[Skiboot] [PATCH 1/2] Fix scom addresses for power9 core checkstop hmi handling.

[Andrew Donnellan]

On 03/07/17 15:08, Mahesh J Salgaonkar wrote:
> From: Mahesh Salgaonkar <mahesh at linux.vnet.ibm.com>
>
> Scom addresses for CORE FIR (Fault Isolation Register) and Malfunction
> Alert Register has changed for Power9. Fixup those while handling core
> checkstop for Power9.
>
> Without this change HMI handler fails to check for correct reason for
> core checkstop on Power9.
>
> Signed-off-by: Mahesh Salgaonkar <mahesh at linux.vnet.ibm.com>

Comments below

> ---
>  core/hmi.c |   61 +++++++++++++++++++++++++++++++++++++++++++++++++++++++-----
>  1 file changed, 56 insertions(+), 5 deletions(-)
>
> diff --git a/core/hmi.c b/core/hmi.c
> index 84f2c2d..8e1ebbd 100644
> --- a/core/hmi.c
> +++ b/core/hmi.c
> @@ -154,7 +154,13 @@
>  		((((1UL) << (t_count)) - 1) << ((s_id) * (t_count)))
>
>  /* xscom addresses for core FIR (Fault Isolation Register) */
> -#define CORE_FIR		0x10013100
> +#define P8_CORE_FIR		0x10013100
> +#define P9_CORE_FIR		0x20010A40
> +
> +/* xscom addresses for pMisc Receive Malfunction Alert Register */
> +#define P8_MALFUNC_ALERT	0x02020011
> +#define P9_MALFUNC_ALERT	0x00090022
> +
>  #define NX_STATUS_REG		0x02013040 /* NX status register */
>  #define NX_DMA_ENGINE_FIR	0x02013100 /* DMA & Engine FIR Data Register */
>  #define NX_PBI_FIR		0x02013080 /* PowerBus Interface FIR Register */
> @@ -215,6 +221,23 @@ static const struct nx_xstop_bit_info nx_pbi_xstop_bits[] = {
>  };
>
>  static struct lock hmi_lock = LOCK_UNLOCKED;
> +static uint32_t malf_alert_scom;
> +
> +static int setup_scom_addresses(void)
> +{
> +	switch (proc_gen) {
> +	case proc_gen_p8:
> +		malf_alert_scom = P8_MALFUNC_ALERT;
> +		return 1;
> +	case proc_gen_p9:
> +		malf_alert_scom = P9_MALFUNC_ALERT;
> +		return 1;
> +	default:
> +		prerror("HMI: %s: Unknown CPU type\n", __func__);
> +		break;
> +	}
> +	return 0;
> +}
>
>  static int queue_hmi_event(struct OpalHMIEvent *hmi_evt, int recover)
>  {
> @@ -244,6 +267,25 @@ static int queue_hmi_event(struct OpalHMIEvent *hmi_evt, int recover)
>  				num_params, (uint64_t *)hmi_evt);
>  }
>
> +static int read_core_fir(uint32_t chip_id, uint32_t core_id, uint64_t *core_fir)
> +{
> +	int rc;
> +
> +	switch (proc_gen) {
> +	case proc_gen_p8:
> +		rc = xscom_read(chip_id,
> +			XSCOM_ADDR_P8_EX(core_id, P8_CORE_FIR), core_fir);
> +		break;
> +	case proc_gen_p9:
> +		rc = xscom_read(chip_id,
> +			XSCOM_ADDR_P9_EC(core_id, P9_CORE_FIR), core_fir);
> +		break;
> +	default:
> +		rc = OPAL_HARDWARE;
> +	}
> +	return rc;
> +}
> +
>  static bool decode_core_fir(struct cpu_thread *cpu,
>  				struct OpalHMIEvent *hmi_evt)
>  {
> @@ -260,8 +302,7 @@ static bool decode_core_fir(struct cpu_thread *cpu,
>  	core_id = pir_to_core_id(cpu->pir);
>
>  	/* Get CORE FIR register value. */
> -	ret = xscom_read(cpu->chip_id, XSCOM_ADDR_P8_EX(core_id, CORE_FIR),
> -			 &core_fir);
> +	ret = read_core_fir(cpu->chip_id, core_id, &core_fir);
>
>  	if (ret == OPAL_HARDWARE) {
>  		prerror("HMI: XSCOM error reading CORE FIR\n");
> @@ -529,14 +570,24 @@ static void decode_malfunction(struct OpalHMIEvent *hmi_evt)
>  	uint64_t malf_alert;
>  	bool event_generated = false;
>
> -	xscom_read(this_cpu()->chip_id, 0x2020011, &malf_alert);
> +	if (!setup_scom_addresses()) {

Any way we can move the call so this only gets called once at startup?

> +		prerror("HMI: Failed to setup scom addresses\n");
> +		/* Send an unknown HMI event. */
> +		hmi_evt->u.xstop_error.xstop_type = CHECKSTOP_TYPE_UNKNOWN;
> +		hmi_evt->u.xstop_error.xstop_reason = 0;
> +		queue_hmi_event(hmi_evt, false);
> +		return;
> +	}
> +
> +	xscom_read(this_cpu()->chip_id, malf_alert_scom, &malf_alert);
>
>  	if (!malf_alert)
>  		return;
>
>  	for (i = 0; i < 64; i++) {
>  		if (malf_alert & PPC_BIT(i)) {
> -			xscom_write(this_cpu()->chip_id, 0x02020011, ~PPC_BIT(i));
> +			xscom_write(this_cpu()->chip_id, malf_alert_scom,
> +								~PPC_BIT(i));
>  			find_capp_checkstop_reason(i, hmi_evt, &event_generated);
>  			find_nx_checkstop_reason(i, hmi_evt, &event_generated);
>  			find_npu_checkstop_reason(i, hmi_evt, &event_generated);
>
> _______________________________________________
> Skiboot mailing list
> Skiboot at lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/skiboot
>

-- 
Andrew Donnellan              OzLabs, ADL Canberra
andrew.donnellan at au1.ibm.com  IBM Australia Limited