[U-Boot] [PATCH 2/2] arm: mvebu: x530: Enable watchdog in SPL and U-Boot

[Chris Packham]

On Fri, 15 Feb 2019, 9:46 PM Stefan Roese <sr at denx.de wrote:

> Hi Chris,
>
> On 15.02.19 03:12, Chris Packham wrote:
> > Enable the hardware watchdog to guard against system lock ups when
> > running in the SPL or U-Boot. Stop the watchdog just before booting so
> > that the OS.
> >
> > Signed-off-by: Chris Packham <judge.packham at gmail.com>
> > ---
> >
> >   arch/arm/dts/armada-385-atl-x530-u-boot.dtsi |  4 ++
> >   board/alliedtelesis/x530/x530.c              | 48 ++++++++++++++++++++
> >   configs/x530_defconfig                       |  5 ++
> >   3 files changed, 57 insertions(+)
> >
> > diff --git a/arch/arm/dts/armada-385-atl-x530-u-boot.dtsi
> b/arch/arm/dts/armada-385-atl-x530-u-boot.dtsi
> > index 7074a73537fa..79b694cb84bc 100644
> > --- a/arch/arm/dts/armada-385-atl-x530-u-boot.dtsi
> > +++ b/arch/arm/dts/armada-385-atl-x530-u-boot.dtsi
> > @@ -11,3 +11,7 @@
> >   &uart0 {
> >       u-boot,dm-pre-reloc;
> >   };
> > +
> > +&watchdog {
> > +     u-boot,dm-pre-reloc;
> > +};
> > diff --git a/board/alliedtelesis/x530/x530.c
> b/board/alliedtelesis/x530/x530.c
> > index d7d1942fe686..1b22b6307cd2 100644
> > --- a/board/alliedtelesis/x530/x530.c
> > +++ b/board/alliedtelesis/x530/x530.c
> > @@ -7,6 +7,7 @@
> >   #include <command.h>
> >   #include <dm.h>
> >   #include <i2c.h>
> > +#include <wdt.h>
> >   #include <asm/gpio.h>
> >   #include <linux/mbus.h>
> >   #include <linux/io.h>
> > @@ -24,6 +25,10 @@ DECLARE_GLOBAL_DATA_PTR;
> >   #define CONFIG_NVS_LOCATION         0xf4800000
> >   #define CONFIG_NVS_SIZE                     (512 << 10)
> >
> > +#ifdef CONFIG_WATCHDOG
> > +static struct udevice *watchdog_dev;
> > +#endif
> > +
> >   static struct serdes_map board_serdes_map[] = {
> >       {PEX0, SERDES_SPEED_5_GBPS, PEX_ROOT_COMPLEX_X1, 0, 0},
> >       {DEFAULT_SERDES, SERDES_SPEED_5_GBPS, SERDES_DEFAULT_MODE, 0, 0},
> > @@ -75,6 +80,10 @@ struct mv_ddr_topology_map
> *mv_ddr_topology_map_get(void)
> >
> >   int board_early_init_f(void)
> >   {
> > +#ifdef CONFIG_WATCHDOG
> > +     watchdog_dev = NULL;
> > +#endif
> > +
> >       /* Configure MPP */
> >       writel(0x00001111, MVEBU_MPP_BASE + 0x00);
> >       writel(0x00000000, MVEBU_MPP_BASE + 0x04);
> > @@ -88,6 +97,17 @@ int board_early_init_f(void)
> >       return 0;
> >   }
> >
> > +void spl_board_init(void)
> > +{
> > +#ifdef CONFIG_WATCHDOG
> > +     int ret;
> > +
> > +     ret = uclass_get_device(UCLASS_WDT, 0, &watchdog_dev);
> > +     if (!ret)
> > +             wdt_start(watchdog_dev, 25000000ULL * 120, 0);
>
> This is CONFIG_SYS_TCLK? Would it make sense to use this macro
> instead here?
>
> > +#endif
> > +}
> > +
> >   int board_init(void)
> >   {
> >       /* address of boot parameters */
> > @@ -100,9 +120,37 @@ int board_init(void)
> >       /* DEV_READYn is not needed for NVS, ignore it when accessing CS1
> */
> >       writel(0x00004001, MVEBU_DEV_BUS_BASE + 0xc8);
> >
> > +     spl_board_init();
> > +
> >       return 0;
> >   }
> >
> > +void arch_preboot_os(void)
> > +{
> > +#ifdef CONFIG_WATCHDOG
> > +     wdt_stop(watchdog_dev);
> > +#endif
> > +}
>
> So you are not using the WDT in the OS (Linux)?
>

In our production systems we are. But I wanted to allow a kernel built
without the driver to boot and not have it reset on me (e.g. if I'm using a
debugger).


> > +
> > +#ifdef CONFIG_WATCHDOG
> > +void watchdog_reset(void)
> > +{
> > +     static ulong next_reset = 0;
> > +     ulong now;
> > +
> > +     if (!watchdog_dev)
> > +             return;
> > +
> > +     now = timer_get_us();
> > +
> > +     /* Do not reset the watchdog too often */
> > +     if (now > next_reset) {
> > +             wdt_reset(watchdog_dev);
> > +             next_reset = now + 1000;
> > +     }
> > +}
> > +#endif
>
> When I recently implemented the watchdog support the MT7688, I
> wondered why there is so much code necessary, that each board
> and platform needs to copy to get this real watchdog working.
> We should definitely rework this at some time, so make this more
> generic with less board specific code that could be shared.
>
> You don't need to change this here. I just wanted to express my
> thoughts, as I'm stumbling over this code again.
>

Yes it's no coincidence that this looks a lot like the code from one of the
turris boards. I was kind of surprised i needed to get the device and start
it but it kind of makes sense as I've decided that my board needs to do it
in the spl where as others might want to use it only to ensure that their
os boots succesfully.

Implementing watchdog_reset was a surprise but again generic code would
need to iterate over all instantiated UCLASS_WDT devices.

May be there's some middle ground with some mvebu specific code that can't
be completely generic but can avoid the repetition.


> Other than this (and your commit text change):
>
> Reviewed-by: Stefan Roese <sr at denx.de>
>
> Thanks,
> Stefan
>