aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2021-08-04arm64: dts: qcom: pm660l: Support SPMI regulators on PMIC sid 3AngeloGioacchino Del Regno1-0/+4
The PM660L PMIC has SPMI regulators on the PMIC SID 3: let's add the compatible in order to probe them. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: pm660l: Add WLED supportKonrad Dybcio1-0/+17
This will enable backlight control on WLED-enabled devices. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: pm660: Support SPMI regulators on PMIC sid 1AngeloGioacchino Del Regno1-0/+11
The PM660 PMIC has SPMI regulators on the PMIC SID 1: let's declare the second pmic subtree and add the spmi vregs compatible to probe them there. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add Adreno 508 GPU configurationAngeloGioacchino Del Regno1-1/+83
The SDM630 SoC features an Adreno 508.0 GPU with a minimum frequency of 160MHz and a maximum of (depending on the speed-bin) 775MHz. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Raise tcsr_mutex_regs sizeKonrad Dybcio1-1/+1
Enlarge the size to make sure all relevant registers can be reached. This will be required to support the modem. Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add ADSP remoteproc configurationKonrad Dybcio1-0/+83
Configure the ADSP remote processor and add a simple sound{} node to make way for future development. Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add thermal-zones configurationKonrad Dybcio1-0/+173
Add a basic thermal-zones configuration to make sure the SoC doesn't overheat itself to death. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] [bjorn: Sorted thermal-zones below "soc"] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add modem/ADSP SMP2P nodesKonrad Dybcio1-0/+40
Add SMP2P nodes that are required for ADSP and modem bringup. Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add TSENS nodeKonrad Dybcio2-0/+12
This will enable temperature reporting for various SoC components. Acked-by: Rob Herring <[email protected]> Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add qcom,adreno-smmu compatibleAngeloGioacchino Del Regno1-1/+2
The Adreno SMMU in SDM630 needs this compatible string for proper context handling and split pagetables support. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add clocks and power domains to SMMU nodesAngeloGioacchino Del Regno1-3/+28
Add the required clocks and power domains for the SMMUs to work. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add GPU Clock Controller nodeAngeloGioacchino Del Regno1-0/+17
Add the GPU Clock Controller in SDM630 and keep it disabled by default. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add interconnect and opp table to sdhc_1AngeloGioacchino Del Regno1-0/+28
The SDHC port 1 has interconnects and can make use of DVFS: define the interconnections and the OPP table in order to optimize performance and power consumption. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] [bjorn: Dropped "sdhc1-" prefix from opp-table node] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add SDHCI2 nodeAngeloGioacchino Del Regno1-0/+50
This will enable usage of (generally) uSD cards. While at it, add accompanying OPP table for DVFS. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] [bjorn: Dropped "sdhci1-" prefix from opp-table node] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Fix TLMM node and pinctrl configurationAngeloGioacchino Del Regno1-73/+139
Previous pinctrl configuration was wrong. Fix it and clean up how multi-pin states are described. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] [bjorn: Polished the commit message] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add USB configurationKonrad Dybcio1-0/+64
This will let us use USB2 on our devices. The SoC supposedly supports USB3, but there are no known cases of devices that actually have USB3 wired up in hardware. Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] [bjorn: Changes dwc3 node name to "usb"] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add qfprom subnodesAngeloGioacchino Del Regno1-0/+10
These will be required for USB and Adreno support. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] [bjorn: y/_/-/ in gpu_speed_bin] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add MDSS nodesKonrad Dybcio1-2/+196
Add MDSS node along with its children to enable display functionality. Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add interconnect provider nodesKonrad Dybcio1-0/+59
