Olof Johansson [Mon, 6 Jan 2020 17:28:56 +0000 (09:28 -0800)]
Merge tag 'renesas-arm64-dt-for-v5.6-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-devel into arm/dt
Renesas ARM64 DT updates for v5.6
- Remove now unused ARCH_R8A7796 config symbol,
- Rename R-Car H3 and M3-W SoC, and ULCB board DTS files to increase
naming consistency,
- Miscellaneous fixes for issues detected by "make dtbs_check",
- Enhance support for R-Car M3-W+,
- Display support for the EK874 board,
- Prepare for split of R-Car H3 ES1.x and ES2.0+ config symbols,
- Minor fixes and improvements.
* tag 'renesas-arm64-dt-for-v5.6-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-devel:
arm64: dts: renesas: Prepare for split of ARCH_R8A7795 into ARCH_R8A7795[01]
arm64: dts: renesas: Sort DTBs in Makefile
arm64: dts: renesas: Drop redundant SoC prefixes from ULCB DTS file names
arm64: dts: renesas: Rename r8a7795{-es1,}* to r8a7795[01]*
arm64: dts: renesas: Add EK874 board with idk-2121wr display support
arm64: dts: renesas: r8a77961: Add SDHI nodes
arm64: dts: renesas: r8a77961: Add I2C nodes
arm64: dts: renesas: r8a77961: Add SYS-DMAC nodes
arm64: dts: renesas: r8a77961: Add RAVB node
arm64: dts: renesas: r8a77961: Add GPIO nodes
arm64: dts: renesas: r8a77961: Add RWDT node
arm64: dts: renesas: r8a77990: ebisu: Remove clkout-lr-synchronous from sound
arm64: dts: renesas: r8a77970: Group tuples in thermal reg property
arm64: dts: renesas: Group tuples in pci ranges and dma-ranges properties
arm64: dts: renesas: Group tuples in interrupt properties
arm64: dts: renesas: Group tuples in regulator-gpio states properties
arm64: dts: renesas: Rename r8a7796* to r8a77960*
arm64: dts: renesas: Remove use of ARCH_R8A7796
Olof Johansson [Mon, 6 Jan 2020 17:28:00 +0000 (09:28 -0800)]
Merge tag 'renesas-arm-dt-for-v5.6-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-devel into arm/dt
Renesas ARM DT updates for v5.6
- Touch screen support for the iwg20d board,
- ARM global timer support on Cortex-A9 MPCore SoCs,
- Miscellaneous fixes for issues detected by "make dtbs_check".
* tag 'renesas-arm-dt-for-v5.6-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-devel:
ARM: dts: sh73a0: Add missing clock-frequency for fixed clocks
ARM: dts: r8a7778: Add missing clock-frequency for fixed clocks
ARM: dts: rcar-gen2: Add missing mmio-sram bus properties
ARM: dts: rcar-gen2: Fix PCI high address in interrupt-map-mask
ARM: dts: renesas: Group tuples in pci ranges and dma-ranges properties
ARM: dts: renesas: Group tuples in interrupt properties
ARM: dts: renesas: Group tuples in regulator-gpio states properties
ARM: dts: r8a7779: Add device node for ARM global timer
ARM: dts: sh73a0: Add device node for ARM global timer
ARM: dts: sh73a0: Rename twd clock to periph clock
ARM: dts: iwg20d-q7-common: Add LCD support
Olof Johansson [Mon, 6 Jan 2020 17:22:26 +0000 (09:22 -0800)]
Merge tag 'ux500-armsoc-v5.6-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-stericsson into arm/dt
Support the Samsung GT-I8190/Golden phone:
- Proper include file for the AB8505 PMIC variant.
- Add a DTS file for the GT-I8190/Golden
- Extend the IMU, touch screen, WiFi and Bluetooth
as separate patches.
* tag 'ux500-armsoc-v5.6-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-stericsson:
ARM: dts: ux500: samsung-golden: Add Bluetooth
ARM: dts: ux500: samsung-golden: Add WiFi
ARM: dts: ux500: samsung-golden: Add touch screen
ARM: dts: ux500: samsung-golden: Add IMU (accelerometer + gyroscope)
ARM: dts: ux500: Add device tree for Samsung Galaxy S III mini (GT-I8190)
dt-bindings: arm: ux500: Document samsung,golden compatible
ARM: dts: ux500: Add device tree include for AB8505
ARM: dts: ux500: Remove unused ste-href-ab8505.dtsi
ARM: dts: sh73a0: Add missing clock-frequency for fixed clocks
"clock-frequency" is a required property for devices nodes compatible
with "fixed-clock", leading to warnings when running
$ make dtbs_check DT_SCHEMA_FILES=Documentation/devicetree/bindings/clock/fixed-clock.yaml
arch/arm/boot/dts/sh73a0-kzm9g.dt.yaml: extcki: 'clock-frequency' is a required property
Fix this by adding the missing "clock-frequency" properties to the various
clocks, to be overridden by the board DTS files when populated.
ARM: dts: r8a7778: Add missing clock-frequency for fixed clocks
"clock-frequency" is a required property for devices nodes compatible
with "fixed-clock", leading to warnings when running
$ make dtbs_check DT_SCHEMA_FILES=Documentation/devicetree/bindings/clock/fixed-clock.yaml
arch/arm/boot/dts/r8a7778-bockw.dt.yaml: audio_clk_a: 'clock-frequency' is a required property
arch/arm/boot/dts/r8a7778-bockw.dt.yaml: audio_clk_b: 'clock-frequency' is a required property
arch/arm/boot/dts/r8a7778-bockw.dt.yaml: audio_clk_c: 'clock-frequency' is a required property
Fix this by adding the missing "clock-frequency" properties to the audio
clocks, to be overridden by board DTS files when populated.
ARM: dts: rcar-gen2: Add missing mmio-sram bus properties
"#address-cells", "#size-cells", and "ranges" are required properties
for devices nodes compatible with "mmio-sram", leading to warnings when
running "make dtbs_check":
$ make dtbs_check DT_SCHEMA_FILES=Documentation/devicetree/bindings/sram/sram.yaml
arch/arm/boot/dts/r8a7791-koelsch.dt.yaml: sram@e63a0000: '#address-cells' is a required property
arch/arm/boot/dts/r8a7791-koelsch.dt.yaml: sram@e63a0000: '#size-cells' is a required property
arch/arm/boot/dts/r8a7791-koelsch.dt.yaml: sram@e63a0000: 'ranges' is a required property
...
Fix this by adding the missing properties to the mmio-sram device nodes
in the DTS files for all affected R-Car Gen2 and RZ/G1 SoCs.
