Pull watchdog updates from Wim Van Sebroeck:
- add marvell GTI watchdog driver
- add support for Amlogic-T7 SoCs
- document the IPQ5018 watchdog compatible
- enable COMPILE_TEST for more watchdog device drivers
- core: stop watchdog when executing poweroff command
- other small improvements and fixes
* tag 'linux-watchdog-6.6-rc1' of git://www.linux-watchdog.org/linux-watchdog: (21 commits)
watchdog: Add support for Amlogic-T7 SoCs
watchdog: Add a new struct for Amlogic-GXBB driver
dt-bindings: watchdog: Add support for Amlogic-T7 SoCs
dt-bindings: watchdog: qcom-wdt: document IPQ5018
watchdog: imx2_wdt: Improve dev_crit() message
watchdog: stm32: Drop unnecessary of_match_ptr()
watchdog: sama5d4: readout initial state
watchdog: intel-mid_wdt: add MODULE_ALIAS() to allow auto-load
watchdog: core: stop watchdog when executing poweroff command
watchdog: pm8916_wdt: Remove redundant of_match_ptr()
watchdog: xilinx_wwdt: Use div_u64() in xilinx_wwdt_start()
watchdog: starfive: Remove #ifdef guards for PM related functions
watchdog: s3c2410: Fix potential deadlock on &wdt->lock
watchdog:rit_wdt: Add support for WDIOF_CARDRESET
dt-bindings: watchdog: ti,rti-wdt: Add support for WDIOF_CARDRESET
watchdog: Enable COMPILE_TEST for more drivers
watchdog: advantech_ec_wdt: fix Kconfig dependencies
watchdog: Explicitly include correct DT includes
Watchdog: Add marvell GTI watchdog driver
dt-bindings: watchdog: marvell GTI system watchdog driver
...
After issuing a "poweroff" command the board goes through the
expected power-off sequence and turns it off completely:
systemd-shutdown[1]: Powering off.
imx2-wdt 30280000.watchdog: Device shutdown: Expect reboot!
reboot: Power down
The "Expect reboot!" message is misleading because in the
power-off case, no reboot is expected to happen at all.
Avoid the confusion by removing the "Expect reboot!" message.
Signed-off-by: Fabio Estevam <festevam@denx.de>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230822135255.1013981-1-festevam@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
On some platforms, for example, Intel SoC FPGA platform, the watchdog
is started in u-boot, and so kernel will create a timer and work to
keep watchdog alive. But when user executes commands "poweroff" or
"shutdown -h", the system is reset when watchdog triggers timeout even
if WDOG_STOP_ON_REBOOT is set in watchdog driver. The root cause is
that the watchdog is not stopped properly because the SYS_POWER_OFF case
is not supported in notify callback function. So, add this case to stop
watchdog when executing poweroff command.
Signed-off-by: Meng Li <Meng.Li@windriver.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230814090616.1853263-1-Meng.Li@windriver.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
After commit f1a43aadb5 ("watchdog: Enable COMPILE_TEST for more
drivers"), it is possible to enable this driver on 32-bit architectures.
When building for those architectures with clang, there is an error due
to a 64-bit division in xilinx_wwdt_start():
ERROR: modpost: "__aeabi_uldivmod" [drivers/watchdog/xilinx_wwdt.ko] undefined!
Use div_u64() to fix this, which takes a 64-bit dividend and 32-bit
divisor. GCC likely avoids the same error due to optimizations it
employs to transform division by a constant into other equivalent
operations, which may be different than what is implemented in clang.
Link: https://github.com/ClangBuiltLinux/linux/issues/1915
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230815-watchdog-xilinx-div_u64-v1-1-20b0b5a65c2e@kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
Use the new PM macros for the suspend and resume functions to be
automatically dropped by the compiler when CONFIG_PM or
CONFIG_PM_SLEEP are disabled, without having to use #ifdef guards.
