summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* dt-bindings: watchdog: renesas,wdt: Add support for RZ/G2LBiju Das2022-01-051-18/+57
| | | | | | | | | | | | Describe the WDT hardware in the RZ/G2L series. Signed-off-by: Biju Das <biju.das.jz@bp.renesas.com> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> Reviewed-by: Rob Herring <robh@kernel.org> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211130195357.18626-2-biju.das.jz@bp.renesas.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: da9063: Add hard dependency on I2CAndrej Picej2022-01-051-0/+1
| | | | | | | | | | | | | | Commit 5ea29919c294 ("watchdog: da9063: use atomic safe i2c transfer in reset handler") implements atomic save i2c transfer which uses i2c functions directly. Add I2C hard dependency which overrides COMPILE_TEST. Reported-by: kernel test robot <lkp@intel.com> Fixes: 968011a291f3 ("watchdog: da9063: use atomic safe i2c transfer in reset handler") Signed-off-by: Andrej Picej <andrej.picej@norik.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211129134938.3273289-1-andrej.picej@norik.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: Add Realtek Otto watchdog timerSander Vanheule2022-01-054-0/+405
| | | | | | | | | | | | | | | | | | | | | | | Realtek MIPS SoCs (platform name Otto) have a watchdog timer with pretimeout notifitication support. The WDT can (partially) hard reset, or soft reset the SoC. This driver implements all features as described in the devicetree binding, except the phase2 interrupt, and also functions as a restart handler. The cpu reset mode is considered to be a "warm" restart, since this mode does not reset all peripherals. Being an embedded system though, the "cpu" and "software" modes will still cause the bootloader to run on restart. It is not known how a forced system reset can be disabled on the supported platforms. This means that the phase2 interrupt will only fire at the same time as reset, so implementing phase2 is of little use. Signed-off-by: Sander Vanheule <sander@svanheule.net> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/6d060bccbdcc709cfa79203485db85aad3c3beb5.1637252610.git.sander@svanheule.net Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* dt-bindings: watchdog: Realtek Otto WDT bindingSander Vanheule2022-01-051-0/+91
| | | | | | | | | | | | | Add a binding description for Realtek's watchdog timer as found on several of their MIPS-based SoCs (codenamed Otto), such as the RTL838x, RTL839x, and RTL930x series of switch SoCs. Signed-off-by: Sander Vanheule <sander@svanheule.net> Reviewed-by: Rob Herring <robh@kernel.org> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/d832d5b02976dd2c2674d46778f61e5cfcd9b651.1637252610.git.sander@svanheule.net Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: s3c2410: Add Exynos850 supportSam Protsenko2021-12-281-1/+63
| | | | | | | | | | | | | | | | | | | | | | | | | Exynos850 is a bit different from SoCs already supported in WDT driver: - AUTOMATIC_WDT_RESET_DISABLE register is removed, so its value is always 0; .disable_auto_reset callback is not set for that reason - MASK_WDT_RESET_REQUEST register is replaced with CLUSTERx_NONCPU_IN_EN register; instead of masking (disabling) WDT reset interrupt it's now enabled with the same value; .mask_reset callback is reused for that functionality though - To make WDT functional, WDT counter needs to be enabled in CLUSTERx_NONCPU_OUT register; it's done using .enable_counter callback Also Exynos850 has two CPU clusters, each has its own dedicated WDT instance. Different PMU registers and bits are used for each cluster. So driver data is now modified in probe, adding needed info depending on cluster index passed from device tree. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211121165647.26706-13-semen.protsenko@linaro.org Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: da9063: use atomic safe i2c transfer in reset handlerYunus Bas2021-12-281-3/+9
| | | | | | | | | | | | | | This patch is based on commit 057b52b4b3d5 ("watchdog: da9062: make restart handler atomic safe"), which uses the atomic transfer capability of the i2c framework. Signed-off-by: Yunus Bas <y.bas@phytec.de> Signed-off-by: Andrej Picej <andrej.picej@norik.com> Reviewed-by: Adam Thomson <Adam.Thomson.Opensource@diasemi.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211124080654.2601135-1-andrej.picej@norik.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: davinci: Use div64_ul instead of do_divChangcheng Deng2021-12-281-1/+1
