Heiko Stuebner [Sat, 3 May 2025 20:15:11 +0000 (22:15 +0200)]
phy: rockchip: inno-usb2: add phy definition for rk3036
The usb2phy on rk3036 is very similar to rk3128 but not 100% identical.
While most registers and bits are the same, a small fraction did get moved
around.
So we can re-use the phy-tuning function, but need a new set a bits.
The biggest change might be that the phy on rk3036 does not support the
charger detection, that rk3128 (and newer phys) have.
Adam Ford [Sun, 4 May 2025 20:40:42 +0000 (15:40 -0500)]
phy: freescale: fsl-samsung-hdmi: Improve LUT search for best clock
Searching the look-up-table runs so long as the frequency in the
table is at or below the desired rate. This works well in most
cases, but the next entry in the LUT might be closer to the
nominal value than the lower one. Add some logic to check
the higer value is any closer to the nominal value and use it.
There are two functions, round_rate and set_rate that duplicate
a lot of the same work, so simplify the code by creating a helper
function that will identify the phy settings for a desired clock
rate and return the structure with the corresponding settings.
>From this structure, the round_rate and set_rate can both get what
they need to achieve the clock setting closest to the desired rate
as possible while minimizing the duplicated code.
Also rename phy_clk_set_rate to fsl_samsung_hdmi_phy_clk_set_rate.
phy_clk_round_rate sounds like a generic helper function. In
reality, it is unique to the phy-fsl-samsung-hdmi. Rename
phy_clk_round_rate to fsl_samsung_hdmi_phy_clk_round_rate.
No functional change intended.
phy: renesas: phy-rcar-gen3-usb2: Sort compatible entries by SoC part number
Reorder the compatible entries in `rcar_gen3_phy_usb2_match_table` to
maintain sorting based on SoC part numbers. Keep the entries ordered
numerically while ensuring the existing generic compatible strings
remain at the bottom.
Document USB2.0 phy bindings for RZ/V2H(P) ("R9A09gG57") SoC.
RZ/V2H(P) USB2.0 phy is similar to one found on the RZ/G2L SoC, but it
needs additional configuration to be done as compared RZ/G2L USB2.0 phy.
To handle this difference a SoC specific compat string is added for
RZ/V2H(P) SoC.
Like the RZ/G2L SoC, the RZ/V2H(P) USB2.0 PHY requires the `resets`
property and has two clocks.
dt-bindings: phy: renesas,usb2-phy: Add clock constraint for RZ/G2L family
The RZ/G2L family requires two clocks for USB2 PHY, which are already
defined in the DTSI files. Add a constraint in the DT binding document
to ensure validation with `dtbs_check`.
Ivaylo Ivanov [Sun, 4 May 2025 14:45:27 +0000 (17:45 +0300)]
phy: exynos5-usbdrd: support Exynos USBDRD 3.2 4nm controller
Add support for the Exynos USB 3.2 DRD 4nm controller. It's used in
recent 4nm SoCs like Exynos2200 and Exynos2400.
This device consists of 3 underlying and independent phys: SEC link
control phy, Synopsys eUSB 2.0 and Synopsys USBDP/SS combophy. Unlike
older device designs, where the internal phy blocks were all IP of
Samsung, Synopsys phys are present. This means that the link controller
is now mapped differently to account for missing bits and registers.
The Synopsys phys also have separate register bases.
As there are non-SEC PHYs present now, it doesn't make much sense to
implement them in this driver. They are expected to be configured
by external drivers, so pass phandles to them. USBDRD3.2 link controller
set up is still required beforehand.
This commit adds the necessary changes for USB HS to work. USB SS and
DisplayPort are out of scope in this commit and will be introduced
in the future.
Ivaylo Ivanov [Sun, 4 May 2025 14:45:26 +0000 (17:45 +0300)]
phy: phy-snps-eusb2: add support for exynos2200
The Exynos2200 SoC reuses the Synopsis eUSB2 PHY IP, alongside an
external repeater, for USB 2.0. Add support for it to the existing
driver, while keeping in mind that it requires enabled more than the
reference clock.
Instead of matching frequencies with a switch and case, introduce
a table-based lookup. This improves readability, reduces redundancy,
and makes it easier to extend support for additional frequencies in
the future.
Ivaylo Ivanov [Sun, 4 May 2025 14:45:23 +0000 (17:45 +0300)]
phy: phy-snps-eusb2: make repeater optional
As described in the device tree bindings, it's not necessary for the
SNPS eUSB2 phy to be connected to a repeater. In configurations where
there are such instances, the driver probing fails and the usb
controller does not work.
Make the repeater optional to avoid that, which also lets us use
the eUSB2 phy when it's connected to a repeater that is not configurable
by the kernel (for example it's missing a driver), as long as it has
been configured beforehand (usually by the bootloader).
Ivaylo Ivanov [Sun, 4 May 2025 14:45:22 +0000 (17:45 +0300)]
phy: phy-snps-eusb2: split phy init code
The current phy init consists of hardware power-up, as well as
QCOM-specific eUSB2 init code. Split it into two parts, to make room
for such non-QCOM init code.