ARM: dts: rcar-gen2: Fix PCI high address in interrupt-map-mask
"make dtbs_check" emits warnings like:
pci@ee090000: interrupt-map-mask:0:0: 65280 is greater than the maximum of 63488
pci@ee0b0000: interrupt-map-mask:0:0: 65280 is greater than the maximum of 63488
pci@ee0d0000: interrupt-map-mask:0:0: 65280 is greater than the maximum of 63488
According to dt-schemas/schemas/pci/pci-bus.yaml, the PCI high address
cell value in the "interrupt-map-mask" property must lie in the range
0..0xf800.
Fix this by correcting the values from 0xff00 to 0xf800 in all affected
R-Car Gen2 and RZ/G1 DTS files.
arm64: dts: renesas: Prepare for split of ARCH_R8A7795 into ARCH_R8A7795[01]
As R-Car H3 ES1.x (R8A77950) and R-Car H3 ES2.0+ (R8A77951) are really
different SoCs, CONFIG_ARCH_R8A7795 will be split in
CONFIG_ARCH_R8A77950 and CONFIG_ARCH_R8A77951.
Relax dependencies by handling both the old and the new symbols.
Sort the entries for the various DTBs in the Makefile by SoC and board
type. Keep Salvator-X(S) together, and do the same for ULCB with and
without Kingfisher extension.
arm64: dts: renesas: Drop redundant SoC prefixes from ULCB DTS file names
Unlike the V3MSK and V3HSK boards, the various "ULCB" boards are really
the same boards, with different SiPs fitted, just like the Salvator-X(S)
boards. Furthermore, the "H3", "M3", and "M3N" prefixes of the "ULCB"
parts in the DTS file names are redundant, as they are implied by the
SoC part numbers, which are also part of the file names.
Hence drop the redundant prefixes, to make the DTS file names consistent
with the file names for the various "Salvator-X(S)" boards.
arm64: dts: renesas: Rename r8a7795{-es1,}* to r8a7795[01]*
Despite using the same compatible values ("r8a7795"-based) because of
historical reasons, R-Car H3 ES1.x (R8A77950) and R-Car H3 ES2.0+
(R8A77951) are really different SoCs, with different part numbers.
Reflect this in the DTS files by changing their base names from
"r8a7795-es1" and "r8a7795" to "r8a77950" resp. "r8a77951".
Drop all "ES" references next to part numbers, as they are implied by
the part numbers, and thus redundant.
Note that DT binding headers, definitions, and compatible values are
not renamed, to preserve backward compatibility.
Stephan Gerhold [Thu, 19 Dec 2019 20:20:51 +0000 (21:20 +0100)]
ARM: dts: ux500: samsung-golden: Add Bluetooth
samsung-golden uses a BCM4334 WiFi+BT combo chip.
The BT part is connected via UART and supported by the hci_bcm
driver in mainline.
Add the necessary device tree changes to make it load correctly.
It requires (seemingly) device-specific firmware that can be
extracted from the stock Android system used on samsung-golden:
- /system/bin/bcm4334.hcd -> /lib/firmware/brcm/BCM4334B0.hcd
On my device, scanning for other Bluetooth devices works just fine,
but for some reason it keeps disconnecting immediately
when attempting to connect to an other device.
Stephan Gerhold [Thu, 19 Dec 2019 20:20:50 +0000 (21:20 +0100)]
ARM: dts: ux500: samsung-golden: Add WiFi
samsung-golden uses a BCM4334 WiFi+BT combo chip, connected to SDIO.
It is supported by the brcmfmac driver in mainline,
so we only need to set up the device tree to make it work correctly.
Note: brcmfmac requires (proprietary) firmware + a device-specific
NVRAM file. Both can be extracted from the stock Android system
used on samsung-golden:
- /system/etc/wifi/bcmdhd_sta.bin_b2 -> /lib/firmware/brcm/brcmfmac4334-sdio.bin
- /system/etc/wifi/nvram_net.txt_GPIO4 -> /lib/firmware/brcm/brcmfmac4334-sdio.samsung,golden.txt
brcmfmac4334-sdio.bin from linux-firmware also seems to work,
but results in occasional errors for some reason.
Stephan Gerhold [Thu, 19 Dec 2019 20:20:49 +0000 (21:20 +0100)]
ARM: dts: ux500: samsung-golden: Add touch screen
samsung-golden has an Atmel mXT224S touch controller connected to I2C.
It is supported by the existing driver for atmel,maxtouch, so all we
need to do to make it work is to define the necessary device tree nodes.
The atmel_mxt_ts driver does not support controlling regulators yet,
so add regulator-always-on for now to turn on the necessary regulators.
samsung-golden has a InvenSense MPU-6051M IMU that provides an
accelerometer and gyroscope. It seems to be functionally compatible
with MPU-6050 so we can easily enable it by adding the necessary
device tree nodes.
Stephan Gerhold [Thu, 19 Dec 2019 20:20:47 +0000 (21:20 +0100)]
ARM: dts: ux500: Add device tree for Samsung Galaxy S III mini (GT-I8190)
The Samsung Galaxy S III mini (GT-I8190) is a smartphone with Ux500 SoC
released in 2012. Thanks to the great mainline support for Ux500,
it can actually run mainline Linux quite well.
Add a new device tree for it with support for:
- Internal Storage (eMMC)
- External Storage (Micro SD card)
- UART
- GPIO Buttons
- Vibrator
Note that the device tree cannot be booted directly with
the original (Samsung) bootloader. It keeps the L2 cache turned on,
which causes the kernel to hang shortly after decompression.
As a workaround I have created a port of (mainline) U-Boot,
which locks the L2 cache before booting Linux. At the moment it does not
replace the Samsung bootloader, instead we let the original bootloader
load U-Boot as an another (intermediate) bootloader.
Another advantage of this is that U-Boot has proper device tree support,
so we do not need to hardcode the kernel command line in the device tree.
Stephan Gerhold [Thu, 19 Dec 2019 20:20:45 +0000 (21:20 +0100)]
ARM: dts: ux500: Add device tree include for AB8505
AB8505 is a slightly newer version of AB8500.
Overall it is quite similar, but there are some differences like
the number of GPIOs and regulators. Therefore we need a separate
device tree definition for devices making use of AB8505.
The AB8500-specific nodes were moved out of ste-dbx5x0.dtsi in
commit a46f7c6762d8 ("ARM: dts: ux500: Move ab8500 nodes to ste-ab8500.dtsi").
Add a new "ste-ab8505.dtsi" device tree include in a similar way.
Keep the battery/charging related sub-devices disabled by default
since they require additional configuration to work correctly.
The pin configuration for HREF boards with AB8505 was added in
commit 77ad9dfc2c7e ("ARM: ux500: move last AB8505 set-up to DT").