| | | | | | | | | | | | | do_div() does a 64-by-32 division. Here the divisor is an unsigned long which on some platforms is 64 bit wide. So use div64_ul instead of do_div to avoid a possible truncation. Reported-by: Zeal Robot <zealci@zte.com.cn> Signed-off-by: Changcheng Deng <deng.changcheng@zte.com.cn> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211125014924.46297-1-deng.changcheng@zte.com.cn Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: Remove BCM63XX_WDTFlorian Fainelli2021-12-283-328/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | Now that we can utilize the BCM7038_WDT driver, remove that one which was not converted to the watchdog APIs. There are a couple of notable differences with how the bcm7038_wdt driver proceeds: - bcm63xx_wdt would register with the ad-hoc BCM63xx hardware timer API, but this would only be used in order to catch the interrupt *before* a SoC reset and make the kernel "die" - bcm6xx_wdt would register a software timer and kick it every second in order to pet the watchdog, thus offering a two step watchdog process. This is not something that is brought over to the bcm7038_wdt as it is deemed unnecessary. If user-space cannot pet the watchdog, but a kernel timer can, the system is still in a bad shape anyway. bcm7038_wdt is simpler in its behavior and behaves as a standard watchdog driver and is not making use of any specific platform APIs, therefore making it more maintainable and extensible. Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211112224636.395101-8-f.fainelli@gmail.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* MIPS: BCM63XX: Provide platform data to watchdog deviceFlorian Fainelli2021-12-281-0/+8
| | | | | | | | | | | | | In order to utilize the bcm7038_wdt.c driver which needs to know the clock name to obtain, pass it via platform data using the bcm7038_wdt_platform_data structure. Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Acked-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211112224636.395101-7-f.fainelli@gmail.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: bcm7038_wdt: Add platform device id for bcm63xx-wdtFlorian Fainelli2021-12-281-0/+7
| | | | | | | | | | | In order to phase out bcm63xx_wdt and use bcm7038_wdt instead, introduce a platform_device_id table that allows both names to be matched. Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211112224636.395101-6-f.fainelli@gmail.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: Allow building BCM7038_WDT for BCM63XXFlorian Fainelli2021-12-281-3/+4
| | | | | | | | | | | | CONFIG_BCM63XX denotes the legacy MIPS-based DSL SoCs which utilize the same piece of hardware as a watchdog, make it possible to select that driver for those platforms. Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211112224636.395101-5-f.fainelli@gmail.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: bcm7038_wdt: Support platform data configurationFlorian Fainelli2021-12-282-1/+15
| | | | | | | | | | | | The BCM7038 watchdog driver needs to be able to obtain a specific clock name on BCM63xx platforms which is the "periph" clock ticking at 50MHz. make it possible to specify the clock name to obtain via platform data. Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211112224636.395101-4-f.fainelli@gmail.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* dt-bindings: watchdog: Add BCM6345 compatible to BCM7038 bindingFlorian Fainelli2021-12-281-2/+4
| | | | | | | | | | | | | The BCM7038 watchdog binding is updated to include a "brcm,bcm6345-wdt" compatible string which is the first instance of a DSL (BCM63xx) SoC seeing the integration of such a watchdog timer block. Reviewed-by: Rob Herring <robh@kernel.org> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211112224636.395101-3-f.fainelli@gmail.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* dt-bindings: watchdog: convert Broadcom's WDT to the json-schemaRafał Miłecki2021-12-282-19/+41