Ivaylo Ivanov [Sun, 4 May 2025 14:45:21 +0000 (17:45 +0300)]
phy: phy-snps-eusb2: refactor constructs names
As the driver now resides outside the phy subdirectory under a different
name, refactor all definitions, structures and functions to explicitly
specify what code is Qualcomm-specific and what is not.
Ivaylo Ivanov [Sun, 4 May 2025 14:45:20 +0000 (17:45 +0300)]
phy: move phy-qcom-snps-eusb2 out of its vendor sub-directory
As not only Qualcomm, but also Samsung is using the Synopsys eUSB2 IP
(albeit with a different register layout) in their newer SoCs, move the
driver out of its vendor sub-directory and rename it to phy-snps-eusb2.
Ivaylo Ivanov [Sun, 4 May 2025 14:45:19 +0000 (17:45 +0300)]
dt-bindings: phy: samsung,usb3-drd-phy: add exynos2200 support
Document support for Exynos2200. As the USBDRD 3.2 4nm controller
consists of Synopsys eUSB2.0 phy and USBDP/SS combophy, which will
be handled by external drivers, define only the bus clocked used
by the link controller.
Ivaylo Ivanov [Sun, 4 May 2025 14:45:18 +0000 (17:45 +0300)]
dt-bindings: phy: add exynos2200 eusb2 phy support
Document the exynos2200 eUSB2 compatible. Unlike the currently documented
Qualcomm SoCs, the driver doesn't make use of reset lines for reset
control and uses more clocks.
The description of TX_VBOOST_LVL is wrong in register PHY_CTRL3
bit[31:29].
The updated description as below:
011: Corresponds to a launch amplitude of 0.844 V.
100: Corresponds to a launch amplitude of 1.008 V.
101: Corresponds to a launch amplitude of 1.156 V.
This will fix the parsing function
phy_tx_vboost_level_from_property() to return correct value.
Fixes: 63c85ad0cd81 ("phy: fsl-imx8mp-usb: add support for phy tuning") Cc: stable@vger.kernel.org Reviewed-by: Jun Li <jun.li@nxp.com> Signed-off-by: Xu Yang <xu.yang_2@nxp.com> Link: https://lore.kernel.org/r/20250430094502.2723983-3-xu.yang_2@nxp.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
Xu Yang [Wed, 30 Apr 2025 09:45:00 +0000 (17:45 +0800)]
dt-bindings: phy: imx8mq-usb: add imx95 tuning support
The parameter value of below 3 properties are USB PHY specific. i.MX8MP
and i.MX95 USB PHY has different meanings. This will enlarge parameters
value and add constraints for them.
The ticket TKT0676370 shows the description of TX_VBOOST_LVL is wrong in
register PHY_CTRL3 bit[31:29].
011: Corresponds to a launch amplitude of 1.12 V.
010: Corresponds to a launch amplitude of 1.04 V.
000: Corresponds to a launch amplitude of 0.88 V.
After updated:
011: Corresponds to a launch amplitude of 0.844 V.
100: Corresponds to a launch amplitude of 1.008 V.
101: Corresponds to a launch amplitude of 1.156 V.
This will correct it accordingly.
Fixes: b2e75563dc39 ("dt-bindings: phy: imx8mq-usb: add phy tuning properties") Cc: stable@vger.kernel.org Reviewed-by: Jun Li <jun.li@nxp.com> Signed-off-by: Xu Yang <xu.yang_2@nxp.com> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Link: https://lore.kernel.org/r/20250430094502.2723983-1-xu.yang_2@nxp.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
phy: qcom-qusb2: reuse the IPQ6018 settings for IPQ5424
With the settings used in the commit 9c56a1de296e ("phy: qcom-qusb2: add
QUSB2 support for IPQ5424"), compliance test cases especially
eye-diagram (Host High-speed Signal Quality) tests are failing.
Reuse the IPQ6018 settings for IPQ5424 as mentioned in the Hardware
Design Document which helps to meet all the complaince requirement test
cases.
Revert "phy: qcom-qusb2: add QUSB2 support for IPQ5424"
With the current settings, compliance tests especially eye diagram
(Host High-speed Signal Quality) tests are failing. Reuse the IPQ6018
settings to overcome this issue, as mentioned in the Hardware Design
Document.
So revert the change which introduced the new settings and reuse the
IPQ6018 settings in the subsequent patch.
André Draszik [Tue, 29 Apr 2025 08:49:25 +0000 (09:49 +0100)]
phy: exynos5-usbdrd: s/FIELD_PREP_CONST/FIELD_PREP where appropriate
Commit 9b6662a0f715 ("phy: exynos5-usbdrd: use GENMASK and FIELD_PREP
for Exynos5 PHY registers") added FIELD_PREP_CONST() in many cases
where FIELD_PREP() would have been more appropriate. It also switched
existing uses of FIELD_PREP() to FIELD_PREP_CONST().