As the commit message notes, it was unused back then and it has
remained so even today, especially considering AB8505 did not have
proper device tree support until recently.
We are now preparing to add support for some Samsung smartphones
that are using AB8505. However, they use different pin configs
because using ste-href-ab8505.dtsi is known to break UART.
There were not many HREFs with AB8505, so at this point it seems
unlikely that we will ever make use of this include. Remove it.
Fabrizio Castro [Tue, 17 Dec 2019 13:46:01 +0000 (13:46 +0000)]
arm64: dts: renesas: Add EK874 board with idk-2121wr display support
The EK874 is advertised as compatible with panel IDK-2121WR from
Advantech, however the panel isn't sold alongside the board.
A new dts, adding everything that's required to get the panel to
to work with the EK874, is the most convenient way to support the
EK874 when it's connected to the IDK-2121WR.
Add device nodes for the I2C and IIC Controllers on the Renesas R-Car
M3-W+ (r8a77961) SoC, including DMA properties linking them to the DMA
controllers.
arm64: dts: renesas: r8a77990: ebisu: Remove clkout-lr-synchronous from sound
rcar_sound doesn't support clkout-lr-synchronous in upstream.
It was supported under out-of-tree rcar_sound.
upstream rcar_sound is supporting
- clkout-lr-synchronous
+ clkout-lr-asynchronous
arm64: dts: renesas: Group tuples in pci ranges and dma-ranges properties
To improve human readability and enable automatic validation, the tuples
in the "ranges" and "dma-ranges" properties of PCI device nodes should
be grouped.
Fix this by grouping the tuples of the "ranges" and "dma-ranges"
properties using angle brackets.
arm64: dts: renesas: Group tuples in interrupt properties
To improve human readability and enable automatic validation, the tuples
in the various properties containing interrupt specifiers should be
grouped. While "make dtbs_check" does not impose this yet for the
"interrupts" property, it does for the "interrupt-map" property.
Fix this by grouping the tuples of the "interrupts" and "interrupt-map"
properties using angle brackets.
arm64: dts: renesas: Group tuples in regulator-gpio states properties
To improve human readability and enable automatic validation, the tuples
in the "states" properties of device nodes compatible with
"regulator-gpio" should be grouped, as reported by "make dtbs_check":
$ make dtbs_check DT_SCHEMA_FILES=Documentation/devicetree/bindings/regulator/gpio-regulator.yaml
arch/arm64/boot/dts/renesas/r8a7795-salvator-x.dt.yaml: regulator-vccq-sdhi0: states:0: Additional items are not allowed (1800000, 0 were unexpected)
arch/arm64/boot/dts/renesas/r8a7795-salvator-x.dt.yaml: regulator-vccq-sdhi0: states:0: [3300000, 1, 1800000, 0] is too long
arch/arm64/boot/dts/renesas/r8a7795-salvator-x.dt.yaml: regulator-vccq-sdhi3: states:0: Additional items are not allowed (1800000, 0 were unexpected)
arch/arm64/boot/dts/renesas/r8a7795-salvator-x.dt.yaml: regulator-vccq-sdhi3: states:0: [3300000, 1, 1800000, 0] is too long
...
Fix this by grouping the tuples using angle brackets.
ARM: dts: renesas: Group tuples in pci ranges and dma-ranges properties
To improve human readability and enable automatic validation, the tuples
in the "ranges" and "dma-ranges" properties of PCI devices nodes should
be grouped. Not doing so causes "make dtbs_check" to emit warnings
like:
ARM: dts: renesas: Group tuples in interrupt properties
To improve human readability and enable automatic validation, the tuples
in the various properties containing interrupt specifiers should be
grouped. While "make dtbs_check" does not impose this yet for the
"interrupts" property, it does for the "interrupt-map" property, leading
to warnings like:
pci@ee090000: interrupt-map:0: [0, 0, 0, 1, 5, 0, 108, 4, 2048, 0, 0, 1, 5, 0, 108, 4, 4096, 0, 0, 2, 5, 0, 108, 4] is too long
pci@ee0d0000: interrupt-map:0: [0, 0, 0, 1, 5, 0, 113, 4, 2048, 0, 0, 1, 5, 0, 113, 4, 4096, 0, 0, 2, 5, 0, 113, 4] is too long
Fix this by grouping the tuples of the "interrupts" and "interrupt-map"
properties using angle brackets.
ARM: dts: renesas: Group tuples in regulator-gpio states properties
To improve human readability and enable automatic validation, the tuples
in the "states" properties of device nodes compatible with
"regulator-gpio" should be grouped, as reported by "make dtbs_check":
$ make dtbs_check DT_SCHEMA_FILES=Documentation/devicetree/bindings/regulator/gpio-regulator.yaml
arch/arm/boot/dts/r8a7791-koelsch.dt.yaml: regulator-vccq-sdhi0: states:0: Additional items are not allowed (1800000, 0 were unexpected)
arch/arm/boot/dts/r8a7791-koelsch.dt.yaml: regulator-vccq-sdhi0: states:0: [3300000, 1, 1800000, 0] is too long
arch/arm/boot/dts/r8a7791-koelsch.dt.yaml: regulator-vccq-sdhi1: states:0: Additional items are not allowed (1800000, 0 were unexpected)
arch/arm/boot/dts/r8a7791-koelsch.dt.yaml: regulator-vccq-sdhi1: states:0: [3300000, 1, 1800000, 0] is too long
arch/arm/boot/dts/r8a7791-koelsch.dt.yaml: regulator-vccq-sdhi2: states:0: Additional items are not allowed (1800000, 0 were unexpected)
arch/arm/boot/dts/r8a7791-koelsch.dt.yaml: regulator-vccq-sdhi2: states:0: [3300000, 1, 1800000, 0] is too long
...
Fix this by grouping the tuples using angle brackets.
ARM: dts: sh73a0: Rename twd clock to periph clock
The "TWD" clock is actually the Cortex-A9 MPCore "PERIPHCLK" clock,
which not only clocks the private timers and watchdogs (TWD), but also
the interrupt controller and global timer.