| | | | | | | | | | | This helps validating DTS files. Signed-off-by: Rafał Miłecki <rafal@milecki.pl> Acked-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Rob Herring <robh@kernel.org> Link: https://lore.kernel.org/r/20211115055354.6089-1-zajec5@gmail.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: meson_gxbb_wdt: remove stop_on_rebootArtem Lapkin2021-12-281-1/+0
| | | | | | | | | | | | | | | | | | | | Remove watchdog_stop_on_reboot() The Meson platform still has some hardware drivers problems for some configurations which can freeze devices on shutdown/reboot. Remove watchdog_stop_on_reboot() to catch this situation and ensure that the reboot happens anyway. Users who still want to stop the watchdog on reboot can still do so using the watchdog.stop_on_reboot=1 module parameter. https://lore.kernel.org/linux-watchdog/20210729072308.1908904-1-art@khadas.com/T/#t Signed-off-by: Artem Lapkin <art@khadas.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211110022518.1676834-1-art@khadas.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: Kconfig: fix help text indentationLuca Ceresoli2021-12-281-24/+24
| | | | | | | | | | | | Some entries indent their help text with 1 tab + 1 space or 1 tab only instead of 1 tab + 2 spaces. Add the missing spaces. Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net> Acked-by: Randy Dunlap <rdunlap@infradead.org> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211111225852.3128201-7-luca@lucaceresoli.net Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* dt-bindings: watchdog: imx7ulp-wdt: Add imx8ulp compatible stringJacky Bai2021-12-281-2/+5
| | | | | | | | | | | | | | The wdog on i.MX8ULP is derived from i.MX7ULP, it uses two compatible strings, so update the compatible string for i.MX8ULP. Reviewed-by: Dong Aisheng <aisheng.dong@nxp.com> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: Jacky Bai <ping.bai@nxp.com> Signed-off-by: Peng Fan <peng.fan@nxp.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211112082930.3809351-7-peng.fan@oss.nxp.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: s3c2410: Remove superfluous err labelSam Protsenko2021-12-281-9/+4
| | | | | | | | | | | | | | | | 'err' label in probe function is not really need, it just returns. Remove it and replace all 'goto' statements with actual returns in place. No functional change here, just a cleanup patch. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211107202943.8859-12-semen.protsenko@linaro.org [groeck: Fixed context conflicts] Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: s3c2410: Support separate source clockSam Protsenko2021-12-281-15/+41
| | | | | | | | | | | | | | | | | | | | | Right now all devices supported in the driver have the single clock: it acts simultaneously as a bus clock (providing register interface clocking) and source clock (driving watchdog counter). Some newer Exynos chips, like Exynos850, have two separate clocks for that. In that case two clocks will be passed to the driver from the resource provider, e.g. Device Tree. Provide necessary infrastructure to support that case: - use source clock's rate for all timer related calculations - use bus clock to gate/ungate the register interface All devices that use the single clock are kept intact: if only one clock is passed from Device Tree, it will be used for both purposes as before. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211107202943.8859-11-semen.protsenko@linaro.org Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: s3c2410: Cleanup PMU related codeSam Protsenko2021-12-281-25/+58
| | | | | | | | | | | | | | | | Now that PMU enablement code was extended for new Exynos SoCs, it doesn't look very cohesive and consistent anymore. Do a bit of renaming, grouping and style changes, to make it look good again. While at it, add quirks documentation as well. No functional change, just a refactoring commit. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211123232613.22438-1-semen.protsenko@linaro.org Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: s3c2410: Add support for WDT counter enable registerSam Protsenko2021-12-281-1/+27
| | | | | | | | | | | | | | | | | On new Exynos chips (e.g. Exynos850) new CLUSTERx_NONCPU_OUT register is introduced, where CNT_EN_WDT bit must be enabled to make watchdog counter running. Add corresponding quirk and proper infrastructure to handle that register if the quirk is set. This commit doesn't bring any functional change to existing devices, but merely provides an infrastructure for upcoming chips support. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211107202943.8859-9-semen.protsenko@linaro.org Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: s3c2410: Implement a way to invert mask reg valueSam Protsenko2021-12-281-1/+4