FIELD_PREP() is the preferred macro to use whenever possible while
FIELD_PREP_CONST() is meant to be used in constant initialisers.
Mike Looijmans [Mon, 28 Apr 2025 06:35:47 +0000 (08:35 +0200)]
phy-zynqmp: Postpone getting clock rate until actually needed
At probe time the driver would display the following error and abort:
xilinx-psgtr fd400000.phy: Invalid rate 0 for reference clock 0
At probe time, the associated GTR driver (e.g. SATA or PCIe) hasn't
initialized the clock yet, so clk_get_rate() likely returns 0 if the clock
is programmable. So this driver only works if the clock is fixed.
The PHY driver doesn't need to know the clock frequency at probe yet, so
wait until the associated driver initializes the lane before requesting the
clock rate setting.
In addition to allowing the driver to be used with programmable clocks,
this also reduces the driver's runtime memory footprint by removing an
array of pointers from struct xpsgtr_phy.
dt-bindings: phy: rockchip: Convert RK3399 Type-C PHY to schema
Convert the Rockchip RK3399 Type-C PHY to DT schema format. Add the
missing "power-domains" property and "port" and "orientation-switch"
properties in the child nodes.
Omit the previously deprecated properties as they aren't used anywhere.
Drop the 2nd example which was pretty much identical to the 1st example.
Chenyuan Yang [Mon, 14 Apr 2025 12:50:50 +0000 (07:50 -0500)]
phy: qcom-qmp-usb: Fix an NULL vs IS_ERR() bug
The qmp_usb_iomap() helper function currently returns the raw result of
devm_ioremap() for non-exclusive mappings. Since devm_ioremap() may return
a NULL pointer and the caller only checks error pointers with IS_ERR(),
NULL could bypass the check and lead to an invalid dereference.
Fix the issue by checking if devm_ioremap() returns NULL. When it does,
qmp_usb_iomap() now returns an error pointer via IOMEM_ERR_PTR(-ENOMEM),
ensuring safe and consistent error handling.
Signed-off-by: Chenyuan Yang <chenyuan0y@gmail.com> Fixes: a5d6b1ac56cb ("phy: qcom-qmp-usb: fix memleak on probe deferral") CC: Johan Hovold <johan@kernel.org> CC: Krzysztof Kozlowski <krzk@kernel.org> Reviewed-by: Johan Hovold <johan+linaro@kernel.org> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Link: https://lore.kernel.org/r/20250414125050.2118619-1-chenyuan0y@gmail.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
The Cadence Torrent SERDES supports USXGMII protocol. Hence, update the
bindings to allow PHY_TYPE_USXGMII. Since PHY_TYPE_USXGMII has the value
of "12" while the existing maximum allowed PHY TYPE is "9", switch back to
using "enum" property in the bindings to account for this discontinuity.
Daniel Golle [Tue, 22 Apr 2025 13:24:29 +0000 (15:24 +0200)]
phy: mediatek: xsphy: support type switch by pericfg
Patch from Sam Shih <sam.shih@mediatek.com> found in MediaTek SDK
released under GPL.
Get syscon and use it to set the PHY type.
Extend support to PCIe and SGMII mode in addition to USB2 and USB3.
Signed-off-by: Daniel Golle <daniel@makrotopia.org> Signed-off-by: Frank Wunderlich <frank-w@public-files.de> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Link: https://lore.kernel.org/r/20250422132438.15735-7-linux@fw-web.de Signed-off-by: Vinod Koul <vkoul@kernel.org>
Frank Wunderlich [Tue, 22 Apr 2025 13:24:27 +0000 (15:24 +0200)]
dt-bindings: phy: mtk-xs-phy: support type switch by pericfg
Add support for type switch by pericfg register between USB3/PCIe.
Signed-off-by: Frank Wunderlich <frank-w@public-files.de> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Link: https://lore.kernel.org/r/20250422132438.15735-5-linux@fw-web.de Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Frank Wunderlich <frank-w@public-files.de> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Link: https://lore.kernel.org/r/20250422132438.15735-4-linux@fw-web.de Signed-off-by: Vinod Koul <vkoul@kernel.org>
The main purpose of the ->rate member of struct rk_hdptx_phy is to
implement rk_hdptx_phy_clk_recalc_rate() by providing the actual rate
programmed in hardware. Hence the current naming is too generic and
rather ambiguous.
phy: rockchip: samsung-hdptx: Restrict altering TMDS char rate via CCF
Although, in theory, the clock provider functionality could be enabled
as a standalone driver feature, in practice it is unlikely that it would
be ever needed separately from the common PHY related features, i.e.
making use of the PHY PLL as an alternative and more accurate clock
source for display modes handling. Which means the PLL will be always
programmed according to the TMDS char rate set via the HDMI PHY
configuration API.
Currently it's possible to freely adjust the rate via the clock API as
well, that is through clk_set_rate(). Making the clock read-only is not
feasible since we need to ensure any rate update done via the PHY
configuration API has been actually programmed into the hardware before
CCF accesses it. This would be normally done during phy_ops.power_on()
or clk_ops.prepare() callbacks, but it might happen that the former gets
fired too late and the latter only once, hence we need to keep handle it
via clk_ops.set_rate() as a fallback approach.
