Re: [PATCH 1/3] misc: m48t59: replace qemu_system_reset_request() call with watchdog_perform_action()

[Peter Maydell <peter.maydell@linaro.org>]

On Fri, 16 Feb 2024 at 13:56, Abhiram Tilak <atp.exp@gmail.com> wrote:
>
> A few watchdog devices use qemu_system_reset_request(). This is not ideal since
> behaviour of watchdog-expiry can't be changed by QMP using `watchdog_action`.
> As stated in BiteSizedTasks wiki page, instead of using qemu_system_reset_request()
> to reset when a watchdog timer expires, let watchdog_perform_action() decide
> what to do.
>
> Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2124
> Signed-off-by: Abhiram Tilak <atp.exp@gmail.com>
> ---
>  hw/rtc/m48t59.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/hw/rtc/m48t59.c b/hw/rtc/m48t59.c
> index aa44c4b20c..ebda084478 100644
> --- a/hw/rtc/m48t59.c
> +++ b/hw/rtc/m48t59.c
> @@ -36,6 +36,7 @@
>  #include "qemu/bcd.h"
>  #include "qemu/module.h"
>  #include "trace.h"
> +#include "sysemu/watchdog.h"
>
>  #include "m48t59-internal.h"
>  #include "migration/vmstate.h"
> @@ -163,8 +164,7 @@ static void watchdog_cb (void *opaque)
>      if (NVRAM->buffer[0x1FF7] & 0x80) {
>          NVRAM->buffer[0x1FF7] = 0x00;
>          NVRAM->buffer[0x1FFC] &= ~0x40;
> -        /* May it be a hw CPU Reset instead ? */
> -        qemu_system_reset_request(SHUTDOWN_CAUSE_GUEST_RESET);
> +        watchdog_perform_action(); /* watchdog-expired action */
>      } else {

I don't think we need the comment, it's clear enough from the
name of the function.

Otherwise
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>

thanks
-- PMM