| | | | | | | | | | | | | | | | | | | | | | On new Exynos chips (like Exynos850) the MASK_WDT_RESET_REQUEST register is replaced with CLUSTERx_NONCPU_INT_EN, and its mask bit value meaning was reversed: for new register the bit value "1" means "Interrupt enabled", while for MASK_WDT_RESET_REQUEST register "1" means "Mask the interrupt" (i.e. "Interrupt disabled"). Introduce "mask_reset_inv" boolean field in driver data structure; when that field is "true", mask register handling function will invert the value before setting it to the register. This commit doesn't bring any functional change to existing devices, but merely provides an infrastructure for upcoming chips support. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211107202943.8859-8-semen.protsenko@linaro.org Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: s3c2410: Extract disable and mask code into separate functionsSam Protsenko2021-12-281-19/+35
| | | | | | | | | | | | | | | | | The s3c2410wdt_mask_and_disable_reset() function content is bound to be changed further. Prepare it for upcoming changes by splitting into separate "mask reset" and "disable reset" functions. But keep s3c2410wdt_mask_and_disable_reset() function present as a facade. This commit doesn't bring any functional change to existing devices, but merely provides an infrastructure for upcoming chips support. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211107202943.8859-7-semen.protsenko@linaro.org Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: s3c2410: Make reset disable register optionalSam Protsenko2021-12-281-9/+13
| | | | | | | | | | | | | | | | | On new Exynos chips (e.g. Exynos850 and Exynos9) the AUTOMATIC_WDT_RESET_DISABLE register was removed, and its value can be thought of as "always 0x0". Add correspondig quirk bit, so that the driver can omit accessing it if it's not present. This commit doesn't bring any functional change to existing devices, but merely provides an infrastructure for upcoming chips support. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211107202943.8859-6-semen.protsenko@linaro.org Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: s3c2410: Let kernel kick watchdogSam Protsenko2021-12-281-11/+15
| | | | | | | | | | | | | | | | | | | | When "tmr_atboot" module param is set, the watchdog is started in driver's probe. In that case, also set WDOG_HW_RUNNING bit to let watchdog core driver know it's running. This way watchdog core can kick the watchdog for us (if CONFIG_WATCHDOG_HANDLE_BOOT_ENABLED option is enabled), until user space takes control. WDOG_HW_RUNNING bit must be set before registering the watchdog. So the "tmr_atboot" handling code is moved before watchdog registration, to avoid performing the same check twice. This is also logical because WDOG_HW_RUNNING bit makes WDT core expect actually running watchdog. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211107202943.8859-5-semen.protsenko@linaro.org Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: s3c2410: Fail probe if can't find valid timeoutSam Protsenko2021-12-281-11/+10
| | | | | | | | | | | | | | | | | Driver can't work properly if there no valid timeout was found in s3c2410wdt_set_heartbeat(). Ideally, that function should be reworked in a way that it's always able to find some valid timeout. As a temporary solution let's for now just fail the driver probe in case the valid timeout can't be found in s3c2410wdt_set_heartbeat() function. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reported-by: Guenter Roeck <linux@roeck-us.net> Suggested-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211107202943.8859-4-semen.protsenko@linaro.org Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* dt-bindings: watchdog: Document Exynos850 watchdog bindingsSam Protsenko2021-12-281-4/+41