Prevent changing the TMDS character rate via CCF by letting
rk_hdptx_phy_clk_round_rate() always return the value set via
phy_configure(). To avoid breaking existing users, i.e. RK DW HDMI QP
bridge driver, until the switch to the HDMI PHY config based approach is
completed, introduce a temporary exception to the rule, toggled via the
new ->restrict_rate_change flag, which indicates whether phy_configure()
has been called or not.
Additionally, revert any unlikely rate change that might have occurred
between the calls to ->round_rate() and ->set_rate().
phy: rockchip: samsung-hdptx: Setup TMDS char rate via phy_configure_opts_hdmi
The current workaround to setup the TMDS character rate relies on the
unconventional usage of phy_set_bus_width().
Make use of the recently introduced HDMI PHY configuration API to
properly handle the setup. The workaround will be dropped as soon as
the switch has been completed on both ends.
Rename rk_hdptx_phy_verify_config() to rk_hdptx_phy_verify_dp_config()
and introduce the rk_hdptx_phy_verify_hdmi_config() helper to check the
HDMI parameters during phy_configure().
phy: rockchip: samsung-hdptx: Avoid Hz<->hHz unit conversion overhead
The ropll_tmds_cfg table used to identify the configuration params for
the supported rates expects the search key, i.e. bit_rate member of
struct ropll_config, to be provided in hHz rather than Hz (1 hHz = 100
Hz). This requires multiple conversions between these units being
performed at runtime.
Improve implementation clarity and efficiency by consistently using the
Hz unit throughout driver's internal data structures and functions.
Also rename the rather misleading struct member.
phy: rockchip: samsung-hdptx: Drop superfluous cfgs driver data
The ->cfgs member has been introduced via commit f08d1c085638 ("phy:
phy-rockchip-samsung-hdptx: Don't use dt aliases to determine phy-id"),
but it is only used during probe() in order to setup ->phy_id.
Use a probe() local variable to store device match data and remove the
now unnecessary member from struct rk_hdptx_phy.
phy: rockchip: samsung-hdptx: Fix clock ratio setup
The switch from 1/10 to 1/40 clock ratio must happen when exceeding the
340 MHz rate limit of HDMI 1.4, i.e. when entering the HDMI 2.0 domain,
and not before.
Therefore, use the correct comparison operator '>' instead of '>=' when
checking the max rate. While at it, introduce a define for this rate
limit constant.
phy: qcom: qmp-pcie: Add PHY register retention support
Some QCOM PCIe PHYs support no_csr reset. Unlike BCR reset which resets the
whole PHY (hardware and register), no_csr reset only resets PHY hardware
but retains register values, which means PHY setting can be skipped during
PHY init if PCIe link is enabled in bootloader and only no_csr is toggled
after that.
Hence, determine whether the PHY has been enabled in bootloader by
verifying QPHY_START_CTRL register. If it's programmed and no_csr reset is
available, skip BCR reset and PHY register setting to establish the PCIe
link with bootloader - programmed PHY settings.
Decide the in-driver logic based on whether the nocsr reset is present
and defer checking the appropriateness of that to dt-bindings to save
on boilerplate.
Reset controller APIs are fine consuming a nullptr, so no additional
checks are necessary there.
Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Signed-off-by: Wenbin Yao <quic_wenbyao@quicinc.com> Reviewed-by: Abel Vesa <abel.vesa@linaro.org> Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de> Tested-by: Aleksandrs Vinarskis <alex.vinarskis@gmail.com> Link: https://lore.kernel.org/r/20250411113120.651363-2-quic_wenbyao@quicinc.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
phy: amlogic: phy-meson-axg-pcie: Fix PHY creation order in axg-pcie probe
Reorder the PHY creation in the axg-pcie probe function to ensure all
the resource is mapped before creating the PHY. This change addresses
the issue where the PHY creation was attempted before
mapping the necessary resources, potentially causing failures.
phy: exynos5-usbdrd: add exynos7870 USBDRD support
Implement support for Exynos7870 USB DRD on top of the existing
exynos5-usbdrd driver.
Exynos7870 has a single USB 2.0 DRD PHY controller and no 3.0 PHYs. Thus,
it only supports the UTMI interface.
Moreover, the PMU register offset for enabling the PHY controller is
different for SoCs such as Exynos7870, where BIT(0) is for the 3.0 PHY and
BIT(1) is for the 2.0 PHY. The phy_isol function for Exynos7870 uses the
appropriate register offsets.
Add the compatible string "samsung,exynos7870-usbdrd-phy" to the
documentation. The devicetree node requires two clocks, named "phy" and
"ref" (same as clocks required by Exynos5).
phy: exynos5-usbdrd: use GENMASK and FIELD_PREP for Exynos5 PHY registers
Most Exynos850 and Exynos9 (GS101) DRD PHY registers use GENMASK for masks
and FIELD_PREP for writing values to registers.