Add interconnect provider nodes to allow for NoC bus scaling. Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add MMCC nodeKonrad Dybcio1-0/+28
Add MultiMedia Clock Controller node to allow for accessing and controlling Multimedia Subsystem clocks by their respective users. Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Add RPMPD nodesKonrad Dybcio1-0/+47
Add the rpmpd node on the sdm630 and define the available levels. Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sdm630: Rewrite memory mapAngeloGioacchino Del Regno1-23/+18
The memory map was wrong. Fix it. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sm8150: Fix incorrect cpu opp table entryThara Gopinath1-1/+1
CPU0 frequency 768MHz is wrongly modeled as 576000000 hz in cpu0_opp_table. Use the correct value 768000000 hz. Fixes: 2b6187abafea ("arm64: dts: qcom: sm8150: Add CPU opp tables") Signed-off-by: Thara Gopinath <[email protected]> Reviewed-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-04arm64: dts: qcom: sm8150-mtp: Add 8150 compatible stringThara Gopinath1-1/+1
Add "qcom,sm8150" as one of the platform compatible strings. This will be later used by cpufreq-dt-platdev to exclude using cpufreq-dt cpufreq driver. Signed-off-by: Thara Gopinath <[email protected]> Reviewed-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-03arm64: dts: qcom: Add support for SONY Xperia X Performance / XZ / XZs ↵AngeloGioacchino Del Regno9-6/+1069
(msm8996, Tone platform) Add support for following boards: - Xperia X Performance (dora) - Xperia XZ (kagura) - Xperia XZs (keyaki) They are all based on the SONY Tone platform and feature largely similar hardware with the most obvious differences being lack of USB-C and ToF sensor on Dora and different camera sensor on Keyaki. Signed-off-by: AngeloGioacchino Del Regno <[email protected]> Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-03arm64: dts: qcom: msm8996-*: Disable HDMI by defaultKonrad Dybcio3-0/+20
Most phones ship without HDMI and leaving it enabled wrecks havoc. Disable it in msm8996.dtsi and re-enable it on the boards that did not disable it previously. Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-03arm64: dts: qcom: Add MSM8996v3.0 DTSI fileKonrad Dybcio1-0/+63
Add an overlay for MSM8996v3.0, which is a pre-final revision of the said SoC. It has some stark differences with regards to GPU, or more specifically its power delivery path. Oh, and of course a different msm-id. Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-08-03arm64: dts: qcom: Add PMI8996 DTSI fileKonrad Dybcio1-0/+15
PMI8996 is *almost* the same hardware as PMI8994, say for some annoyances: - Boards equipped with PMI8996 now have to include pmic-id (which wasn't the case before) - Different qpnp-ibb-discharge-resistor value (will be addressed after LABIBB is introduced) - Different inhibit-derating-ua value (will be addressed after BCL is introduced) - Different ramp_up_step value (will be addressed after [if?] QPNP Flash LED is introduced) This DTSI is supposed to be included >>ON TOP OF<< pmi8994.dtsi, like this: -- msm8996-nice-device.dts -- \#include "pmi8994.dtsi" \#include "pmi8996.dtsi" or more likely like this: -- msm8996-some-phone.dts -- \#include "msm8996.dtsi" ... \#include "pmi8994.dtsi" -- msm8996-pmi8996-some-phone.dts -- \#include "msm8996-some-phone.dts" \#include "pmi8996.dtsi" So that we only have to keep 2 DTs for devices that were shipped with both ones, instead of what would be three (device base + pmi8994 + pmi8996) Signed-off-by: Konrad Dybcio <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-21arm64: dts: qcom: sc7180: trogdor: Update audio codec to Max98360AJudy Hsiao1-3/+3
Use max98360a dts node to correctly describe the hardware. Signed-off-by: Judy Hsiao <[email protected]> Reviewed-by: Matthias Kaehlcke <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-20arm64: dts: qcom: sc7280: Add USB nodes for IDP boardSandeep Maheswaram1-0/+39
Add USB nodes for sc7280 IDP board. Signed-off-by: Sandeep Maheswaram <[email protected]> Reviewed-by: Matthias Kaehlcke <[email protected]> Reviewed-by: Stephen Boyd <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-20arm64: dts: qcom: sc7280: Add USB related nodesSandeep Maheswaram1-0/+164
Add nodes for DWC3 USB controller, QMP and HS USB PHYs in sc7280 SOC. Signed-off-by: Sandeep Maheswaram <[email protected]> Reviewed-by: Matthias Kaehlcke <[email protected]> Reviewed-by: Stephen Boyd <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-20arm64: dts: qcom: sc7280: Add nodes for eMMC and SD cardShaik Sajida Bhanu2-0/+257