| | | | | | | | | | | | | | | | | | | | | | | Exynos850 SoC has two CPU clusters: - cluster 0: contains CPUs #0, #1, #2, #3 - cluster 1: contains CPUs #4, #5, #6, #7 Each cluster has its own dedicated watchdog timer. Those WDT instances are controlled using different bits in PMU registers, new "samsung,index" property is added to tell the driver which bits to use for defined watchdog node. Also on Exynos850 the peripheral clock and the source clock are two different clocks. Provide a way to specify two clocks in watchdog device tree node. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Rob Herring <robh@kernel.org> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211107202943.8859-3-semen.protsenko@linaro.org Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* dt-bindings: watchdog: Require samsung,syscon-phandle for Exynos7Sam Protsenko2021-12-281-2/+3
| | | | | | | | | | | | | | Exynos7 watchdog driver is clearly indicating that its dts node must define syscon phandle property. That was probably forgotten, so add it. Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Fixes: 2b9366b66967 ("watchdog: s3c2410_wdt: Add support for Watchdog device on Exynos7") Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com> Reviewed-by: Rob Herring <robh@kernel.org> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211107202943.8859-2-semen.protsenko@linaro.org Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: f71808e_wdt: Add F81966 supportAaeonIot2021-12-281-2/+8
| | | | | | | | | | | | This adds watchdog support the Fintek F81966 Super I/O chip. Testing was done on the Aaeon SSE-OPTI Signed-off-by: AaeonIot <sophiehu@aaeon.com.tw> Signed-off-by: Chia-Lin Kao (AceLan) <acelan.kao@canonical.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211117024052.2427539-1-acelan.kao@canonical.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* watchdog: Kconfig: enable MTK watchdogKevin Hilman2021-12-281-0/+1
| | | | | | | | | | | | | | Enable CONFIG_MEDIATEK_WATCHDOG when ARCH_MEDIATEK is enabled. On some platforms (e.g. mt8183-pumpkin), watchdog is enabled by bootloader, so kernel driver needs to be enabled to avoid watchdog firing and causing reboot part way through kernel boot. Signed-off-by: Kevin Hilman <khilman@baylibre.com> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20211103230354.915658-1-khilman@baylibre.com Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
* Linux 5.16-rc7v5.16-rc7Linus Torvalds2021-12-261-1/+1
|
* Merge tag 'x86_urgent_for_v5.16_rc7' of ↵Linus Torvalds2021-12-262-44/+32
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull x86 fixes from Borislav Petkov: - Prevent potential undefined behavior due to shifting pkey constants into the sign bit - Move the EFI memory reservation code *after* the efi= cmdline parsing has happened - Revert two commits which turned out to be the wrong direction to chase when accommodating early memblock reservations consolidation and command line parameters parsing * tag 'x86_urgent_for_v5.16_rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86/pkey: Fix undefined behaviour with PKRU_WD_BIT x86/boot: Move EFI range reservation after cmdline parsing Revert "x86/boot: Pull up cmdline preparation and early param parsing" Revert "x86/boot: Mark prepare_command_line() __init"
| * x86/pkey: Fix undefined behaviour with PKRU_WD_BITAndrew Cooper2021-12-191-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Both __pkru_allows_write() and arch_set_user_pkey_access() shift PKRU_WD_BIT (a signed constant) by up to 30 bits, hitting the sign bit. Use unsigned constants instead. Clearly pkey 15 has not been used in combination with UBSAN yet. Noticed by code inspection only. I can't actually provoke the compiler into generating incorrect logic as far as this shift is concerned. [ dhansen: add stable@ tag, plus minor changelog massaging, For anyone doing backports, these #defines were in arch/x86/include/asm/pgtable.h before 784a46618f6. ] Fixes: 33a709b25a76 ("mm/gup, x86/mm/pkeys: Check VMAs and PTEs for protection keys") Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com> Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Signed-off-by: Borislav Petkov <bp@suse.de> Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/20211216000856.4480-1-andrew.cooper3@citrix.com