This has the advantage of always compiling these functions in,
independently of any Kconfig option. Thanks to that, bugs and other
regressions are subsequently easier to catch.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230706082928.10869-1-paul@crapouillou.net
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
As &wdt->lock is acquired by hard irq s3c2410wdt_irq(),
other acquisition of the same lock under process context should
disable irq, otherwise deadlock could happen if the
irq preempt the execution while the lock is held in process context
on the same CPU.
[Deadlock Scenario]
s3c2410wdt_suspend()
-> s3c2410wdt_stop()
-> spin_lock(&wdt->lock)
<irq iterrupt>
-> s3c2410wdt_irq()
-> s3c2410wdt_keepalive()
-> spin_lock(&wdt->lock) (deadlock here)
[Deadlock Scenario]
s3c2410wdt_probe()
-> s3c2410wdt_start()
-> spin_lock(&wdt->lock)
<irq iterrupt>
-> s3c2410wdt_irq()
-> s3c2410wdt_keepalive()
-> spin_lock(&wdt->lock) (deadlock here)
[Deadlock Scenario]
s3c2410wdt_keepalive()
-> spin_lock(&wdt->lock)
<irq iterrupt>
-> s3c2410wdt_irq()
-> s3c2410wdt_keepalive()
-> spin_lock(&wdt->lock) (deadlock here)
This flaw was found by an experimental static analysis tool I am
developing for irq-related deadlock, which reported the above
warning when analyzing the linux kernel 6.4-rc7 release.
The tentative patch fix the potential deadlock by spin_lock_irqsave()
under process context.
Signed-off-by: Chengfeng Ye <dg573847474@gmail.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230705090951.63762-1-dg573847474@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
This patch adds the WDIOF_CARDRESET support for the platform watchdog
whose hardware does not support this feature, to know if the board
reboot is due to a watchdog reset.
This is done via reserved memory(RAM), which indicates if specific
info saved, triggering the watchdog reset in last boot.
Signed-off-by: Li Hua Qian <huaqian.li@siemens.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230718021007.1338761-4-huaqian.li@siemens.com
[groeck: vaddr == NULL --> !vaddr]
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
The simatic-ipc driver no longer depends on PCI, but its watchdog portion
still needs it, otherwise P2SB runs into a build failure:
WARNING: unmet direct dependencies detected for P2SB
Depends on [n]: PCI [=n] && X86 [=y]
Selected by [m]:
- SIEMENS_SIMATIC_IPC_WDT [=m] && WATCHDOG [=y] && SIEMENS_SIMATIC_IPC [=y]
drivers/platform/x86/p2sb.c:121:3: error: call to undeclared function 'pci_bus_write_config_dword'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
pci_bus_write_config_dword(bus, devfn_p2sb, P2SBC, 0);
Add back the minimum dependendency to make it build in random
configurations again.
Fixes: b72da71ce2 ("platform/x86: simatic-ipc: drop PCI runtime depends and header")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230814073924.1066390-1-arnd@kernel.org
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
The DT of_device.h and of_platform.h date back to the separate
of_platform_bus_type before it was merged into the regular platform bus.
As part of that merge prepping Arm DT support 13 years ago, they
"temporarily" include each other. They also include platform_device.h
and of.h. As a result, there's a pretty much random mix of those include
files used throughout the tree. In order to detangle these headers and
replace the implicit includes with struct declarations, users need to
explicitly include the correct includes.
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Acked-by: Alim Akhtar <alim.akhtar@samsung.com>
Signed-off-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20230726233302.3812749-1-robh@kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
This patch add support for Marvell GTI watchdog. Global timer
unit (GTI) support hardware watchdog timer. Software programs
watchdog timer to generate interrupt on first timeout, second
timeout is configured to be ignored and system reboots on
third timeout.