Add nodes for eMMC and SD card on sc7280. Signed-off-by: Shaik Sajida Bhanu <[email protected]> Reviewed-by: Stephen Boyd <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-20arm64: dts: qcom: Fix usb entries for SA8155p adp boardBhupesh Sharma1-9/+51
SA8155p adp board has two USB A-type receptacles called USB-portB and USB-portC respectively. While USB-portB is a USB High-Speed connector/interface, the USB-portC one is a USB 3.1 Super-Speed connector/interface. Also the USB-portB is used as the USB emergency download port (for image download purposes). Enable both the ports on the board in USB Host mode (since all the USB interfaces are brought out to USB Type A connectors). Cc: Bjorn Andersson <[email protected]> Signed-off-by: Bhupesh Sharma <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-20arm64: dts: qcom: sm8150: Sort dc_noc and gem_noc nodesBhupesh Sharma1-14/+14
Nodes should be sorted by address, so move the dc_noc and gem_noc nodes to their correct place. Cc: Bjorn Andersson <[email protected]> Signed-off-by: Bhupesh Sharma <[email protected]> Link: https://lore.kernel.org/r/[email protected] [bjorn: Adjusted order slightly more] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-20arm64: dts: qcom: Use correct naming for dwc3 usb nodes in dts filesBhupesh Sharma5-7/+7
The dwc3 usb nodes in several arm64 qcom dts are currently named differently, somewhere as 'usb@<addr>' and somewhere as 'dwc3@<addr>', leading to some confusion when one sees the entries in sysfs or dmesg: [ 1.943482] dwc3 a600000.usb: Adding to iommu group 1 [ 2.266127] dwc3 a800000.dwc3: Adding to iommu group 2 Name the usb nodes as 'usb@<addr>' for consistency, which is the correct convention as per the 'snps,dwc3' dt-binding as well (see [1]). [1]. Documentation/devicetree/bindings/usb/snps,dwc3.yaml Cc: Bjorn Andersson <[email protected]> Signed-off-by: Bhupesh Sharma <[email protected]> Link: https://lore.kernel.org/r/[email protected] [bjorn: Extended to also fix ipq6018] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-20arm64: dts: qcom: sm8150: Add UFS ICE capabilityBhupesh Sharma1-4/+9
Add support for UFS ICE (Qualcomm Inline Crypto Engine) in sm8150 SoC dts. I tested this on SA8155p-adp board, which is a publicly available development board that uses the sa8155p Qualcomm Snapdragon SoC. SA8155p platform is similar to the SM8150, so use this as base for now. I tested the UFS ICE feature using 'fscrypt' test utility. Cc: Bjorn Andersson <[email protected]> Cc: Eric Biggers <[email protected]> Signed-off-by: Bhupesh Sharma <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-20arm64: dts: qcom: sm8250: fix usb2 qmp phy nodeDmitry Baryshkov1-1/+1
Use 'lanes' as SuperSpeed lanes device node instead of just 'lane' to fix issues with TypeC support. Fixes: 46a6f297d7dd ("arm64: dts: qcom: sm8250: Add USB and PHY device nodes") Cc: [email protected] Cc: [email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Tested-by: Bryan O'Donoghue <[email protected]> Signed-off-by: Bryan O'Donoghue <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-19arm64: dts: qcom: sc7180: Set adau wakeup delay to 80 msJudy Hsiao1-1/+1
Set audu wakeup delay to 80 ms for fixing pop noise during capture begin. Fixes: ba5f9b5d7ff3 ("arm64: dts: qcom: sc7180: Add wakeup delay for adau codec") Signed-off-by: Judy Hsiao <[email protected]> Reviewed-by: Stephen Boyd <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-19arm64: dts: qcom: msm8916-wingtech-wt88047: Add IMUStephan Gerhold1-0/+30
The Xiaomi Redmi 2 has an InvenSense MPU-6880 supported by the inv_mpu6050 driver in Linux. Add it to the device tree. Signed-off-by: Stephan Gerhold <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-19arm64: dts: qcom: msm8916-wingtech-wt88047: Add notification LEDStephan Gerhold1-0/+35
The Xiaomi Redmi 2 has a notification LED connected to an Awinic AW2013 controller, add it to the device tree. Signed-off-by: Stephan Gerhold <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-19arm64: dts: qcom: msm8916-wingtech-wt88047: Add touchscreenStephan Gerhold1-0/+40
The Xiaomi Redmi 2 has a FocalTech touchscreen. The exact model is not mentioned anywhere, but it works just fine with the mainline edt-ft5x06 driver when using the "edt,edt-ft5506" compatible. Signed-off-by: Stephan Gerhold <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-19arm64: dts: qcom: Add device tree for Xiaomi Redmi 2 (wingtech-wt88047)Stephan Gerhold2-0/+209