| * x86/boot: Move EFI range reservation after cmdline parsingMike Rapoport2021-12-151-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The memory reservation in arch/x86/platform/efi/efi.c depends on at least two command line parameters. Put it back later in the boot process and move efi_memblock_x86_reserve_range() out of early_memory_reserve(). An attempt to fix this was done in 8d48bf8206f7 ("x86/boot: Pull up cmdline preparation and early param parsing") but that caused other troubles so it got reverted. The bug this is addressing is: Dan reports that Anjaneya Chagam can no longer use the efi=nosoftreserve kernel command line parameter to suppress "soft reservation" behavior. This is due to the fact that the following call-chain happens at boot: early_reserve_memory |-> efi_memblock_x86_reserve_range |-> efi_fake_memmap_early which does if (!efi_soft_reserve_enabled()) return; and that would have set EFI_MEM_NO_SOFT_RESERVE after having parsed "nosoftreserve". However, parse_early_param() gets called *after* it, leading to the boot cmdline not being taken into account. See also https://lore.kernel.org/r/e8dd8993c38702ee6dd73b3c11f158617e665607.camel@intel.com [ bp: Turn into a proper patch. ] Signed-off-by: Mike Rapoport <rppt@kernel.org> Signed-off-by: Borislav Petkov <bp@suse.de> Link: https://lore.kernel.org/r/20211213112757.2612-4-bp@alien8.de
| * Revert "x86/boot: Pull up cmdline preparation and early param parsing"Borislav Petkov2021-12-151-39/+27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This reverts commit 8d48bf8206f77aa8687f0e241e901e5197e52423. It turned out to be a bad idea as it broke supplying mem= cmdline parameters due to parse_memopt() requiring preparatory work like setting up the e820 table in e820__memory_setup() in order to be able to exclude the range specified by mem=. Pulling that up would've broken Xen PV again, see threads at https://lkml.kernel.org/r/20210920120421.29276-1-jgross@suse.com due to xen_memory_setup() needing the first reservations in early_reserve_memory() - kernel and initrd - to have happened already. This could be fixed again by having Xen do those reservations itself... Long story short, revert this and do a simpler fix in a later patch. Signed-off-by: Borislav Petkov <bp@suse.de> Link: https://lore.kernel.org/r/20211213112757.2612-3-bp@alien8.de
| * Revert "x86/boot: Mark prepare_command_line() __init"Borislav Petkov2021-12-151-1/+1
| | | | | | | | | | | | | | This reverts commit c0f2077baa4113f38f008b8e912b9fb3ff8d43df. Signed-off-by: Borislav Petkov <bp@suse.de> Link: https://lore.kernel.org/r/20211213112757.2612-2-bp@alien8.de
* | Merge tag 'objtool_urgent_for_v5.16_rc7' of ↵Linus Torvalds2021-12-264-4/+6
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull objtool fixes from Borislav Petkov: - Prevent clang from reordering the reachable annotation in an inline asm statement without inputs - Fix objtool builds on non-glibc systems due to undefined __always_inline * tag 'objtool_urgent_for_v5.16_rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: compiler.h: Fix annotation macro misplacement with Clang uapi: Fix undefined __always_inline on non-glibc systems
| * | compiler.h: Fix annotation macro misplacement with ClangJosh Poimboeuf2021-12-212-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When building with Clang and CONFIG_TRACE_BRANCH_PROFILING, there are a lot of unreachable warnings, like: arch/x86/kernel/traps.o: warning: objtool: handle_xfd_event()+0x134: unreachable instruction Without an input to the inline asm, 'volatile' is ignored for some reason and Clang feels free to move the reachable() annotation away from its intended location. Fix that by re-adding the counter value to the inputs. Fixes: f1069a8756b9 ("compiler.h: Avoid using inline asm operand modifiers") Fixes: c199f64ff93c ("instrumentation.h: Avoid using inline asm operand modifiers") Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lore.kernel.org/r/0417e96909b97a406323409210de7bf13df0b170.1636410380.git.jpoimboe@redhat.com Cc: Peter Zijlstra <peterz@infradead.org> Cc: x86@kernel.org Cc: Vasily Gorbik <gor@linux.ibm.com> Cc: Miroslav Benes <mbenes@suse.cz>