Rewrite the register definitions which don't follow this approach to follow
it as much as possible. This patch doesn't introduce any fixes or
functional changes, it's merely an attempt to introduce some uniformity and
consistency in the driver code.
The CRPORT SuperSpeed control registers have been exempted from this
change. Since the writing of register values do not require any masking
operations, implementing it would unnecessarily complicate things.
tools/include: make uapi/linux/types.h usable from assembly
The "real" linux/types.h UAPI header gracefully degrades to a NOOP when
included from assembly code.
Mirror this behaviour in the tools/ variant.
Test for __ASSEMBLER__ over __ASSEMBLY__ as the former is provided by the
toolchain automatically.
Reported-by: Mark Brown <broonie@kernel.org> Closes: https://lore.kernel.org/lkml/af553c62-ca2f-4956-932c-dd6e3a126f58@sirena.org.uk/ Fixes: c9fbaa879508 ("selftests: vDSO: parse_vdso: Use UAPI headers instead of libc headers") Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de> Link: https://patch.msgid.link/20250321-uapi-consistency-v1-1-439070118dc0@linutronix.de Signed-off-by: Mark Brown <broonie@kernel.org> Reviewed-by: Mark Brown <broonie@kernel.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge tag 'soundwire-6.15-rc1-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/soundwire
Pull soundwire fix from Vinod Koul:
- add missing config symbol CONFIG_SND_HDA_EXT_CORE required for asoc
driver CONFIG_SND_SOF_SOF_HDA_SDW_BPT
* tag 'soundwire-6.15-rc1-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/soundwire:
ASoC: SOF: Intel: Let SND_SOF_SOF_HDA_SDW_BPT select SND_HDA_EXT_CORE
Len Brown [Sun, 6 Apr 2025 18:49:20 +0000 (14:49 -0400)]
tools/power turbostat: v2025.05.06
Support up to 8192 processors
Add cpuidle governor debug telemetry, disabled by default
Update default output to exclude cpuidle invocation counts
Bug fixes
Merge tag 'perf-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull perf event fix from Ingo Molnar:
"Fix a perf events time accounting bug"
* tag 'perf-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/core: Fix child_total_time_enabled accounting bug at task exit
Merge tag 'sched-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull scheduler fixes from Ingo Molnar:
- Fix a nonsensical Kconfig combination
- Remove an unnecessary rseq-notification
* tag 'sched-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
rseq: Eliminate useless task_work on execve
sched/isolation: Make CONFIG_CPU_ISOLATION depend on CONFIG_SMP
... and don't error out so hard on missing module descriptions.
Before commit 6c6c1fc09de3 ("modpost: require a MODULE_DESCRIPTION()")
we used to warn about missing module descriptions, but only when
building with extra warnigns (ie 'W=1').
After that commit the warning became an unconditional hard error.
And it turns out not all modules have been converted despite the claims
to the contrary. As reported by Damian Tometzki, the slub KUnit test
didn't have a module description, and apparently nobody ever really
noticed.
The reason nobody noticed seems to be that the slub KUnit tests get
disabled by SLUB_TINY, which also ends up disabling a lot of other code,
both in tests and in slub itself. And so anybody doing full build tests
didn't actually see this failre.
So let's disable SLUB_TINY for build-only tests, since it clearly ends
up limiting build coverage. Also turn the missing module descriptions
error back into a warning, but let's keep it around for non-'W=1'
builds.
Len Brown [Sun, 6 Apr 2025 15:18:39 +0000 (11:18 -0400)]
tools/power turbostat: report CoreThr per measurement interval
The CoreThr column displays total thermal throttling events
since boot time.
Change it to report events during the measurement interval.
This is more useful for showing a user the current conditions.
Total events since boot time are still available to the user via
/sys/devices/system/cpu/cpu*/thermal_throttle/*
Document CoreThr on turbostat.8
Fixes: eae97e053fe30 ("turbostat: Support thermal throttle count print") Reported-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Len Brown <len.brown@intel.com> Cc: Chen Yu <yu.c.chen@intel.com>
Justin Ernst [Wed, 19 Mar 2025 20:27:31 +0000 (15:27 -0500)]
tools/power turbostat: Increase CPU_SUBSET_MAXCPUS to 8192
On systems with >= 1024 cpus (in my case 1152), turbostat fails with the error output:
"turbostat: /sys/fs/cgroup/cpuset.cpus.effective: cpu str malformat 0-1151"
A similar error appears with the use of turbostat --cpu when the inputted cpu
range contains a cpu number >= 1024:
# turbostat -c 1100-1151
"--cpu 1100-1151" malformed
...
Both errors are caused by parse_cpu_str() reaching its limit of CPU_SUBSET_MAXCPUS.
It's a good idea to limit the maximum cpu number being parsed, but 1024 is too low.
For a small increase in compute and allocated memory, increasing CPU_SUBSET_MAXCPUS
brings support for parsing cpu numbers >= 1024.
Increase CPU_SUBSET_MAXCPUS to 8192, a common setting for CONFIG_NR_CPUS on x86_64.