The Xiaomi Redmi 2 is a MSM8916 smartphone that was made by Wingtech (codename: wt88047). It's fairly similar to the other MSM8916 devices. The device tree contains initial support for the Xiaomi Redmi 2 with: - UART (untested, probably available via some test points) - eMMC/SD card - Buttons - Vibrator - WiFi/Bluetooth (WCNSS) - USB Note that the Xiaomi Redmi 2 is available in variants with different names (e.g. Redmi 2 Prime, Redmi 2 Pro, ...). As far as I know the main difference between those is the amount of RAM and supported LTE bands. This difference is automatically handled by bootloader/modem firmware so there is no need for separate device trees. Signed-off-by: Stephan Gerhold <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-19arm64: dts: qcom: sm8150: Add CPU opp tablesThara Gopinath1-0/+315
Add OPP tables to scale DDR and L3 with CPUs for SM8150 SoCs. This gives a significant performance boost for cpu-ddr loads. Below is the results for mbw benchmark with and without the opp tables. /mbw 1500 Without l3/ddr scaling With l3 ddr/scaling MEMCPY(MiB/s) 3574 10448 DUMB(MiB/s) 3500 11721 MCBLOCK(MiB/s) 8976 23595 Signed-off-by: Thara Gopinath <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-19arm64: dts: qcom: ipq8074: Add scm on IPQ8074 SoCsGokul Sriram Palanisamy1-0/+6
Enables scm support, clock is not needed for enabling scm interface. Signed-off-by: Gokul Sriram Palanisamy <[email protected]> Signed-off-by: Sricharan R <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-19arm64: dts: qcom: ipq6018: Update WCSS PIL driver compatibleGokul Sriram Palanisamy1-1/+1
Updated WCSS PIL driver node with IPQ6018 specific compatible to enable SoC specific driver data. Signed-off-by: Gokul Sriram Palanisamy <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-19arm64: dts: qcom: ipq6018: enable USB2 supportKathiravan T2-0/+56
Signed-off-by: Kathiravan T <[email protected]> [baruch: adjust regs address/size; drop binding updates; drop unsupported quirk properties] Signed-off-by: Baruch Siach <[email protected]> Link: https://lore.kernel.org/r/59a0d43f34b69406cd320f16edc4e7fabe022bfd.1611756920.git.baruch@tkos.co.il Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-19arm64: dts: qcom: sc7180: Add DisplayPort nodeKuogee Hsieh2-0/+85
Add DP device node on sc7180. Signed-off-by: Kuogee Hsieh <[email protected]> Reviewed-by: Bjorn Andersson <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2021-07-11Linux 5.14-rc1Linus Torvalds1-2/+2
2021-07-11mm/rmap: try_to_migrate() skip zone_device !device_privateHugh Dickins1-3/+3
I know nothing about zone_device pages and !device_private pages; but if try_to_migrate_one() will do nothing for them, then it's better that try_to_migrate() filter them first, than trawl through all their vmas. Signed-off-by: Hugh Dickins <[email protected]> Reviewed-by: Shakeel Butt <[email protected]> Reviewed-by: Alistair Popple <[email protected]> Link: https://lore.kernel.org/lkml/[email protected]/ Cc: Andrew Morton <[email protected]> Cc: Jason Gunthorpe <[email protected]> Cc: Ralph Campbell <[email protected]> Cc: Christoph Hellwig <[email protected]> Cc: Yang Shi <[email protected]> Cc: Kirill A. Shutemov <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2021-07-11mm/rmap: fix new bug: premature return from page_mlock_one()Hugh Dickins1-6/+5
In the unlikely race case that page_mlock_one() finds VM_LOCKED has been cleared by the time it got page table lock, page_vma_mapped_walk_done() must be called before returning, either explicitly, or by a final call to page_vma_mapped_walk() - otherwise the page table remains locked. Fixes: cd62734ca60d ("mm/rmap: split try_to_munlock from try_to_unmap") Signed-off-by: Hugh Dickins <[email protected]> Reviewed-by: Alistair Popple <[email protected]> Reviewed-by: Shakeel Butt <[email protected]> Reported-by: kernel test robot <[email protected]> Link: https://lore.kernel.org/lkml/20210711151446.GB4070@xsang-OptiPlex-9020/ Link: https://lore.kernel.org/lkml/[email protected]/ Cc: Andrew Morton <[email protected]> Cc: Jason Gunthorpe <[email protected]> Cc: Ralph Campbell <[email protected]> Cc: Christoph Hellwig <[email protected]> Cc: Yang Shi <[email protected]> Cc: Kirill A. Shutemov <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.