| * | uapi: Fix undefined __always_inline on non-glibc systemsIsmael Luceno2021-12-212-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This macro is defined by glibc itself, which makes the issue go unnoticed on those systems. On non-glibc systems it causes build failures on several utilities and libraries, like bpftool and objtool. Fixes: 1d509f2a6ebc ("x86/insn: Support big endian cross-compiles") Fixes: 2d7ce0e8a704 ("tools/virtio: more stubs") Fixes: 3fb321fde22d ("selftests/net: ipv6 flowlabel") Fixes: 50b3ed57dee9 ("selftests/bpf: test bpf flow dissection") Fixes: 9cacf81f8161 ("bpf: Remove extra lock_sock for TCP_ZEROCOPY_RECEIVE") Fixes: a4b2061242ec ("tools include uapi: Grab a copy of linux/in.h") Fixes: b12d6ec09730 ("bpf: btf: add btf print functionality") Fixes: c0dd967818a2 ("tools, include: Grab a copy of linux/erspan.h") Fixes: c4b6014e8bb0 ("tools: Add copy of perf_event.h to tools/include/linux/") Signed-off-by: Ismael Luceno <ismael@iodev.co.uk> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lore.kernel.org/r/20211115134647.1921-1-ismael@iodev.co.uk Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Vasily Gorbik <gor@linux.ibm.com>
* | | Merge tag 'pinctrl-v5.16-3' of ↵Linus Torvalds2021-12-253-19/+26
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl Pull pin control fixes from Linus Walleij: "Some hopefully final pin control fixes for the v5.16 kernel: - Fix an out-of-bounds bug in the Mediatek driver - Fix an init order bug in the Broadcom BCM2835 driver - Fix a GPIO offset bug in the STM32 driver" * tag 'pinctrl-v5.16-3' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl: pinctrl: stm32: consider the GPIO offset to expose all the GPIO lines pinctrl: bcm2835: Change init order for gpio hogs pinctrl: mediatek: fix global-out-of-bounds issue
| * | | pinctrl: stm32: consider the GPIO offset to expose all the GPIO linesFabien Dessenne2021-12-161-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Consider the GPIO controller offset (from "gpio-ranges") to compute the maximum GPIO line number. This fixes an issue where gpio-ranges uses a non-null offset. e.g.: gpio-ranges = <&pinctrl 6 86 10> In that case the last valid GPIO line is not 9 but 15 (6 + 10 - 1) Cc: stable@vger.kernel.org Fixes: 67e2996f72c7 ("pinctrl: stm32: fix the reported number of GPIO lines per bank") Reported-by: Christoph Fritz <chf.fritz@googlemail.com> Signed-off-by: Fabien Dessenne <fabien.dessenne@foss.st.com> Link: https://lore.kernel.org/r/20211215095808.621716-1-fabien.dessenne@foss.st.com Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
| * | | pinctrl: bcm2835: Change init order for gpio hogsPhil Elwell2021-12-101-13/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ...and gpio-ranges pinctrl-bcm2835 is a combined pinctrl/gpio driver. Currently the gpio side is registered first, but this breaks gpio hogs (which are configured during gpiochip_add_data). Part of the hog initialisation is a call to pinctrl_gpio_request, and since the pinctrl driver hasn't yet been registered this results in an -EPROBE_DEFER from which it can never recover. Change the initialisation sequence to register the pinctrl driver first. This also solves a similar problem with the gpio-ranges property, which is required in order for released pins to be returned to inputs. Fixes: 73345a18d464b ("pinctrl: bcm2835: Pass irqchip when adding gpiochip") Signed-off-by: Phil Elwell <phil@raspberrypi.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Link: https://lore.kernel.org/r/20211206092237.4105895-2-phil@raspberrypi.com Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
| * | | pinctrl: mediatek: fix global-out-of-bounds issueGuodong Liu2021-11-241-2/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When eint virtual eint number is greater than gpio number, it maybe produce 'desc[eint_n]' size globle-out-of-bounds issue. Signed-off-by: Guodong Liu <guodong.liu@mediatek.corp-partner.google.com> Signed-off-by: Zhiyong Tao <zhiyong.tao@mediatek.com> Reviewed-by: Chen-Yu Tsai <wenst@chromium.org> Link: https://lore.kernel.org/r/20211110071900.4490-2-zhiyong.tao@mediatek.com Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
* | | | Merge tag 'hwmon-for-v5.16-rc7' of ↵Linus Torvalds2021-12-251-44/+62
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging Pull hwmon fixes from Guenter Roeck: "A couple of lm90 driver fixes. None of them are critical, but they should nevertheless be fixed" * tag 'hwmon-for-v5.16-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging: hwmon: (lm90) Do not report 'busy' status bit as alarm hwmom: (lm90) Fix citical alarm status for MAX6680/MAX6681 hwmon: (lm90) Drop critical attribute support for MAX6654 hwmon: (lm90) Prevent integer overflow/underflow in hysteresis calculations hwmon: (lm90) Fix usage of CONFIG2 register in detect function