Signed-off-by: Justin Ernst <justin.ernst@hpe.com> Signed-off-by: Len Brown <len.brown@intel.com>
Merge tag 'timers-cleanups-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer cleanups from Thomas Gleixner:
"A set of final cleanups for the timer subsystem:
- Convert all del_timer[_sync]() instances over to the new
timer_delete[_sync]() API and remove the legacy wrappers.
Conversion was done with coccinelle plus some manual fixups as
coccinelle chokes on scoped_guard().
- The final cleanup of the hrtimer_init() to hrtimer_setup()
conversion.
This has been delayed to the end of the merge window, so that all
patches which have been merged through other trees are in mainline
and all new users are catched.
Doing this right before rc1 ensures that new code which is merged post
rc1 is not introducing new instances of the original functionality"
* tag 'timers-cleanups-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
tracing/timers: Rename the hrtimer_init event to hrtimer_setup
hrtimers: Rename debug_init_on_stack() to debug_setup_on_stack()
hrtimers: Rename debug_init() to debug_setup()
hrtimers: Rename __hrtimer_init_sleeper() to __hrtimer_setup_sleeper()
hrtimers: Remove unnecessary NULL check in hrtimer_start_range_ns()
hrtimers: Make callback function pointer private
hrtimers: Merge __hrtimer_init() into __hrtimer_setup()
hrtimers: Switch to use __htimer_setup()
hrtimers: Delete hrtimer_init()
treewide: Convert new and leftover hrtimer_init() users
treewide: Switch/rename to timer_delete[_sync]()
Merge tag 'irq-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull more irq updates from Thomas Gleixner:
"A set of updates for the interrupt subsystem:
- A treewide cleanup for the irq_domain code, which makes the naming
consistent and gets rid of the original oddity of naming domains
'host'.
This is a trivial mechanical change and is done late to ensure that
all instances have been catched and new code merged post rc1 wont
reintroduce new instances.
- A trivial consistency fix in the migration code
The recent introduction of irq_force_complete_move() in the core
code, causes a problem for the nostalgia crowd who maintains ia64
out of tree.
The code assumes that hierarchical interrupt domains are enabled
and dereferences irq_data::parent_data unconditionally. That works
in mainline because both architectures which enable that code have
hierarchical domains enabled. Though it breaks the ia64 build,
which enables the functionality, but does not have hierarchical
domains.
While it's not really a problem for mainline today, this
unconditional dereference is inconsistent and trivially fixable by
using the existing helper function irqd_get_parent_data(), which
has the appropriate #ifdeffery in place"
* tag 'irq-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
genirq/migration: Use irqd_get_parent_data() in irq_force_complete_move()
irqdomain: Stop using 'host' for domain
irqdomain: Rename irq_get_default_host() to irq_get_default_domain()
irqdomain: Rename irq_set_default_host() to irq_set_default_domain()
Merge tag 'timers-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer fix from Thomas Gleixner:
"A revert to fix a adjtimex() regression:
The recent change to prevent that time goes backwards for the coarse
time getters due to immediate multiplier adjustments via adjtimex(),
changed the way how the timekeeping core treats that.
That change result in a regression on the adjtimex() side, which is
user space visible:
1) The forwarding of the base time moves the update out of the
original period and establishes a new one. That's changing the
behaviour of the [PF]LL control, which user space expects to be
applied periodically.
2) The clearing of the accumulated NTP error due to #1, changes the
behaviour as well.
An attempt to delay the multiplier/frequency update to the next tick
did not solve the problem as userspace expects that the multiplier or
frequency updates are in effect, when the syscall returns.
There is a different solution for the coarse time problem available,
so revert the offending commit to restore the existing adjtimex()
behaviour"
* tag 'timers-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
Revert "timekeeping: Fix possible inconsistencies in _COARSE clockids"
Merge tag 'sh-for-v6.15-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/glaubitz/sh-linux
Pull sh updates from John Paul Adrian Glaubitz:
"One important fix and one small configuration update.
The first patch by Artur Rojek fixes an issue with the J2 firmware
loader not being able to find the location of the device tree blob due
to insufficient alignment of the .bss section which rendered J2 boards
unbootable.
The second patch by Johan Korsnes updates the defconfigs on sh to drop
the CONFIG_NET_CLS_TCINDEX configuration option which became obsolete
after 8c710f75256b ("net/sched: Retire tcindex classifier").