Olof Johansson [Tue, 17 Dec 2019 18:03:56 +0000 (10:03 -0800)]
Merge tag 'ux500-armsoc-v5.6-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-stericsson into arm/dt
First set of Ux500 DTS changes for the v5.6 kernel:
- Add the GPADC IIO channels
- Factor out generic pin configuration
- Add the gpio_in_nopull configuration
- Tighten up I2C and SPI buses
- Clean up some compatibles
- Extract a generic DB8500 DTSI
- Add HREF520 DTS and the associated DB8520 DTSI
- Split TVK R2 and R3 to separate DTSI files
* tag 'ux500-armsoc-v5.6-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-stericsson:
ARM: dts: ux500: Add devicetree for HREF520
ARM: dts: ux500: Split TVK DTSI files in two
ARM: dts: ux500: Break out DB8500 DTSI
ARM: dts: ux500: Drop pulls on I2C buses
ARM: dts: ux500: Use "arm,pl031" compatible for PL031
ARM: dts: ux500: Add "simple-bus" compatible to soc node
ARM: dts: ux500: Remove ux500_ prefix from ux500_serial* labels
ARM: dts: ux500: Move serial aliases to ste-dbx5x0.dtsi
ARM: dts: ux500: Add aliases for I2C and SPI buses
ARM: dts: ux500: Disable I2C/SPI buses by default
ARM: dts: ux500: nomadik-pinctrl: Add &gpio_in_nopull
ARM: dts: ux500: Add pin configs for UART1 CTS/RTS pins
ARM: dts: ux500: Add alternative SDI pin configs
ARM: dts: ux500: Rename generic pin configs according to pin group
ARM: dts: ux500: Move generic pin configs out of ste-href-family-pinctrl.dtsi
dt-bindings: arm: Document compatibles for Ux500 boards
ARM: dts: ux500: snowball: Remove unused PRCMU cpufreq node
ARM: dts: ux500: declare GPADC IIO ADC channels
CONFIG_ARCH_R8A7796 was renamed to CONFIG_ARCH_R8A77960 in commit 39e57e14d7eaf818 ("soc: renesas: Add ARCH_R8A77960 for existing R-Car
M3-W"), to avoid confusion between R-Car M3-W (R8A77960) and R-Car M3-W+
(R8A77961).
Rename the DTS files as well, for consistency.
Note that DT binding headers, definitions, and compatible values were
not renamed, to preserve backward compatibility.
CONFIG_ARCH_R8A7796 was renamed to CONFIG_ARCH_R8A77960 in commit 39e57e14d7eaf818 ("soc: renesas: Add ARCH_R8A77960 for existing R-Car
M3-W"), so its users can be removed.
Olof Johansson [Mon, 9 Dec 2019 17:22:47 +0000 (09:22 -0800)]
Merge tag 'socfpga_dts_updates_for_v5.5_part2' of git://git.kernel.org/pub/scm/linux/kernel/git/dinguyen/linux into arm/dt
SoCFPGA DTS updates for v5.5, part 2
- Add NAND support for both Agilex and Stratix10 boards
- Agilex
- Add FPGA manager and Service Layer support
- Add EDAC support
- Add System manager
- Add System manager property to ethernet nodes
* tag 'socfpga_dts_updates_for_v5.5_part2' of git://git.kernel.org/pub/scm/linux/kernel/git/dinguyen/linux:
arm64: dts: agilex: Add SysMgr to Ethernet nodes
arm64: dts: agilex: Add SysMgr compatible
arm64: dts: agilex: Add EDAC Device Tree
arm64: dts: add NAND board files for Stratix10 and Agilex
arm64: dts: agilex: add NAND IP to base dts
Olof Johansson [Mon, 9 Dec 2019 17:21:43 +0000 (09:21 -0800)]
Merge tag 'samsung-dt-5.5-2' of https://git.kernel.org/pub/scm/linux/kernel/git/krzk/linux into arm/dt
Samsung DTS ARM changes for v5.5, part 2
1. Cleanup by adjusting DTS to bindings,
2. Add touch-sensitive buttons to Midas (Galaxy S III family phones),
3. Add GPU/Mali to Exynos542x and Odroid XU3/XU4 family.
* tag 'samsung-dt-5.5-2' of https://git.kernel.org/pub/scm/linux/kernel/git/krzk/linux:
ARM: dts: exynos: Add Mali/GPU node on Exynos5420 and enable it on Odroid XU3/4
ARM: dts: exynos: Add support for the touch-sensitive buttons on Midas family
ARM: dts: exynos: Rename children of SysRAM node to "sram"
Linus Walleij [Tue, 26 Nov 2019 12:47:37 +0000 (13:47 +0100)]
ARM: dts: ux500: Add devicetree for HREF520
This reference design is very similar to the others just that
it has a different display mounted on the user interface
board, and some GPIOs where shuffled around.
As this is the first board that uses DB8520 we create the
DB8520-specific DTSI file here.
Linus Walleij [Tue, 26 Nov 2019 12:47:35 +0000 (13:47 +0100)]
ARM: dts: ux500: Break out DB8500 DTSI
The DB8500 exists in an enhanced variant named DB8520
for some machines. To clearly distinguish between the
different machines, create an explicit db8500.dtsi
and move the operating points (only known difference so
far) to that file, so we can add an explicit db8520.dtsi
after this.
Linus Walleij [Tue, 26 Nov 2019 12:31:16 +0000 (13:31 +0100)]
ARM: dts: ux500: Drop pulls on I2C buses
The I2C block in the Ux500 uses internal pull-ups on the
SoC, in fact it has to: in HS mode, the I2C block will need
to autonomously take control over the pull-up line to do
its job. This can be clearly seen from the SoC manual which
states that the silicon has a line named "en_cspu_hs" which
enables current source pull-up for high speed mode. Another
hint is that the vendor code tree never enabled the pull
up on these lines, despite being deployed on boards that
lack external pull-up resistors.
Tested on the Ux500 reference designs without any problems.
Cc: Stephan Gerhold <stephan@gerhold.net> Reported-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20191126123116.56244-1-linus.walleij@linaro.org Reviewed-by: Stephan Gerhold <stephan@gerhold.net> Tested-by: Stephan Gerhold <stephan@gerhold.net> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Stephan Gerhold [Sun, 24 Nov 2019 20:51:10 +0000 (21:51 +0100)]
ARM: dts: ux500: Use "arm,pl031" compatible for PL031
The Ux500 device tree uses "arm,rtc-pl031" as compatible for PL031.
All other boards in Linux describe it using "arm,pl031" instead.
This works because the compatible is not actually used in Linux:
AMBA devices get probed based on "arm,primecell" and their peripheral ID.
Nevertheless, some other projects (e.g. U-Boot) rely on the compatible
to probe the device with the correct driver. Those will look for
"arm,pl031" instead of "arm,rtc-pl031", preventing the RTC from being
probed.
Change it to "arm,pl031" to match all other boards.
Stephan Gerhold [Sun, 24 Nov 2019 19:57:28 +0000 (20:57 +0100)]
ARM: dts: ux500: Add "simple-bus" compatible to soc node
The "soc" node in the Ux500 device tree does not need any special
handling - it is just a simple I/O bus that can be accessed without
additional configuration.
