mount-notify and statmount tests.
Linking with utils.c drags in a dependecy with libcap, so add it to the
Makefile of the tests.
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Link: https://lore.kernel.org/20250509133240.529330-7-amir73il@gmail.com
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Copy the fanotify uapi header files to the tools include dir
and define __kernel_fsid_t to decouple dependency with headers_install
and then remove the redundant re-definitions of fanotify macros.
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Link: https://lore.kernel.org/20250509133240.529330-6-amir73il@gmail.com
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Which are already defined in wrappers.h.
For now, the syscall defintions of mount_settattr() itself
remain in the test, which is the only test to use them.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Link: https://lore.kernel.org/20250509133240.529330-5-amir73il@gmail.com
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
There was already duplicity in some of the defintions.
Remove syscall number defintions for __ia64__ that are
both stale and incorrect.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Link: https://lore.kernel.org/20250509133240.529330-4-amir73il@gmail.com
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Copy the required headers files (mount.h, nsfs.h) to the tools
include dir and define the statmount/listmount syscall numbers
to decouple dependency with headers_install for the common cases.
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Link: https://lore.kernel.org/20250509133240.529330-3-amir73il@gmail.com
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
This is not an overlayfs specific header.
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Link: https://lore.kernel.org/20250509133240.529330-2-amir73il@gmail.com
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Filesystem too small for a journal
mount: /mnt/D/: mount failed: Operation not permitted.
mount_setattr_test.c:1076:idmap_mount_tree_invalid:Expected system("mount -o loop -t ext4 /mnt/C/ext4.img /mnt/D/") (256) == 0 (0)
Signed-off-by: Christian Brauner <brauner@kernel.org>
CC mount_setattr_test
In file included from mount_setattr_test.c:24:
mount_setattr_test.c: In function ‘mount_setattr_mount_detached_mount_on_detached_mount_and_attach’:
mount_setattr_test.c:1850:60: error: ‘STATX_MNT_ID_UNIQUE’ undeclared (first use in this function); did you mean ‘STATX_MNT_ID’?
1850 | ASSERT_EQ(statx(fd_tree_subdir, "", AT_EMPTY_PATH, STATX_MNT_ID_UNIQUE, &stx), 0);
| ^~~~~~~~~~~~~~~~~~~
../kselftest_harness.h:757:20: note: in definition of macro ‘__EXPECT’
757 | __typeof__(_expected) __exp = (_expected); \
| ^~~~~~~~~
mount_setattr_test.c:1850:9: note: in expansion of macro ‘ASSERT_EQ’
1850 | ASSERT_EQ(statx(fd_tree_subdir, "", AT_EMPTY_PATH, STATX_MNT_ID_UNIQUE, &stx), 0);
| ^~~~~~~~~
mount_setattr_test.c:1850:60: note: each undeclared identifier is reported only once for each function it appears in
1850 | ASSERT_EQ(statx(fd_tree_subdir, "", AT_EMPTY_PATH, STATX_MNT_ID_UNIQUE, &stx), 0);
| ^~~~~~~~~~~~~~~~~~~
../kselftest_harness.h:757:20: note: in definition of macro ‘__EXPECT’
757 | __typeof__(_expected) __exp = (_expected); \
| ^~~~~~~~~
mount_setattr_test.c:1850:9: note: in expansion of macro ‘ASSERT_EQ’
1850 | ASSERT_EQ(statx(fd_tree_subdir, "", AT_EMPTY_PATH, STATX_MNT_ID_UNIQUE, &stx), 0);
| ^~~~~~~~~
Signed-off-by: Christian Brauner <brauner@kernel.org>
CC mount_setattr_test
mount_setattr_test.c:176:19: error: redefinition of ‘sys_open_tree’
176 | static inline int sys_open_tree(int dfd, const char *filename, unsigned int flags)
| ^~~~~~~~~~~~~
In file included from mount_setattr_test.c:23:
../filesystems/overlayfs/wrappers.h:59:19: note: previous definition of ‘sys_open_tree’ with type ‘int(int, const char *, unsigned int)’
59 | static inline int sys_open_tree(int dfd, const char *filename, unsigned int flags)
Signed-off-by: Christian Brauner <brauner@kernel.org>
Use the BIT() and BIT_ULL() macros in the new audit code instead of
explicit shifts to improve readability. Use bitmask instead of modulo
operation to simplify code.
Add test_range1_rand15() and test_range2_rand15() KUnit tests to improve
get_id_range() coverage.
Signed-off-by: Günther Noack <gnoack3000@gmail.com>
Link: https://lore.kernel.org/r/20250512093732.1408485-1-mic@digikod.net
Signed-off-by: Mickaël Salaün <mic@digikod.net>
Due to changes in allocator, the size of the allocation for
contiguous region is not rounded up to a power-of-two and
instead allocated as is. Thus, change the part of test that
expected the allocation to fail.
Signed-off-by: Mikolaj Wasiak <mikolaj.wasiak@intel.com>
Reviewed-by: Krzysztof Karas <krzysztof.karas@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Reviewed-by: Krzysztof Niemiec <krzysztof.niemiec@intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://lore.kernel.org/r/fptevdfbclvg2lbfuys5ibffbl2baouywkutnr7vdsy5tzcqfk@mpflwlh6jxfd