Summary:
- sh: defconfig: Drop obsolete CONFIG_NET_CLS_TCINDEX
- sh: Align .bss section padding to 8-byte boundary"
* tag 'sh-for-v6.15-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/glaubitz/sh-linux:
sh: defconfig: Drop obsolete CONFIG_NET_CLS_TCINDEX
sh: Align .bss section padding to 8-byte boundary
Merge tag 'kbuild-v6.15' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild
Pull Kbuild updates from Masahiro Yamada:
- Improve performance in gendwarfksyms
- Remove deprecated EXTRA_*FLAGS and KBUILD_ENABLE_EXTRA_GCC_CHECKS
- Support CONFIG_HEADERS_INSTALL for ARCH=um
- Use more relative paths to sources files for better reproducibility
- Support the loong64 Debian architecture
- Add Kbuild bash completion
- Introduce intermediate vmlinux.unstripped for architectures that need
static relocations to be stripped from the final vmlinux
- Fix versioning in Debian packages for -rc releases
- Treat missing MODULE_DESCRIPTION() as an error
- Convert Nios2 Makefiles to use the generic rule for built-in DTB
- Add debuginfo support to the RPM package
* tag 'kbuild-v6.15' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild: (40 commits)
kbuild: rpm-pkg: build a debuginfo RPM
kconfig: merge_config: use an empty file as initfile
nios2: migrate to the generic rule for built-in DTB
rust: kbuild: skip `--remap-path-prefix` for `rustdoc`
kbuild: pacman-pkg: hardcode module installation path
kbuild: deb-pkg: don't set KBUILD_BUILD_VERSION unconditionally
modpost: require a MODULE_DESCRIPTION()
kbuild: make all file references relative to source root
x86: drop unnecessary prefix map configuration
kbuild: deb-pkg: add comment about future removal of KDEB_COMPRESS
kbuild: Add a help message for "headers"
kbuild: deb-pkg: remove "version" variable in mkdebian
kbuild: deb-pkg: fix versioning for -rc releases
Documentation/kbuild: Fix indentation in modules.rst example
x86: Get rid of Makefile.postlink
kbuild: Create intermediate vmlinux build with relocations preserved
kbuild: Introduce Kconfig symbol for linking vmlinux with relocations
kbuild: link-vmlinux.sh: Make output file name configurable
kbuild: do not generate .tmp_vmlinux*.map when CONFIG_VMLINUX_MAP=y
Revert "kheaders: Ignore silly-rename files"
...
Merge tag 'drm-next-2025-04-05' of https://gitlab.freedesktop.org/drm/kernel
Pull drm fixes from Dave Airlie:
"Weekly fixes, mostly from the end of last week, this week was very
quiet, maybe you scared everyone away. It's mostly amdgpu, and xe,
with some i915, adp and bridge bits, since I think this is overly
quiet I'd expect rc2 to be a bit more lively.
bridge:
- tda998x: Select CONFIG_DRM_KMS_HELPER
amdgpu:
- Guard against potential division by 0 in fan code
- Zero RPM support for SMU 14.0.2
- Properly handle SI and CIK support being disabled
- PSR fixes
- DML2 fixes
- DP Link training fix
- Vblank fixes
- RAS fixes
- Partitioning fix
- SDMA fix
- SMU 13.0.x fixes
- Rom fetching fix
- MES fixes
- Queue reset fix
xe:
- Fix NULL pointer dereference on error path
- Add missing HW workaround for BMG
- Fix survivability mode not triggering
- Fix build warning when DRM_FBDEV_EMULATION is not set
i915:
- Bounds check for scalers in DSC prefill latency computation
- Fix build by adding a missing include
adp:
- Fix error handling in plane setup"
# -----BEGIN PGP SIGNATURE-----
* tag 'drm-next-2025-04-05' of https://gitlab.freedesktop.org/drm/kernel: (34 commits)
drm/i2c: tda998x: select CONFIG_DRM_KMS_HELPER
drm/amdgpu/gfx12: fix num_mec
drm/amdgpu/gfx11: fix num_mec
drm/amd/pm: Add gpu_metrics_v1_8
drm/amdgpu: Prefer shadow rom when available
drm/amd/pm: Update smu metrics table for smu_v13_0_6
drm/amd/pm: Remove host limit metrics support
Remove unnecessary firmware version check for gc v9_4_2
drm/amdgpu: stop unmapping MQD for kernel queues v3
Revert "drm/amdgpu/sdma_v4_4_2: update VM flush implementation for SDMA"
drm/amdgpu: Parse all deferred errors with UMC aca handle
drm/amdgpu: Update ta ras block
drm/amdgpu: Add NPS2 to DPX compatible mode
drm/amdgpu: Use correct gfx deferred error count
drm/amd/display: Actually do immediate vblank disable
drm/amd/display: prevent hang on link training fail
Revert "drm/amd/display: dml2 soc dscclk use DPM table clk setting"
drm/amd/display: Increase vblank offdelay for PSR panels
drm/amd: Handle being compiled without SI or CIK support better
drm/amd/pm: Add zero RPM enabled OD setting support for SMU14.0.2
...
Uday Shankar [Mon, 31 Mar 2025 22:46:32 +0000 (16:46 -0600)]
kbuild: rpm-pkg: build a debuginfo RPM
The rpm-pkg make target currently suffers from a few issues related to
debuginfo:
1. debuginfo for things built into the kernel (vmlinux) is not available
in any RPM produced by make rpm-pkg. This makes using tools like
systemtap against a make rpm-pkg kernel impossible.
2. debug source for the kernel is not available. This means that
commands like 'disas /s' in gdb, which display source intermixed with
assembly, can only print file names/line numbers which then must be
painstakingly resolved to actual source in a separate editor.