Therefore we can additionally describe it as compatible with "simple-bus".
This can be used by platforms to probe devices under the soc node without
special handling for our custom "stericsson,db8500" compatible
(e.g. in U-Boot).
Stephan Gerhold [Mon, 25 Nov 2019 17:04:28 +0000 (18:04 +0100)]
ARM: dts: ux500: Remove ux500_ prefix from ux500_serial* labels
ux500_serial{0,1,2} are the only labels with ux500_ prefix in
ste-dbx5x0.dtsi, the other labels (gpio0, msp, ...) do not use
any prefix. Remove it for consistency.
Stephan Gerhold [Mon, 25 Nov 2019 17:04:27 +0000 (18:04 +0100)]
ARM: dts: ux500: Move serial aliases to ste-dbx5x0.dtsi
Now that we have aliases for I2C and SPI in ste-dbx5x0.dtsi,
it does not make much sense to keep only the aliases for UART
separately in each board device tree.
Considering that all boards set the same aliases for the serial
ports there is no reason to keep them separated either.
Move them to ste-dbx5x0.dtsi and remove the aliases from the
board-specific device tree parts.
Stephan Gerhold [Mon, 25 Nov 2019 17:04:26 +0000 (18:04 +0100)]
ARM: dts: ux500: Add aliases for I2C and SPI buses
Now that we disable the I2C/SPI buses by default, is is even more
important to assign aliases to the I2C/SPI device nodes.
Otherwise, enabling/disabling one of them will potentially change
all device IDs, e.g. i2c2 will be named i2c-0 if it is the only
enabled I2C bus.
Add aliases for the I2C and SPI buses to avoid this.
Stephan Gerhold [Mon, 25 Nov 2019 17:04:25 +0000 (18:04 +0100)]
ARM: dts: ux500: Disable I2C/SPI buses by default
At the moment, all 5 I2C and 6 SPI buses are probed and exposed
to user-space by default - even if they are not muxed to any pins
on the board. This means that user-space sees an I2C/SPI bus that
cannot be actually used properly.
In some cases this was used to put the corresponding pins into
a low power sleep mode - but even then the pins first need to be
configured by the board-specific device tree part.
Avoid exposing unconfigured devices to user-space by disabling
the I2C/SPI buses by default. Enable them in the board device trees
when needed.
ste-nomadik-pinctrl.dtsi already defines in_nopull and gpio_in_pu/pd,
but there is no node to configure a pin as GPIO without pull up/down.
Add a new &gpio_in_nopull node for this.
Stephan Gerhold [Mon, 25 Nov 2019 12:22:54 +0000 (13:22 +0100)]
ARM: dts: ux500: Add alternative SDI pin configs
SDI0/SDI1 can be used in configurations where some of the pins
(e.g. direction control) are not used. The pinctrl driver has
separate pin groups for them.
Add new pin configurations for:
- mc0_a_2: like mc0_a_1, but without CMDDIR/DAT0DIR/DAT2DIR
- mc1_a_2: like mc1_a_1, but without FBCLK
Stephan Gerhold [Mon, 25 Nov 2019 12:22:53 +0000 (13:22 +0100)]
ARM: dts: ux500: Rename generic pin configs according to pin group
Some components (e.g. SDI, I2C) can be used with different pin assignments.
Before we can add the alternative configurations, we need to rename the
current configurations to more generic names.
Each pin configuration usually configures one specific pin group.
Therefore we rename the configurations to use the pin group as name.
Make up for the slightly longer names by removing the "_mode" suffix.
Stephan Gerhold [Mon, 25 Nov 2019 12:22:52 +0000 (13:22 +0100)]
ARM: dts: ux500: Move generic pin configs out of ste-href-family-pinctrl.dtsi
All existing Ux500 boards make use of ste-href-family-pinctrl.dtsi,
which contains shared pin configurations for UART, I2C and SDI.
Most of these can be also used for devices not based on HREF.
Move the generic pin configs into a new device tree include
"ste-dbx5x0-pinctrl.dtsi". There is no functional change (yet),
as a next step we will rename the pin configs to use more generic
names.
Commit a435adbec264 ("ARM: dts: augment Ux500 to use DT cpufreq")
switched the Ux500 device tree to use the generic DT cpufreq driver
and removed the PRCMU cpufreq node.
The snowball DTS still references it, without effect, since cpufreq
is now enabled by default. Remove the unused node.
Linus Walleij [Tue, 10 Jan 2017 22:50:52 +0000 (23:50 +0100)]
ARM: dts: ux500: declare GPADC IIO ADC channels
This adds the IIO channels for the GPADC after converting it
to using the standard IIO ADC bindings and moving the driver
over to the IIO subsystem. We also add IIO hwmon standard
driver node to support reading channels in a standard manner.
1) More jumbo frame fixes in r8169, from Heiner Kallweit.
2) Fix bpf build in minimal configuration, from Alexei Starovoitov.
3) Use after free in slcan driver, from Jouni Hogander.
4) Flower classifier port ranges don't work properly in the HW offload
case, from Yoshiki Komachi.
5) Use after free in hns3_nic_maybe_stop_tx(), from Yunsheng Lin.
6) Out of bounds access in mqprio_dump(), from Vladyslav Tarasiuk.
7) Fix flow dissection in dsa TX path, from Alexander Lobakin.
8) Stale syncookie timestampe fixes from Guillaume Nault.
[ Did an evil merge to silence a warning introduced by this pull - Linus ]
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (84 commits)
r8169: fix rtl_hw_jumbo_disable for RTL8168evl
net_sched: validate TCA_KIND attribute in tc_chain_tmplt_add()
r8169: add missing RX enabling for WoL on RTL8125
vhost/vsock: accept only packets with the right dst_cid
net: phy: dp83867: fix hfs boot in rgmii mode
net: ethernet: ti: cpsw: fix extra rx interrupt
inet: protect against too small mtu values.
gre: refetch erspan header from skb->data after pskb_may_pull()
pppoe: remove redundant BUG_ON() check in pppoe_pernet
tcp: Protect accesses to .ts_recent_stamp with {READ,WRITE}_ONCE()
tcp: tighten acceptance of ACKs not matching a child socket
tcp: fix rejected syncookies due to stale timestamps
lpc_eth: kernel BUG on remove
tcp: md5: fix potential overestimation of TCP option space
net: sched: allow indirect blocks to bind to clsact in TC
net: core: rename indirect block ingress cb function
net-sysfs: Call dev_hold always in netdev_queue_add_kobject
net: dsa: fix flow dissection on Tx path
net/tls: Fix return values to avoid ENOTSUPP
net: avoid an indirect call in ____sys_recvmsg()
...