| * | | | hwmon: (lm90) Do not report 'busy' status bit as alarmGuenter Roeck2021-12-121-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Bit 7 of the status register indicates that the chip is busy doing a conversion. It does not indicate an alarm status. Stop reporting it as alarm status bit. Signed-off-by: Guenter Roeck <linux@roeck-us.net>
| * | | | hwmom: (lm90) Fix citical alarm status for MAX6680/MAX6681Guenter Roeck2021-12-121-2/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Tests with a real chip and a closer look into the datasheet reveals that the local and remote critical alarm status bits are swapped for MAX6680/MAX6681. Signed-off-by: Guenter Roeck <linux@roeck-us.net>
| * | | | hwmon: (lm90) Drop critical attribute support for MAX6654Guenter Roeck2021-12-121-37/+49
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Tests with a real chip and a closer look into the datasheet show that MAX6654 does not support CRIT/THERM/OVERTEMP limits, so drop support of the respective attributes for this chip. Introduce LM90_HAVE_CRIT flag and use it to instantiate critical limit attributes to solve the problem. Cc: Josh Lehan <krellan@google.com> Fixes: 229d495d8189 ("hwmon: (lm90) Add max6654 support to lm90 driver") Signed-off-by: Guenter Roeck <linux@roeck-us.net>
| * | | | hwmon: (lm90) Prevent integer overflow/underflow in hysteresis calculationsGuenter Roeck2021-12-121-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Commit b50aa49638c7 ("hwmon: (lm90) Prevent integer underflows of temperature calculations") addressed a number of underflow situations when writing temperature limits. However, it missed one situation, seen when an attempt is made to set the hysteresis value to MAX_LONG and the critical temperature limit is negative. Use clamp_val() when setting the hysteresis temperature to ensure that the provided value can never overflow or underflow. Fixes: b50aa49638c7 ("hwmon: (lm90) Prevent integer underflows of temperature calculations") Cc: Dmitry Osipenko <digetx@gmail.com> Reviewed-by: Dmitry Osipenko <digetx@gmail.com> Signed-off-by: Guenter Roeck <linux@roeck-us.net>
| * | | | hwmon: (lm90) Fix usage of CONFIG2 register in detect functionGuenter Roeck2021-12-121-3/+2
| | |_|/ | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The detect function had a comment "Make compiler happy" when id did not read the second configuration register. As it turns out, the code was checking the contents of this register for manufacturer ID 0xA1 (NXP Semiconductor/Philips), but never actually read the register. So it wasn't surprising that the compiler complained, and it indeed had a point. Fix the code to read the register contents for manufacturer ID 0xa1. At the same time, the code was reading the register for manufacturer ID 0x41 (Analog Devices), but it was not using the results. In effect it was just checking if reading the register returned an error. That doesn't really add much if any value, so stop doing that. Fixes: f90be42fb383 ("hwmon: (lm90) Refactor reading of config2 register") Signed-off-by: Guenter Roeck <linux@roeck-us.net>
* | | | Merge branch 'for-linus' of ↵Linus Torvalds2021-12-2510-37/+151
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input Pull input fixes from Dmitry Torokhov: "A few small updates to drivers. Of note we are now deferring probes of i8042 on some Asus devices as the controller is not ready to respond to queries first time around when the driver is compiled into the kernel" * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input: Input: elants_i2c - do not check Remark ID on eKTH3900/eKTH5312 Input: atmel_mxt_ts - fix double free in mxt_read_info_block Input: goodix - fix memory leak in goodix_firmware_upload Input: goodix - add id->model mapping for the "9111" model Input: goodix - try not to touch the reset-pin on x86/ACPI devices Input: i8042 - enable deferred probe quirk for ASUS UM325UA Input: elantech - fix stack out of bound access in elantech_change_report_id() Input: iqs626a - prohibit inlining of channel parsing functions Input: i8042 - add deferred probe support
generated by cgit v1.2.3 (git 2.25.1) at 2024-07-14 06:02:04 +0000