3. debuginfo for modules is available, but it remains bundled with the
.ko files that contain module code, in the main kernel RPM. This is a
waste of space for users who do not need to debug the kernel (i.e.
most users).
Address all of these issues by additionally building a debuginfo RPM
when the kernel configuration allows for it, in line with standard
patterns followed by RPM distributors. With these changes:
1. systemtap now works (when these changes are backported to 6.11, since
systemtap lags a bit behind in compatibility), as verified by the
following simple test script:
129 op_str = blk_op_name[op];
130
131 return op_str;
132 }
0xffffffff814c8768 <+40>: jmp 0xffffffff81d01360 <__x86_return_thunk>
End of assembler dump.
3. The size of the main kernel package goes down substantially,
especially if many modules are built (quite typical). Here is a
comparison of installed size of the kernel package (configured with
allmodconfig, dwarf4 debuginfo, and module compression turned off)
before and after this patch:
# rpm -qi kernel-6.13* | grep -E '^(Version|Size)'
Version : 6.13.0postpatch+
Size : 1382874089
Version : 6.13.0prepatch+
Size : 17870795887
This is a ~92% size reduction.
Note that a debuginfo package can only be produced if the following
configs are set:
- CONFIG_DEBUG_INFO=y
- CONFIG_MODULE_COMPRESS=n
- CONFIG_DEBUG_INFO_SPLIT=n
The first of these is obvious - we can't produce debuginfo if the build
does not generate it. The second two requirements can in principle be
removed, but doing so is difficult with the current approach, which uses
a generic rpmbuild script find-debuginfo.sh that processes all packaged
executables. If we want to remove those requirements the best path
forward is likely to add some debuginfo extraction/installation logic to
the modules_install target (controllable by flags). That way, it's
easier to operate on modules before they're compressed, and the logic
can be reused by all packaging targets.
Daniel Gomez [Fri, 28 Mar 2025 14:28:37 +0000 (14:28 +0000)]
kconfig: merge_config: use an empty file as initfile
The scripts/kconfig/merge_config.sh script requires an existing
$INITFILE (or the $1 argument) as a base file for merging Kconfig
fragments. However, an empty $INITFILE can serve as an initial starting
point, later referenced by the KCONFIG_ALLCONFIG Makefile variable
if -m is not used. This variable can point to any configuration file
containing preset config symbols (the merged output) as stated in
Documentation/kbuild/kconfig.rst. When -m is used $INITFILE will
contain just the merge output requiring the user to run make (i.e.
KCONFIG_ALLCONFIG=<$INITFILE> make <allnoconfig/alldefconfig> or make
olddefconfig).
Instead of failing when `$INITFILE` is missing, create an empty file and
use it as the starting point for merges.
Signed-off-by: Daniel Gomez <da.gomez@samsung.com> Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Masahiro Yamada [Sun, 22 Dec 2024 00:30:53 +0000 (09:30 +0900)]
nios2: migrate to the generic rule for built-in DTB
Commit 654102df2ac2 ("kbuild: add generic support for built-in boot
DTBs") introduced generic support for built-in DTBs.
Select GENERIC_BUILTIN_DTB when built-in DTB support is enabled.
To keep consistency across architectures, this commit also renames
CONFIG_NIOS2_DTB_SOURCE_BOOL to CONFIG_BUILTIN_DTB, and
CONFIG_NIOS2_DTB_SOURCE to CONFIG_BUILTIN_DTB_NAME.
Johan Korsnes [Sun, 23 Mar 2025 19:13:30 +0000 (20:13 +0100)]
sh: defconfig: Drop obsolete CONFIG_NET_CLS_TCINDEX
This option was removed from Kconfig in 8c710f75256b ("net/sched:
Retire tcindex classifier") but from the defconfigs.
Fixes: 8c710f75256b ("net/sched: Retire tcindex classifier") Signed-off-by: Johan Korsnes <johan.korsnes@gmail.com> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Cc: Rich Felker <dalias@libc.org> Cc: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Artur Rojek [Sun, 16 Feb 2025 17:55:44 +0000 (18:55 +0100)]
sh: Align .bss section padding to 8-byte boundary
J2-based devices expect to find a device tree blob at the end of the
.bss section. As of a77725a9a3c5 ("scripts/dtc: Update to upstream
version v1.6.1-19-g0a3a9d3449c8"), libfdt enforces 8-byte alignment
for the DTB, causing J2 devices to fail early in sh_fdt_init().
As the J2 loader firmware calculates the DTB location based on the kernel
image .bss section size rather than the __bss_stop symbol offset, the
required alignment can't be enforced with BSS_SECTION(0, PAGE_SIZE, 8).
To fix this, inline a modified version of the above macro which grows
.bss by the required size. While this change affects all existing SH
boards, it should be benign on platforms which don't need this alignment.
Signed-off-by: Artur Rojek <contact@artur-rojek.eu> Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> Tested-by: Rob Landley <rob@landley.net> Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
projects / thirdparty / linux.git / log