Linus Torvalds [Sun, 8 Dec 2019 20:23:42 +0000 (12:23 -0800)]
Merge tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Pull more SCSI updates from James Bottomley:
"Eleven patches, all in drivers (no core changes) that are either minor
cleanups or small fixes.
They were late arriving, but still safe for -rc1"
* tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
scsi: MAINTAINERS: Add the linux-scsi mailing list to the ISCSI entry
scsi: megaraid_sas: Make poll_aen_lock static
scsi: sd_zbc: Improve report zones error printout
scsi: qla2xxx: Fix qla2x00_request_irqs() for MSI
scsi: qla2xxx: unregister ports after GPN_FT failure
scsi: qla2xxx: fix rports not being mark as lost in sync fabric scan
scsi: pm80xx: Remove unused include of linux/version.h
scsi: pm80xx: fix logic to break out of loop when register value is 2 or 3
scsi: scsi_transport_sas: Fix memory leak when removing devices
scsi: lpfc: size cpu map by last cpu id set
scsi: ibmvscsi_tgt: Remove unneeded variable rc
Linus Torvalds [Sun, 8 Dec 2019 20:12:18 +0000 (12:12 -0800)]
Merge tag '5.5-rc-smb3-fixes-part2' of git://git.samba.org/sfrench/cifs-2.6
Pull cifs fixes from Steve French:
"Nine cifs/smb3 fixes:
- one fix for stable (oops during oplock break)
- two timestamp fixes including important one for updating mtime at
close to avoid stale metadata caching issue on dirty files (also
improves perf by using SMB2_CLOSE_FLAG_POSTQUERY_ATTRIB over the
wire)
- two fixes for "modefromsid" mount option for file create (now
allows mode bits to be set more atomically and accurately on create
by adding "sd_context" on create when modefromsid specified on
mount)
- two fixes for multichannel found in testing this week against
different servers
- two small cleanup patches"
* tag '5.5-rc-smb3-fixes-part2' of git://git.samba.org/sfrench/cifs-2.6:
smb3: improve check for when we send the security descriptor context on create
smb3: fix mode passed in on create for modetosid mount option
cifs: fix possible uninitialized access and race on iface_list
cifs: Fix lookup of SMB connections on multichannel
smb3: query attributes on file close
smb3: remove unused flag passed into close functions
cifs: remove redundant assignment to pointer pneg_ctxt
fs: cifs: Fix atime update check vs mtime
CIFS: Fix NULL-pointer dereference in smb2_push_mandatory_locks
Linus Torvalds [Sun, 8 Dec 2019 19:08:28 +0000 (11:08 -0800)]
Merge branch 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull misc vfs cleanups from Al Viro:
"No common topic, just three cleanups".
* 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
make __d_alloc() static
fs/namespace: add __user to open_tree and move_mount syscalls
fs/fnctl: fix missing __user in fcntl_rw_hint()
Linus Torvalds [Sun, 8 Dec 2019 01:07:18 +0000 (17:07 -0800)]
Merge tag 'iomap-5.5-merge-14' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux
Pull iomap fixes from Darrick Wong:
"Fix a race condition and a use-after-free error:
- Fix a UAF when reporting writeback errors
- Fix a race condition when handling page uptodate on fragmented file
with blocksize < pagesize"
* tag 'iomap-5.5-merge-14' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
iomap: stop using ioend after it's been freed in iomap_finish_ioend()
iomap: fix sub-page uptodate handling
Linus Torvalds [Sun, 8 Dec 2019 01:05:33 +0000 (17:05 -0800)]
Merge tag 'xfs-5.5-merge-17' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux
Pull xfs fixes from Darrick Wong:
"Fix a couple of resource management errors and a hang:
- fix a crash in the log setup code when log mounting fails
- fix a hang when allocating space on the realtime device
- fix a block leak when freeing space on the realtime device"
* tag 'xfs-5.5-merge-17' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
xfs: fix mount failure crash on invalid iclog memory access
xfs: don't check for AG deadlock for realtime files in bunmapi
xfs: fix realtime file data space leak
Linus Torvalds [Sun, 8 Dec 2019 00:59:25 +0000 (16:59 -0800)]
Merge tag 'for-linus-5.5-ofs1' of git://git.kernel.org/pub/scm/linux/kernel/git/hubcap/linux
Pull orangefs update from Mike Marshall:
"orangefs: posix open permission checking...
Orangefs has no open, and orangefs checks file permissions on each
file access. Posix requires that file permissions be checked on open
and nowhere else. Orangefs-through-the-kernel needs to seem posix
compliant.
The VFS opens files, even if the filesystem provides no method. We can
see if a file was successfully opened for read and or for write by
looking at file->f_mode.
When writes are flowing from the page cache, file is no longer
available. We can trust the VFS to have checked file->f_mode before
writing to the page cache.
The mode of a file might change between when it is opened and IO
commences, or it might be created with an arbitrary mode.
We'll make sure we don't hit EACCES during the IO stage by using
UID 0"
[ This is "posixish", but not a great solution in the long run, since a
proper secure network server shouldn't really trust the client like this.
But proper and secure POSIX behavior requires an open method and a
resulting cookie for IO of some kind, or similar. - Linus ]
* tag 'for-linus-5.5-ofs1' of git://git.kernel.org/pub/scm/linux/kernel/git/hubcap/linux:
orangefs: posix open permission checking...
Linus Torvalds [Sun, 8 Dec 2019 00:56:00 +0000 (16:56 -0800)]
Merge tag 'nfsd-5.5' of git://linux-nfs.org/~bfields/linux
Pull nfsd updates from Bruce Fields:
"This is a relatively quiet cycle for nfsd, mainly various bugfixes.
Possibly most interesting is Trond's fixes for some callback races
that were due to my incomplete understanding of rpc client shutdown.
Unfortunately at the last minute I've started noticing a new
intermittent failure to send callbacks. As the logic seems basically
correct, I'm leaving Trond's patches in for now, and hope to find a
fix in the next week so I don't have to revert those patches"
* tag 'nfsd-5.5' of git://linux-nfs.org/~bfields/linux: (24 commits)
nfsd: depend on CRYPTO_MD5 for legacy client tracking
NFSD fixing possible null pointer derefering in copy offload
nfsd: check for EBUSY from vfs_rmdir/vfs_unink.
nfsd: Ensure CLONE persists data and metadata changes to the target file
SUNRPC: Fix backchannel latency metrics
nfsd: restore NFSv3 ACL support
nfsd: v4 support requires CRYPTO_SHA256
nfsd: Fix cld_net->cn_tfm initialization
lockd: remove __KERNEL__ ifdefs
sunrpc: remove __KERNEL__ ifdefs
race in exportfs_decode_fh()
nfsd: Drop LIST_HEAD where the variable it declares is never used.
nfsd: document callback_wq serialization of callback code
nfsd: mark cb path down on unknown errors
nfsd: Fix races between nfsd4_cb_release() and nfsd4_shutdown_callback()
nfsd: minor 4.1 callback cleanup
SUNRPC: Fix svcauth_gss_proxy_init()
SUNRPC: Trace gssproxy upcall results
sunrpc: fix crash when cache_head become valid before update
nfsd: remove private bin2hex implementation
...
