Hi Shawn, On Mittwoch, 14. Juli 2021 11:40:46 CET Shawn Guo wrote: > Hi Luca, > > On Sun, Jul 11, 2021 at 11:57:25AM +0200, Luca Weiss wrote: > > Hi Shawn, > > > > On Dienstag, 29. Juni 2021 05:05:09 CEST Shawn Guo wrote: > > > On some devices, e.g. Sony Xperia M4 Aqua, warm reset is used to reboot > > > device into bootloader and recovery mode. Instead of always doing hard > > > reset, add a check on reboot_mode for possible warm reset. > > > > > > Signed-off-by: Shawn Guo <shawn.guo@xxxxxxxxxx> > > > --- > > > > > > drivers/input/misc/pm8941-pwrkey.c | 6 +++++- > > > 1 file changed, 5 insertions(+), 1 deletion(-) > > > > > > diff --git a/drivers/input/misc/pm8941-pwrkey.c > > > b/drivers/input/misc/pm8941-pwrkey.c index cf8104454e74..9b14d6eb1918 > > > 100644 > > > --- a/drivers/input/misc/pm8941-pwrkey.c > > > +++ b/drivers/input/misc/pm8941-pwrkey.c > > > @@ -27,6 +27,7 @@ > > > > > > #define PON_PS_HOLD_RST_CTL2 0x5b > > > #define PON_PS_HOLD_ENABLE BIT(7) > > > #define PON_PS_HOLD_TYPE_MASK 0x0f > > > > > > +#define PON_PS_HOLD_TYPE_WARM_RESET 1 > > > > > > #define PON_PS_HOLD_TYPE_SHUTDOWN 4 > > > #define PON_PS_HOLD_TYPE_HARD_RESET 7 > > > > > > @@ -93,7 +94,10 @@ static int pm8941_reboot_notify(struct notifier_block > > > *nb, break; > > > > > > case SYS_RESTART: > > > > > > default: > > > - reset_type = PON_PS_HOLD_TYPE_HARD_RESET; > > > + if (reboot_mode == REBOOT_WARM) > > > > This doesn't compile with CONFIG_INPUT_PM8941_PWRKEY=m > > > > ERROR: modpost: "reboot_mode" [drivers/input/misc/pm8941-pwrkey.ko] > > undefined! > Thanks for the report! I will add a patch to export the symbol. > > > Also just to clarify, this is supposed to trigger when rebooting with > > LINUX_REBOOT_CMD_RESTART2 and adding an argument that way, right? > > With either of the following two methods, 'reboot_mode' will be turned > into REBOOT_WARM. > > - Boot kernel with 'reboot=warm' on cmdline > > - Set warm mode via sysfs entry > $ echo warm > /sys/kernel/reboot/mode +CC Sebastian Reichel Do you have an idea how this situation should look from user space? Just using LINUX_REBOOT_CMD_RESTART2 with the param won't work if the reboot mode also has to be set to work - but that's also not the case on all boards. Just LINUX_REBOOT_CMD_RESTART2 works fine on newer qcom SoCs that iirc store it in a pmic register that seems to survive a hard reboot. Would it be a good idea to introduce a new dt property for e.g. syscon-reboot-mode that makes it switch to warm reboot mode when a reboot mode is passed to the kernel? That way user space wouldn't need to care whether a particular board needs warm reboot or can just use the default hard reboot. Regards Luca > Shawn