Signed-off-by: Bharat Bhushan <bbhushan2@marvell.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230803032523.6242-2-bbhushan2@marvell.com
[groeck: MODULE_LICENSE GPL v2 -> GPL since that is sufficient]
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
Versal watchdog driver uses window watchdog mode. Window watchdog
timer(WWDT) contains closed(first) and open(second) window with
32 bit width. Write to the watchdog timer within predefined window
periods of time. This means a period that is not too soon and a
period that is not too late. The WWDT has to be restarted within
the open window time. If software tries to restart WWDT outside of
the open window time period, it generates a reset.
Signed-off-by: Srinivas Neeli <srinivas.neeli@amd.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lkml.kernel.org/r/20230420104231.2243079-4-srinivas.neeli@amd.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
When the starfive watchdog driver uses 'pm_runtime_put_sync()' as probe
return value at last and 'early_enable' is disabled, it could return the
error '-ENOSYS' if the CONFIG_PM is disabled, but the driver should works
normally.
Drop the 'return' and keep the 'pm_runtime_put_sync()', but do not use it
as the return value.
Fixes: db728ea9c7 ("drivers: watchdog: Add StarFive Watchdog driver")
Signed-off-by: Xingyu Wu <xingyu.wu@starfivetech.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230426031216.37981-1-xingyu.wu@starfivetech.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
When setting the timeout for the menz069_wdt watchdog driver, we
erroneously read from the 'watchdog value register' (WVR) instead of the
'watchdog timer register' (WTR) and then write the value back into WTR.
This can potentially lead to wrong timeouts and wrong enable bit settings.
Signed-off-by: Johannes Thumshirn <jth@kernel.org>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230418172531.177349-3-jth@kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
Doing a 'cat /dev/watchdog0' with menz069_wdt as watchdog0 will result in
a NULL pointer dereference.
This happens because we're passing the wrong pointer to
watchdog_register_device(). Fix this by getting rid of the static
watchdog_device structure and use the one embedded into the driver's
per-instance private data.
Signed-off-by: Johannes Thumshirn <jth@kernel.org>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230418172531.177349-2-jth@kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
The watchdog countdown is supposed to begin when the device file is
opened. Instead, it would begin countdown upon the first write to or
close of the device file. Now, the ping operation is called within the
start operation which ensures the countdown begins. From experimenation,
it does not appear possible to do this with a single write including
both the start bit and the trigger bit. So, it is done as two distinct
writes.
Signed-off-by: Gregory Oakes <gregory.oakes@amd.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230316201312.17538-1-gregory.oakes@amd.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
Make use of dev_err_probe() also for error paths that don't have to
handle -EPROBE_DEFER. While the code handing -EPROBE_DEFER isn't used
for these error paths, it still simpler as it cares for pretty printing
the error code and usually needs one code line less as it combines
message emitting and error returning. This also unifies the format of
the error messages.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230307065603.2253054-3-u.kleine-koenig@pengutronix.de
[groeck: Split long line to avoid checkpatch warning]
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
This is a preparation for making more use of dev_err_probe(). The idea
is that s3c2410_get_wdt_drv_data() (as it's called only by .probe()) can
make effective use of dev_err_probe() only if it returns an int. For
that the assignment to wdt->drv_data has to happen in the function. The
caller can then just pass on the return value in the error case.
This seems to be nicer for the compiler: bloatometer reports for an
ARCH=arm s3c6400_defconfig build:
add/remove: 1/1 grow/shrink: 0/1 up/down: 4/-64 (-60)
Function old new delta
__initcall__kmod_s3c2410_wdt__209_821_s3c2410wdt_driver_init6 - 4 +4
__initcall__kmod_s3c2410_wdt__209_819_s3c2410wdt_driver_init6 4 - -4
s3c2410wdt_probe 1332 1272 -60
There is no semantical change. (Just one minor difference: Before this
patch wdt->drv_data was always assigned, now that only happens in the
non-error case. That doesn't matter however as *wdt is freed in the
error case.)
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230307065603.2253054-2-u.kleine-koenig@pengutronix.de
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