Linus Torvalds [Sun, 8 Dec 2019 00:50:55 +0000 (16:50 -0800)]
Merge tag 'nfs-for-5.5-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs
Pull NFS client updates from Trond Myklebust:
"Highlights include:
Features:
- NFSv4.2 now supports cross device offloaded copy (i.e. offloaded
copy of a file from one source server to a different target
server).
- New RDMA tracepoints for debugging congestion control and Local
Invalidate WRs.
Bugfixes and cleanups
- Drop the NFSv4.1 session slot if nfs4_delegreturn_prepare waits for
layoutreturn
- Handle bad/dead sessions correctly in nfs41_sequence_process()
- Various bugfixes to the delegation return operation.
- Various bugfixes pertaining to delegations that have been revoked.
- Cleanups to the NFS timespec code to avoid unnecessary conversions
between timespec and timespec64.
- Fix unstable RDMA connections after a reconnect
- Close race between waking an RDMA sender and posting a receive
- Wake pending RDMA tasks if connection fails
- Fix MR list corruption, and clean up MR usage
- Fix another RPCSEC_GSS issue with MIC buffer space"
* tag 'nfs-for-5.5-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs: (79 commits)
SUNRPC: Capture completion of all RPC tasks
SUNRPC: Fix another issue with MIC buffer space
NFS4: Trace lock reclaims
NFS4: Trace state recovery operation
NFSv4.2 fix memory leak in nfs42_ssc_open
NFSv4.2 fix kfree in __nfs42_copy_file_range
NFS: remove duplicated include from nfs4file.c
NFSv4: Make _nfs42_proc_copy_notify() static
NFS: Fallocate should use the nfs4_fattr_bitmap
NFS: Return -ETXTBSY when attempting to write to a swapfile
fs: nfs: sysfs: Remove NULL check before kfree
NFS: remove unneeded semicolon
NFSv4: add declaration of current_stateid
NFSv4.x: Drop the slot if nfs4_delegreturn_prepare waits for layoutreturn
NFSv4.x: Handle bad/dead sessions correctly in nfs41_sequence_process()
nfsv4: Move NFSPROC4_CLNT_COPY_NOTIFY to end of list
SUNRPC: Avoid RPC delays when exiting suspend
NFS: Add a tracepoint in nfs_fh_to_dentry()
NFSv4: Don't retry the GETATTR on old stateid in nfs4_delegreturn_done()
NFSv4: Handle NFS4ERR_OLD_STATEID in delegreturn
...
Steve French [Sat, 7 Dec 2019 23:38:22 +0000 (17:38 -0600)]
smb3: improve check for when we send the security descriptor context on create
We had cases in the previous patch where we were sending the security
descriptor context on SMB3 open (file create) in cases when we hadn't
mounted with with "modefromsid" mount option.
Add check for that mount flag before calling ad_sd_context in
open init.
Signed-off-by: Steve French <stfrench@microsoft.com> Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
Linus Torvalds [Sat, 7 Dec 2019 22:51:04 +0000 (14:51 -0800)]
Merge tag 'vfio-v5.5-rc1' of git://github.com/awilliam/linux-vfio
Pull VFIO updates from Alex Williamson:
- Remove hugepage checks for reserved pfns (Ben Luo)
- Fix irq-bypass unregister ordering (Jiang Yi)
* tag 'vfio-v5.5-rc1' of git://github.com/awilliam/linux-vfio:
vfio/pci: call irq_bypass_unregister_producer() before freeing irq
vfio/type1: remove hugepage checks in is_invalid_reserved_pfn()
Linus Torvalds [Sat, 7 Dec 2019 22:49:20 +0000 (14:49 -0800)]
Merge tag 'for-linus-5.5b-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip
Pull more xen updates from Juergen Gross:
- a patch to fix a build warning
- a cleanup of no longer needed code in the Xen event handling
- a small series for the Xen grant driver avoiding high order
allocations and replacing an insane global limit by a per-call one
- a small series fixing Xen frontend/backend module referencing
* tag 'for-linus-5.5b-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
xen-blkback: allow module to be cleanly unloaded
xen/xenbus: reference count registered modules
xen/gntdev: switch from kcalloc() to kvcalloc()
xen/gntdev: replace global limit of mapped pages by limit per call
xen/gntdev: remove redundant non-zero check on ret
xen/events: remove event handling recursion detection
Linus Torvalds [Sat, 7 Dec 2019 22:43:46 +0000 (14:43 -0800)]
Merge branch 'akpm' (patches from Andrew)
Merge misc Kconfig updates from Andrew Morton:
"A number of changes to Kconfig files under lib/ from Changbin Du and
Krzysztof Kozlowski"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
lib/: fix Kconfig indentation
kernel-hacking: move DEBUG_FS to 'Generic Kernel Debugging Instruments'
kernel-hacking: move DEBUG_BUGVERBOSE to 'printk and dmesg options'
kernel-hacking: create a submenu for scheduler debugging options
kernel-hacking: move SCHED_STACK_END_CHECK after DEBUG_STACK_USAGE
kernel-hacking: move Oops into 'Lockups and Hangs'
kernel-hacking: move kernel testing and coverage options to same submenu
kernel-hacking: group kernel data structures debugging together
kernel-hacking: create submenu for arch special debugging options
kernel-hacking: group sysrq/kgdb/ubsan into 'Generic Kernel Debugging Instruments'
Heiner Kallweit [Sat, 7 Dec 2019 21:21:52 +0000 (22:21 +0100)]
r8169: fix rtl_hw_jumbo_disable for RTL8168evl
In referenced fix we removed the RTL8168e-specific jumbo config for
RTL8168evl in rtl_hw_jumbo_enable(). We have to do the same in
rtl_hw_jumbo_disable().
v2: fix referenced commit id
Fixes: 14012c9f3bb9 ("r8169: fix jumbo configuration for RTL8168evl") Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Linus Torvalds [Sat, 7 Dec 2019 21:53:09 +0000 (13:53 -0800)]
pipe: don't use 'pipe_wait() for basic pipe IO
pipe_wait() may be simple, but since it relies on the pipe lock, it
means that we have to do the wakeup while holding the lock. That's
unfortunate, because the very first thing the waked entity will want to
do is to get the pipe lock for itself.
