Re: [PATCH v2 2/9] remoteproc: mediatek: Support hanlding scp core 1 wdt timeout

From: Mathieu Poirier
Date: Mon Aug 29 2022 - 13:40:32 EST


Hi Tinghan,

I have started reviewing this set and I expect comments to be spread out over a few
days. I will tell you when I am done.

Please see below for comments...

On Wed, Jun 08, 2022 at 04:35:46PM +0800, Tinghan Shen wrote:
> MT8195 SCP is a dual-core processor. The SCP core 1 watchdog timeout
> interrupt uses the same interrupt line of SCP core 0 watchdog timeout
> interrupt.
>
> Add support for handling SCP core 1 watchdog timeout interrupt in the
> SCP IRQ handler.
>
> Signed-off-by: Tinghan Shen <tinghan.shen@xxxxxxxxxxxx>
> ---
> drivers/remoteproc/mtk_common.h | 4 ++++
> drivers/remoteproc/mtk_scp.c | 27 ++++++++++++++++++++++++++-
> 2 files changed, 30 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/remoteproc/mtk_common.h b/drivers/remoteproc/mtk_common.h
> index ea6fa1100a00..73e8adf00de3 100644
> --- a/drivers/remoteproc/mtk_common.h
> +++ b/drivers/remoteproc/mtk_common.h
> @@ -54,6 +54,10 @@
> #define MT8192_CORE0_WDT_IRQ 0x10030
> #define MT8192_CORE0_WDT_CFG 0x10034
>
> +#define MT8195_SYS_STATUS 0x4004
> +#define MT8195_CORE0_WDT BIT(16)
> +#define MT8195_CORE1_WDT BIT(17)
> +
> #define MT8195_L1TCM_SRAM_PDN_RESERVED_RSI_BITS GENMASK(7, 4)
>
> #define SCP_FW_VER_LEN 32
> diff --git a/drivers/remoteproc/mtk_scp.c b/drivers/remoteproc/mtk_scp.c
> index 47b2a40e1b4a..3510c6d0bbc8 100644
> --- a/drivers/remoteproc/mtk_scp.c
> +++ b/drivers/remoteproc/mtk_scp.c
> @@ -212,6 +212,31 @@ static void mt8192_scp_irq_handler(struct mtk_scp *scp)
> }
> }
>
> +static void mt8195_scp_irq_handler(struct mtk_scp *scp)
> +{
> + u32 scp_to_host;
> +
> + scp_to_host = readl(scp->reg_base + MT8192_SCP2APMCU_IPC_SET);
> +
> + if (scp_to_host & MT8192_SCP_IPC_INT_BIT) {
> + scp_ipi_handler(scp);
> +
> + /*
> + * SCP won't send another interrupt until we clear
> + * MT8192_SCP2APMCU_IPC.
> + */
> + writel(MT8192_SCP_IPC_INT_BIT,
> + scp->reg_base + MT8192_SCP2APMCU_IPC_CLR);
> + } else {
> + if (readl(scp->reg_base + MT8195_SYS_STATUS) & MT8195_CORE1_WDT) {
> + writel(1, scp->reg_base + MT8195_CORE1_WDT_IRQ);
> + } else {
> + writel(1, scp->reg_base + MT8192_CORE0_WDT_IRQ);
> + scp_wdt_handler(scp, scp_to_host);

Why is scp_wdt_handler() not called when CORE1 signals a watchdog failure? If
this is the intended behaviour there is no way for anyone but you to know that
it is the case.

> + }
> + }
> +}
> +
> static irqreturn_t scp_irq_handler(int irq, void *priv)
> {
> struct mtk_scp *scp = priv;
> @@ -961,7 +986,7 @@ static const struct mtk_scp_of_data mt8192_of_data = {
> static const struct mtk_scp_of_data mt8195_of_data = {
> .scp_clk_get = mt8195_scp_clk_get,
> .scp_before_load = mt8195_scp_before_load,
> - .scp_irq_handler = mt8192_scp_irq_handler,
> + .scp_irq_handler = mt8195_scp_irq_handler,
> .scp_reset_assert = mt8192_scp_reset_assert,
> .scp_reset_deassert = mt8192_scp_reset_deassert,
> .scp_stop = mt8195_scp_stop,
> --
> 2.18.0
>