So get rid of the pipe_wait() usage by simply releasing the pipe lock,
doing the wakeup (if required) and then using wait_event_interruptible()
to wait on the right condition instead.
wait_event_interruptible() handles races on its own by comparing the
wakeup condition before and after adding itself to the wait queue, so
you can use an optimistic unlocked condition for it.
Cc: David Howells <dhowells@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds [Sat, 7 Dec 2019 21:21:01 +0000 (13:21 -0800)]
pipe: remove 'waiting_writers' merging logic
This code is ancient, and goes back to when we only had a single page
for the pipe buffers. The exact history is hidden in the mists of time
(ie "before git", and in fact predates the BK repository too).
At that long-ago point in time, it actually helped to try to merge big
back-and-forth pipe reads and writes, and not limit pipe reads to the
single pipe buffer in length just because that was all we had at a time.
However, since then we've expanded the pipe buffers to multiple pages,
and this logic really doesn't seem to make sense. And a lot of it is
somewhat questionable (ie "hmm, the user asked for a non-blocking read,
but we see that there's a writer pending, so let's wait anyway to get
the extra data that the writer will have").
But more importantly, it makes the "go to sleep" logic much less
obvious, and considering the wakeup issues we've had, I want to make for
less of those kinds of things.
Cc: David Howells <dhowells@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds [Sat, 7 Dec 2019 20:54:26 +0000 (12:54 -0800)]
pipe: fix and clarify pipe read wakeup logic
This is the read side version of the previous commit: it simplifies the
logic to only wake up waiting writers when necessary, and makes sure to
use a synchronous wakeup. This time not so much for GNU make jobserver
reasons (that pipe never fills up), but simply to get the writer going
quickly again.
A bit less verbose commentary this time, if only because I assume that
the write side commentary isn't going to be ignored if you touch this
code.
Cc: David Howells <dhowells@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds [Sat, 7 Dec 2019 20:14:28 +0000 (12:14 -0800)]
pipe: fix and clarify pipe write wakeup logic
The pipe rework ends up having been extra painful, partly becaused of
actual bugs with ordering and caching of the pipe state, but also
because of subtle performance issues.
In particular, the pipe rework caused the kernel build to inexplicably
slow down.
The reason turns out to be that the GNU make jobserver (which limits the
parallelism of the build) uses a pipe to implement a "token" system: a
parallel submake will read a character from the pipe to get the job
token before starting a new job, and will write a character back to the
pipe when it is done. The overall job limit is thus easily controlled
by just writing the appropriate number of initial token characters into
the pipe.
But to work well, that really means that the old behavior of write
wakeups being synchronous (WF_SYNC) is very important - when the pipe
writer wakes up a reader, we want the reader to actually get scheduled
immediately. Otherwise you lose the parallelism of the build.
The pipe rework lost that synchronous wakeup on write, and we had
clearly all forgotten the reasons and rules for it.
This rewrites the pipe write wakeup logic to do the required Wsync
wakeups, but also clarifies the logic and avoids extraneous wakeups.
It also ends up addign a number of comments about what oit does and why,
so that we hopefully don't end up forgetting about this next time we
change this code.
Cc: David Howells <dhowells@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Heiner Kallweit [Fri, 6 Dec 2019 22:27:15 +0000 (23:27 +0100)]
r8169: add missing RX enabling for WoL on RTL8125
RTL8125 also requires to enable RX for WoL.
v2: add missing Fixes tag
Fixes: f1bce4ad2f1c ("r8169: add support for RTL8125") Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The commit ef87f7da6b28 ("net: phy: dp83867: move dt parsing to probe")
causes regression on TI dra71x-evm and dra72x-evm, where DP83867 PHY is
used in "rgmii-id" mode - the networking stops working.
Unfortunately, it's not enough to just move DT parsing code to .probe() as
it depends on phydev->interface value, which is set to correct value abter
the .probe() is completed and before calling .config_init(). So, RGMII
configuration can't be loaded from DT.
To fix and issue
- move RGMII validation code to .config_init()
- parse RGMII parameters in dp83867_of_init(), but consider them as
optional.
Fixes: ef87f7da6b28 ("net: phy: dp83867: move dt parsing to probe") Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Now RX interrupt is triggered twice every time, because in
cpsw_rx_interrupt() it is asked first and then disabled. So there will be
pending interrupt always, when RX interrupt is enabled again in NAPI
handler.
Fix it by first disabling IRQ and then do ask.
Fixes: 870915feabdc ("drivers: net: cpsw: remove disable_irq/enable_irq as irq can be masked from cpsw itself") Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Fri, 6 Dec 2019 04:43:46 +0000 (20:43 -0800)]
inet: protect against too small mtu values.
syzbot was once again able to crash a host by setting a very small mtu
on loopback device.
Let's make inetdev_valid_mtu() available in include/net/ip.h,
and use it in ip_setup_cork(), so that we protect both ip_append_page()
and __ip_append_data()
Also add a READ_ONCE() when the device mtu is read.
Pairs this lockless read with one WRITE_ONCE() in __dev_set_mtu(),
even if other code paths might write over this field.
Add a big comment in include/linux/netdevice.h about dev->mtu
needing READ_ONCE()/WRITE_ONCE() annotations.
Hopefully we will add the missing ones in followup patches.
Fixes: 1470ddf7f8ce ("inet: Remove explicit write references to sk/inet in ip_append_data") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Cong Wang [Fri, 6 Dec 2019 03:39:02 +0000 (19:39 -0800)]
gre: refetch erspan header from skb->data after pskb_may_pull()
After pskb_may_pull() we should always refetch the header
pointers from the skb->data in case it got reallocated.
In gre_parse_header(), the erspan header is still fetched
from the 'options' pointer which is fetched before
pskb_may_pull().
Found this during code review of a KMSAN bug report.
Fixes: cb73ee40b1b3 ("net: ip_gre: use erspan key field for tunnel lookup") Cc: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Acked-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> Acked-by: William Tu <u9012063@gmail.com> Reviewed-by: Simon Horman <simon.horman@netronome.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Aditya Pakki [Thu, 5 Dec 2019 23:04:49 +0000 (17:04 -0600)]
pppoe: remove redundant BUG_ON() check in pppoe_pernet
Passing NULL to pppoe_pernet causes a crash via BUG_ON.
Dereferencing net in net_generici() also has the same effect. This patch
removes the redundant BUG_ON check on the same parameter.
Signed-off-by: Aditya Pakki <pakki001@umn.edu> Signed-off-by: David S. Miller <davem@davemloft.net>